A little bit of a riddle and challenge for all you mathematicians out there! Could you use maths to create an equation to give each unit a point value set around stats? Basically, have 40k work like a spreadsheet. You add the units toughness, number of shots, weapon type, bs, ect... into some boxes and the maths converts them into points. Then you do the same to every unit and boom! Balance game.
Obviously extra rules and bonuses on top would throw everything off but for argument sake let's pretend they don't exist.
One issue would be how some stats synergies with each other. A wound on a toughness 1 model is not worth the same as a wound on a T8 one, same with saves
Attacks are worth more on models with better move/strength/WS.
While it might be possible to make a complex formula, it would not be easy, or even possible to set up so it could not be gamed/broken.
And that’s not even going into things like access to gear, or role. Marines pay for a statline they only use part of for each type of squad. But they do get some utility for being well rounded.
Yes, you could. It would take some work, but you could.
The problem would be extraneous effects, spells, stratagems, etc. Those would complicate everything. But a basic stat-line and basic weapon components could be costed out. The initial values would be somewhat arbitrary of course.
I have a character creation method for one of the games I sell. It's not nearly as complex, but each of the six stats has a simple cost associated with it - costs increased for more "important" stats (shooting, fighting, etc.). I use the same method when I create actual characters for the game.
As the depth increases the arbitrary decisions would increase - but as a basic premise, yes it's possible.
This is not remotely feasible. You can do a pretty good job just identifying some known well-priced units and then comparing what you've got to them and making small adjustments, but figuring out an all-inclusive formula for whatever oddball combination of stats you can come up with is just not worth attempting, especially if you're trying to include things like units where models can be equipped differently, etc.
There's also no one right way to do it. Relative costs for very different units are to some extent arbitrary. Tanks and light infantry are vulnerable to very different kinds of weapons, and so the durability of tanks relative to light infantry depends on the costs of anti-tank and anti-infantry weapons, as well as on the meta you're trying to cultivate.
Probably not across all units and factions, no. That's simply too much data for one formula.
GW has somewhat tried this with the "one wargear per codex" concept. However there's non-points based balance here, like the Rhino not being able to use twin lascannons.
lolman1c wrote: A little bit of a riddle and challenge for all you mathematicians out there! Could you use maths to create an equation to give each unit a point value set around stats? Basically, have 40k work like a spreadsheet. You add the units toughness, number of shots, weapon type, bs, ect... into some boxes and the maths converts them into points. Then you do the same to every unit and boom! Balance game.
Obviously extra rules and bonuses on top would throw everything off but for argument sake let's pretend they don't exist.
Almost certainly not. There's far too much contextual variability in 40k, which is why balance always seems so weird.
How do you balance taking a powersword vs a powermaul on a squad sergeant in a game involving titans? How do you value long range weapons that dont need LoS when tables can be barren and empty or densely packed with terrain? What about playing on a 4x4 table versus a 6x4? Etc ad nauseum.
There's a reason GW doesn't stick doggedly to some formula or set of equations. Most of it is kinda subjective feel and comparison to equivalents.
You can in theory do it. In practice the amount of work required will be vastly more than what it would take to do the conventional iterative playtesting process, where you start with a rough guess of a unit's point cost, and alternate playing test games and making point changes until you get it right.
What you’re describing is effectively a pricing model. How does your insurance company determine the price of insurance? Pricing model. Same concept here; anyone who says that a bound system like 40k is more complex than the real world is wrong. In the real world, financial analysts, actuaries, and other people typically associated with the finance sector will undertake the activity described.
If we take a few samples of units that are considered balanced as the baseline, we can check the effectiveness of our model against other units (by comparing the rule book price to our model’s guess). When we’re happy with the quality (read: accuracy of guesses), we can use the model to determine what the pricing should be on new and arbitrary stat blocks.
Sorcererbob wrote: What you’re describing is effectively a pricing model. How does your insurance company determine the price of insurance? Pricing model. Same concept here; anyone who says that a bound system like 40k is more complex than the real world is wrong. In the real world, financial analysts, actuaries, and other people typically associated with the finance sector will undertake the activity described.
If we take a few samples of units that are considered balanced as the baseline, we can check the effectiveness of our model against other units (by comparing the rule book price to our model’s guess). When we’re happy with the quality (read: accuracy of guesses), we can use the model to determine what the pricing should be on new and arbitrary stat blocks.
The difference is that in the real world, averages are good enough.
How do you price a 1-shot, 2+ to hit rerolling 1s, strength 60 100 damage weapon with Macro that ignores Void Shields? It's got about a 50% chance of killing a 3pt Brimstone Horror and about a 100% chance of killing a 6000 point Warlord Titan.
The average is going to be horrible (well, the average between 2pts and 6000 points is like 3001 points, right? Clearly this weapon should cost 3000 points), with a points efficiency of like 0.0013 against conscripts.
Sorcererbob wrote: What you’re describing is effectively a pricing model. How does your insurance company determine the price of insurance? Pricing model. Same concept here; anyone who says that a bound system like 40k is more complex than the real world is wrong. In the real world, financial analysts, actuaries, and other people typically associated with the finance sector will undertake the activity described.
If we take a few samples of units that are considered balanced as the baseline, we can check the effectiveness of our model against other units (by comparing the rule book price to our model’s guess). When we’re happy with the quality (read: accuracy of guesses), we can use the model to determine what the pricing should be on new and arbitrary stat blocks.
The big problem (beyond the degree to which you need different prices for weapons on units depending on a number of variables) is weighting range and movement properly, but I've got some ideas on that front.
I figure if you wanted to pick a few baseline units to use as a pivot you could theoretically figure out which units, when chosen as the pivot, require the fewest changes to the rest of the game, and then make your baseline assumption based on that.
Automatically Appended Next Post:
Unit1126PLL wrote: ...How do you price a 1-shot, 2+ to hit rerolling 1s, strength 60 100 damage weapon with Macro that ignores Void Shields? It's got about a 50% chance of killing a 3pt Brimstone Horror and about a 100% chance of killing a 6000 point Warlord Titan...
My theory on that is to figure out an average distribution of targets (based on some kind of archive of army lists) and then value the weapon based on how effective it is against a "normal" distribution. Not perfect (not only because the act of repricing things based on that model might change the distribution of army lists), but it's a place to start asking the question.
What is a "normal distribution" of units? Armies change all the time; heck, you'd have 6 months of data in this presumed "archive" from 8th edition in it right now and the top tournament lists have been 3 Lords of War fighting against 400 conscripts and Guilliman, and shifted considerably from the 5 Stormravens of yore.
I'm not sure you could make a "normal distribution" that would give that weapon a fair price ever.
If the question is "can it be done with the current 40K to achieve ultimate ballance" then the answer is no, because you have abilities that depend not on your stat or trait, but on the TARGETED ENEMY stats or traits. In such environment all you can do is to cost abilities on statistical utility in a "meta", i.e. how likely you are to face an enemy against whom such ability is usefull. But not only meta changes constantly, it is also localized, so a single world-wide point costs won't get rid of all FLGSs rage. This was more pronounced during 7th than now (how much is Haywire USR worth agains Tyranids?), but is still a thing. You would need something like constatnly updated "stock market" for point costs.
But if the question is "can such system be utlised to develop a game from scratch" then this is largely an open question. You can do such formulas for simple games but with added complexity such formulas get complicated quite quickly. Depending on how exactly you count "unique factions", 40K has more than 400 faction vs faction matchups alone. Then you must take into account, that 40K is based on a premise, that every faction should feel different enough from all others. Then add sandbox nature of listbuilding, and "environment conditions" such as table size, terrain setup style, winning conditions etc... Seeking mathematical ballance is pretty much the same goal as solving increasingly larger GO boards. Even in modern days of cheap computing power availability, 5x5 board was only solved in 2002, largest solved GO board now is 6x6 while standard game takes place on a 19x19. And this is a game with very, very simple rules and only one type of "unit" and very basic "unit interactions".
No you can not because there are too many variables involved. How do you value a weapon that ignores LOS? That depends on terrain which is not standardized. How do you value a unit with the conscript stat line? The value is different in an army like AM with good shooting, than it is for orks. It is also different depending on buffs available. Balance can only really be approximated with math, then would need adjusting through testing.
Assuming you factored in enough variables, you could probably get the basic cost of a given unit to be pretty well balanced using such a "unit builder." However, this would really only apply to unit's statline and weapons (assuming you didn't just rely on the codex prices for existing weapons).
The thing is, you have to start making a lot of assumptions in order to have any kind of pricing framework at all. X points might be a great cost for a given buffing aura on a given unit assuming you have Y of those units in range of the aura for most of the game. But does that aura become significantly better if you build a list around it?
Actually, here's a more specific example: A captain has an aura that lets marines reroll to-hit rolls while they're around him. Obviously this is much more potent when the captain is standing near missile launcher devs or centurions than when he's standing around regular bolter marines. So do you price the ability based on the assumption that the captain is going to be surrounded by devastators all game, or do you assume the player won't be optimizing quite as hard? If you assume the former, then you're basically overcharging any player that wants a captain to hang out with their tacs and assault marines or whatever.
Now you can set assumptions about a lot of these considerations. You can assume that a captain will frequently be within range of X units that will do about A, B, and C more wounds to GEQs, MEQs, and vehicles as a result of the aura. But that's still not really ideal and will fall apart to context.
A lascannon devastator might be perfectly balanced for its cost using mathematical formula, but it's woefully overpriced when fielded against wave after wave of gaunts.
tldr; You can use such a formula to balance out stats if you make certain assumptions, but it's very easy for any balance you might achieve to quickly become skewed if your assumptions don't happen to hold up for a given match.
You guys are making this more complex than it needs to be. Your pricing model is as strong as the inputs. You want it to consider buffs, line of sight, the impact of movement? Great! You can!
You simply need a way to record and quantify the actions of individual units / kitouts over thousands/millions/billions of games. The best way to do this is probably to create a game simulator with all of the modern rules, and allow the computer to play itself several billion times ala AlphaZero (Google’s chess bot).
Feed the data from that back into your pricing model, adjust pricing, rinse and repeat until everything is within a balance tolerance.
For those of you saying this is impossible, how does this approach fail? It’s infeasible, not impossible.
And that's not even looking into things that are simply incomparable
How do you price a +6" shooting range compared to +1 weapon strength?
+2" movement range, or +1 to WS.
Then the more unique abilities that bend the rules.
Shooting without line of sight.
The whole fly mechanic
Advance and charge
Free vertical movement
Etc
None of these can be mathed out to point efficiency like simple shooting comparisons are.
And to make matters worse, the game devs CAN'T know what kind of battlefield you are on.
Free vertical movement is worth a whole lot more in tables with a lot of multi level ruins, or outright cityscape than it is in a woodland table or a desert table.
Likewise ignoring line of sight is rather useless when there is hardly anything that blocks lines of sight.
All these situational abilities range from useless to amazing, depending on factors you can't possibly know during development/balancing as they are not constants nor predictable.
You either have them cost right when the ability is weak, and then it's situationaly OP, or cost right under the proper conditions and then they are usually useless.
The advantage video game RTS games has is that the realm of possibility is far more narrow, and devs can tell about how likely these conditions are to happen, or even enforce every match to have a given amount of these conditions that is enough to matter, but not enough to dominate.
Add the fact "army building" is far more reactive there, so you can afford to have your games have units who are only useful in odd conditions, as players (at least good ones) can tell by map and opponent when said conditions apply and THEN build these units.
You want all the oddities to be even remotely balancable, you need to change the way the game is played.
If lists are made only after you see the table and know what faction you are facing, they could balance all the "silver bullet" choices to be good only in given scenarios, and it will be fine, as the will only ever show up under said scenario.
Sorcererbob wrote: You guys are making this more complex than it needs to be. Your pricing model is as strong as the inputs. You want it to consider buffs, line of sight, the impact of movement? Great! You can!
You simply need a way to record and quantify the actions of individual units / kitouts over thousands/millions/billions of games. The best way to do this is probably to create a game simulator with all of the modern rules, and allow the computer to play itself several billion times ala AlphaZero (Google’s chess bot).
Feed the data from that back into your pricing model, adjust pricing, rinse and repeat until everything is within a balance tolerance.
For those of you saying this is impossible, how does this approach fail? It’s infeasible, not impossible.
with google and microsoft farming out AI tools to different companies now, it might not be as infeasable as you think for GW to work with one of them to apply that kind of AI approach to game analysis, the thing is they'd need at least two people reviewing the data that play the game to ensure the resulting stuff is actually "fun"
Given a particular formula for pricing everything in the game, how would you even know if the formula had a mistake?
Proving that the formula is correct is nearly impossible - I don't even think we could get past the step of defining what balance is. How many games need to be simulated (there's no way you're playing enough games by hand)? How do you know you covered enough army lists? How do you validate your game-playing AI?
What if I spotted a flaw, and I just wanted to prove the formula is wrong? What counts as a counter-example? Is showing that one list nearly always beats another enough? How many games do I have to play? Do joke lists count? How do you test if a list is a joke?
Let's say my formula is that every model costs zero points. How do you even prove that's unbalanced?
Elbows wrote: Yes, you could. It would take some work, but you could.
The problem would be extraneous effects, spells, stratagems, etc. Those would complicate everything. But a basic stat-line and basic weapon components could be costed out. The initial values would be somewhat arbitrary of course.
I have a character creation method for one of the games I sell. It's not nearly as complex, but each of the six stats has a simple cost associated with it - costs increased for more "important" stats (shooting, fighting, etc.). I use the same method when I create actual characters for the game.
As the depth increases the arbitrary decisions would increase - but as a basic premise, yes it's possible.
It honestly doesn't matter if they get 'perfect mathematical balance'.
Elbows wrote: Yes, you could. It would take some work, but you could.
The problem would be extraneous effects, spells, stratagems, etc. Those would complicate everything. But a basic stat-line and basic weapon components could be costed out. The initial values would be somewhat arbitrary of course.
I have a character creation method for one of the games I sell. It's not nearly as complex, but each of the six stats has a simple cost associated with it - costs increased for more "important" stats (shooting, fighting, etc.). I use the same method when I create actual characters for the game.
As the depth increases the arbitrary decisions would increase - but as a basic premise, yes it's possible.
It honestly doesn't matter if they get 'perfect mathematical balance'.
Players would still SAY it was imbalance.
The basic problem with the question (and with the responses I'm seeing) is that you can produce decent mathematical balance for a general case of a "40k game", but everything is still going to be matchup-dependent to enough of a degree that something that's "balanced" in a general sense may be unbalanced in the context of a specific game.
A mathematical approach isn't going to produce a perfect answer, but I do think it could produce something better than what we've got in a number of cases.
You can’t expect a perfect answer; many of the responders in this thread are seeking a beautiful unison of mathematically perfect moving parts. It won’t happen.
To those comments saying “you can’t compare X to Y”, yes you can. Is a bishop or a rook more valuable going into the chess end game? How do you compare moving diagonally to moving up and down? The answer is in the impact the piece has over a large number of games.
The original poster wanted to specifically omit complexities such as buffs. Within the confines of the original question, it’s entirely possible to create a model that will tell you which units are strong for their points and which units are weak for their points based on stat line online.
You cannot balance 40k mathematically due to so many subjective factors that go into the efficiency of things.
There are too many issues where stuff can be good or bad depending on the scenario. Yes Grapnel Launchers are great... when there is adequate terrain and LOS blocking buildings to take advantage of it. Open field? Completely worthless. Then you have the efficiency of things like weapons being costed differently for different units based on a mathematical vacuum. You can't price something correctly in a vacuum. In a vacuum, psykers with full smite were priced correctly. When used in an army and spammed multiple times, it became a problem. So now what? Do you price them where they should be assuming people won't abuse them or do you price them out of viability to prevent it? A balanced game is completely subjective and there is no objective answer. Chess is arguably one of the most balanced games in the history of gaming but is still arguably unbalanced in the favor of the first turn player.
Automatically Appended Next Post:
Sorcererbob wrote: You can’t expect a perfect answer; many of the responders in this thread are seeking a beautiful unison of mathematically perfect moving parts. It won’t happen.
To those comments saying “you can’t compare X to Y”, yes you can. Is a bishop or a rook more valuable going into the chess end game? How do you compare moving diagonally to moving up and down? The answer is in the impact the piece has over a large number of games.
The original poster wanted to specifically omit complexities such as buffs. Within the confines of the original question, it’s entirely possible to create a model that will tell you which units are strong for their points and which units are weak for their points based on stat line online.
Actually a rook is worth more since you can checkmate someone with two rooks alone and not two bishops alone. From what you're saying, there is an objective end game in chess that is static. 40k isn't. That's what people need to understand.
Hoodwink wrote: ...A balanced game is completely subjective and there is no objective answer. Chess is arguably one of the most balanced games in the history of gaming but is still arguably unbalanced in the favor of the first turn player...
There is an objective answer. The problem with balancing 40k is that it isn't a "game" in the sense that there's a defined set of conditions that exist and a clear game tree, it's a set of conditions used to construct a "game". The "game" starts with army selection, which (assuming both players have brought a list not knowing what the game is going to be) is a matter of guesswork. Comparisons to chess are sort of wonky for that reason.
Chess has a single objective condition in a static environment. 40k is not remotely close. How do you quantify the cost of a lascannon's range when you might deploy on the long side or short side. Its usefulness is absolutely less when deployed long side of the board. Its range outranges the table. You could just split the difference down the middle, but then you force people to play a static game. What if my area enjoys deploying on the long side of the board? Am I supposed to just say the extra cost associated with the extra negligible range is still balanced?
You can't perfectly balance a game like 40k that includes so many subjective environmental factors. You can achieve a very good balance. But it will never be a perfect balance. You can only balance the game for those who play within a set and defined list of parameters. For everyone else, it will be unbalanced.
Sorcererbob wrote: You guys are making this more complex than it needs to be. Your pricing model is as strong as the inputs. You want it to consider buffs, line of sight, the impact of movement? Great! You can!
You simply need a way to record and quantify the actions of individual units / kitouts over thousands/millions/billions of games. The best way to do this is probably to create a game simulator with all of the modern rules, and allow the computer to play itself several billion times ala AlphaZero (Google’s chess bot).
Feed the data from that back into your pricing model, adjust pricing, rinse and repeat until everything is within a balance tolerance.
For those of you saying this is impossible, how does this approach fail? It’s infeasible, not impossible.
The issue is chess has a limited set of possible moves in any given scenario, 40k does not. Further terrain will be different, there are millions of possible army combinations, different points values etc. so maybe given unlimited time it is possible, but with all the options you could probably have the thing play a billion games and not even fully test every possible unit in the space marine book, in every combination at multiple points values. There are something like 170 load outs for a 5 man tactical squad alone
Also let's not forget that units vary in efficiency based on the matchups. A T8+ model is absolutely better against an army like Grey Knights than melta Sisters of Battle. So if I have an army almost solely T8, I will absolutely do better on matchup alone against Grey Knights. Balance would dictate a 50/50 scenario. On the flip side, Grey Knights will do much better than melta Sisters against a Daemon army.
The thing about balancing games is that the only way to get as close to balance as possible is to create imbalances.
Sorcererbob wrote: What you’re describing is effectively a pricing model. How does your insurance company determine the price of insurance? Pricing model. Same concept here; anyone who says that a bound system like 40k is more complex than the real world is wrong. In the real world, financial analysts, actuaries, and other people typically associated with the finance sector will undertake the activity described.
If we take a few samples of units that are considered balanced as the baseline, we can check the effectiveness of our model against other units (by comparing the rule book price to our model’s guess). When we’re happy with the quality (read: accuracy of guesses), we can use the model to determine what the pricing should be on new and arbitrary stat blocks.
Sooooort of. The problem is that with 40k, you're essentially smashing multiple markets and segments into the same thing and treating it as one whole. At my work, I'm not trying to price aftermarket motors alongside brazing torches or refrigerant, I'm not trying to compare furnace filters and FLIR instruments, my pricing on heat pumps is completely unrelated to my pricing on tool bags, I'm not going to even bother trying to price out some things individually but rather by the ton, while other items may only be feasible as individual items, they're their own unique markets and segments and scales. With 40k, the game fundamentally mashes everything together, mixing scales of combat freely, and has widely variable contexts in which stuff may be employed.
Elbows wrote: Yes, you could. It would take some work, but you could.
The problem would be extraneous effects, spells, stratagems, etc. Those would complicate everything. But a basic stat-line and basic weapon components could be costed out. The initial values would be somewhat arbitrary of course.
I have a character creation method for one of the games I sell. It's not nearly as complex, but each of the six stats has a simple cost associated with it - costs increased for more "important" stats (shooting, fighting, etc.). I use the same method when I create actual characters for the game.
As the depth increases the arbitrary decisions would increase - but as a basic premise, yes it's possible.
No you can't. Any formula attempt to balance things out don't work out. Never been, never will. Or else it would have to be super complex as everything would have to depend on everything. Value of attack would have to depend on WS, S, weapons it can carry and survivability etc.
This is not true. Balance dictates a roughly 50/50 win rate over a large number of games against a diverse range of opposing lists, not that every single matchup be 50/50. For example, having a 60/40 matchup against one list is fine if it's balanced out by a 40/60 matchup against another list.
Automatically Appended Next Post:
Wyldhunt wrote: Actually, here's a more specific example: A captain has an aura that lets marines reroll to-hit rolls while they're around him. Obviously this is much more potent when the captain is standing near missile launcher devs or centurions than when he's standing around regular bolter marines. So do you price the ability based on the assumption that the captain is going to be surrounded by devastators all game, or do you assume the player won't be optimizing quite as hard? If you assume the former, then you're basically overcharging any player that wants a captain to hang out with their tacs and assault marines or whatever.
Easy answer: you price based on buffing the devastators. Balance can not and should not account for stupid plays, if a player wishes to use a poor strategy and put their captain in a less-effective location then that's their fault. Pricing units based on anything but their ideal usage means that there will be an overpowered use available for those players who wish to abuse it.
Automatically Appended Next Post:
Breng77 wrote: No you can not because there are too many variables involved. How do you value a weapon that ignores LOS? That depends on terrain which is not standardized. How do you value a unit with the conscript stat line? The value is different in an army like AM with good shooting, than it is for orks. It is also different depending on buffs available. Balance can only really be approximated with math, then would need adjusting through testing.
These variables are not an obstacle to math. You handle it the same way you handle it with playtesting, by evaluating a unit/list/whatever against a diverse metagame. As a very rough approximation you'd have something like point cost = (best-case value)x(probability of seeing best-case scenario) + (worst-case value)x(probability of seeing worst-case scenario), where each situation contributes a share of the point cost approximately equal to how frequently it occurs. And if you can figure out the relative frequency of each situation well enough to have meaningful playtesting then you can incorporate this into your mathematical model.
If we're talking about a simulation, even a casual simulation, you should want different units to be good in different contexts. The value of individual units should change based on those contexts. Terrain is the most obvious example. A busy board with lots of LOS blockers should devalue a tank like the Predator, which has long range weapons requiring LOS, and begins to miss mkre when on the move.
In order to achieve "complete mathematical balance" tactical and strategic context needs to be tossed. But if you do that, you don't have a wargame.
Probably not. And even trying to do it would almost certainly require scaling/sliding points for different game-sizes and multiplicator effects.
Mortarion at 400something points is a very different beast in a 500 or 1000 point game than he is in a 2000 point or, just for fun, a 5000 Apoc game. This probably no longer the case, but I remember reading some opinion piece or so during 5th or 6th edition that made a pretty good case on how quite a few issues could at least partly be explained by the "normal game" of 40K in the UK being 1500 points usually (and thus presumably what GW writers would play in most cases), where as the "normal game" in the US was often 1850 or 2000 points, which simply played out differently for the effectivenss of many units, the importance of the 1st turn, and a bunch of other things.
Re-rolls or other bonuses change the relative effectiveness of units and would probably need to be priced relative to the units that benefit from it in a given game, not a fixed point cost for a character that brings it.
And of course, differences in missions also bring out different things. All those (surely well-intentioned) ITC houserules and the habit of mixing Maelstrom and Eternal War missions popular in the US also really changes the dynamic of the game, usually placing (even) greater emphasis on board control and large blobs. It's surprising how good Primaris armies actually are/can be in UK tournaments, while also being utter trash in the US scene.
Sunny Side Up wrote: Probably not. And even trying to do it would almost certainly require scaling/sliding points for different game-sizes and multiplicator effects.
Mortarion at 400something points is a very different beast in a 500 or 1000 point game than he is in a 2000 point or, just for fun, a 5000 Apoc game. This probably no longer the case, but I remember reading some opinion piece or so during 5th or 6th edition that made a pretty good case on how quite a few issues could at least partly be explained by the "normal game" of 40K in the UK being 1500 points usually (and thus presumably what GW writers would play in most cases), where as the "normal game" in the US was often 1850 or 2000 points, which simply played out differently for the effectivenss of many units, the importance of the 1st turn, and a bunch of other things.
Yeah obviously anything you flat out can't take more than 1 period is going to have notable scalability issue. Even something as simple as tactical marines would have different impact based on point size if they were max 1 unit ever never mind big thing like primarch!
Similarly psykers increase or decrease in value based on points. Yeah the darn smite stays about constant(though varies in worthyness based on opposing army) but all the others...one -1LD spell in 500 pts is lot more impactfull than same in 5000 pts game.
Looks unfeasible.
I guess that balancing 40k is an NP-hard problem.
In the first place, it requires a statistical model with states and probabilities. This model would be rather large. I guess the studies/questions asked would be posed on a higher level so that statistical reasoning and inference would be necessary. Have a look into the hidden Markov model for speech recognition which is the simplest nontrivial Bayesian networks.
Probably, it is theoretically doable, but the complexity is too high for it to be practical. Certainly it would require much more than a simple spreadsheet But most importantly, there is the problem of the shifting meta. To be balanced, our system of balancing would need to take the meta in accord. The meta is different from place to place, but this theoretically could be dealt with. However, our very act of balancing will change the meta on which its balance is based, thus creating imbalance once more. Therefore, balancing in such a way is a self-defeating process, and true balance ultimately can never be achieved outside of a vacuum.
Iron_Captain wrote: Probably, it is theoretically doable, but the complexity is too high for it to be practical. Certainly it would require much more than a simple spreadsheet
Well sure. Formula just needs to account(not even comprehensive list) for one unit:
game size
scenario
exact terrain(shapes, sizes, density, any special rules etc)
exact opposition army
rest of YOUR army
stats of units keeping in mind value of stats depends on other stats, weapons etc
Of course when this formula then alters one unit it has habit of changing your army which in turn results in formula altering value of other units. GG
Formula sounds like hard work, how do you facture things into a formula that aren't quantifiable? Special Rules, Strategems, Chapter traits,Relics... etc.
I'm sure that technically you could mathematically balance 40k, but it would take way more time and effort than the classic methods.
I helped test and work on a game a friend was making that was heavily math based in the balancing department. He'd constantly track how many turns the game would run, when the first shots would fly, how much damage was done each round by groups of weapons, so on and so forth. The result was an oddly balanced point system, but it required an excel spreadsheet programmed with the formulas. Weapons would cost differently based on what arcs you put them in, the speed of the ship, and other secondary subsystems. It was interesting, but generally not feasible for mass consumption.
Better to just eyeball balance with 40k at first, then refine through playtesting and basic mathhammering.
Breng77 wrote: No you can not because there are too many variables involved. How do you value a weapon that ignores LOS? That depends on terrain which is not standardized. How do you value a unit with the conscript stat line? The value is different in an army like AM with good shooting, than it is for orks. It is also different depending on buffs available. Balance can only really be approximated with math, then would need adjusting through testing.
These variables are not an obstacle to math. You handle it the same way you handle it with playtesting, by evaluating a unit/list/whatever against a diverse metagame. As a very rough approximation you'd have something like point cost = (best-case value)x(probability of seeing best-case scenario) + (worst-case value)x(probability of seeing worst-case scenario), where each situation contributes a share of the point cost approximately equal to how frequently it occurs. And if you can figure out the relative frequency of each situation well enough to have meaningful playtesting then you can incorporate this into your mathematical model.
Sorry but they absolutely are an obstacle in math, the issue is that given the lack of limitations or standardization in the game the possible metagame against which you need to account for is simply too large. For instance what is the best case value? How do you determine the probability of facing that best case scenario. For instance a best case scenario for devestators might be having a ruin in their deployment zone that allows them to see the whole battlefield, then depending on their weapon choice would determine their best case opponent. How do you determine the probability of that happening, when there is no standardized terrain set up, or even existing recommendation about the level of terrain that should be used in the game. Further in a game that allows things like a force comprised entirely of T8+ 3+ save units, and one comprised entirely of T3 5+ save units there is simply too much to account for to truly have balance. In a game where skew lists are possible it is literally impossible to balance the game because it is not designed in a way that can be balanced. The assumption that you can determine the relative frequency to have meaningful playtesting is laughable. You cannot as the game is currently designed have meaningful playtesting because there is not enough time in the world to test out close to every possible match-up. The only way you can have anything resembling meaningful playtesting is that some options are garbage, and some are obviously good, so testers don't even need to touch some, and always use others. As I pointed out a tactical squad alone has hundreds of possible load outs, you really think anyone tests all of the combinations? And then does so along side every combination of other units in the book, or in the imperium faction?
For math to even begin to be usable GW would need to
1.) Define standard terrain layouts, or at least amount of table coverage, % of LOS blockage, % cover to an approximate level as for it to not be a large factor in unit effectiveness
2.) Define a points cost at which the game is supposed to be balanced (units are not going to be balanced at both 1000 points and 3000 points)
3.) Put more limitations on list building to curb skew lists- if you don't do this hard counters/bad match-ups to lists will basically always exist, because as a list designer you have no way of knowing exactly what you are going to face (especially if GW actually ever gets closer to cross faction balance, bad balance actually makes preparing for a meta easier as I only really need to account for the top several list builds)
4.)Reduce wargear options for units. Unless the idea is that one load out will be the "optimal load out" and everything else fluff or red herring having tons of wargear options makes balancing units much more difficult. A unit with few or no options will always be easiest to balance because you know exactly how it will perform against a variety of targets. Intercessors are easier to balance than tactical marines because you can make assumptions about how effective they will be against say a leman russ, where as a tactical squad with Melta gun, combi-melta, and multi-melta will perform quite a bit differently than one with flamer, combi-flamer and heavy bolter. And before you say that they have different targets in game, if a list of all russes exists both have to be balanced to face it.
5.) Either remove allies, or publish a statement saying factions will be balanced based on their entire ally set. Either marines are balanced standing on their own, or balanced assuming you are taking guard, you cannot have both be functional.
I still think the nay-sayers are being unimaginative.
One poster postulated that a unit had a certain value in a particular context, and another value in a different one. Good point. Would be impossible, for a given game board and mission, for a system to build two balanced forces?
The requirement isn’t that a system can be created that assigns a number of point to each thing that holds balance in every context, the requirement is that the game can be balanced.
I think I’ve found a mechanism to solve the context problem; context is just another input into the model. All this stuff takes is a little creativity.
With regard to the chess comments, I never said that chess was balanced. And of course rooks are better in the end game; we can prove this to be statistically true based on the results of thousands of games... just like I suggested we could for 40k (although with more games required to account for complexity).
Sorcererbob wrote: I still think the nay-sayers are being unimaginative.
One poster postulated that a unit had a certain value in a particular context, and another value in a different one. Good point. Would be impossible, for a given game board and mission, for a system to build two balanced forces?
The requirement isn’t that a system can be created that assigns a number of point to each thing that holds balance in every context, the requirement is that the game can be balanced.
I think I’ve found a mechanism to solve the context problem; context is just another input into the model. All this stuff takes is a little creativity.
With regard to the chess comments, I never said that chess was balanced. And of course rooks are better in the end game; we can prove this to be statistically true based on the results of thousands of games... just like I suggested we could for 40k (although with more games required to account for complexity).
Well if we throw out the requirement to have balance across all options in all situations then the game is balanced now. Given any mission and game board I can build 2 forces that are balanced to each other. What I cannot do is do so with variety, or without knowing the table and mission, or having the ability to design both forces ensure any semblance of balance. Sorry simply put you are wrong because there are too many contexts that impact balance and too much player choice involved. I think you are not giving enough credit to how much more complex 40k is than chess as far as balance is concerned. I think you would need near infinite games to achieve balance without instituting as I said previously a bunch of rules and restrictions to shrink the number of variables. You can never build a game where a unit is balanced at 500 points and 5000 points. It is akin to saying chess would have the same balance if you changed the board size, or which pieces were available to each side.
Sorcererbob wrote: I still think the nay-sayers are being unimaginative.
One poster postulated that a unit had a certain value in a particular context, and another value in a different one. Good point. Would be impossible, for a given game board and mission, for a system to build two balanced forces?
You realize there's infinite number of contexes? Every player has basically different terrain collection so it's impossible for GW to know how that affects. Then what about enemy army? That's something GW obviously can't factor in. Point value of your unit changes if opposing chaos army brings instead of 10 chaos marines 30 cultists. Or howabout 5 marines and 15 cultists...Then add in predator. Your units value changes again. Land raider instead? Again changes.
Good luck creating any sort of balanced formula. Only way to get balanced games would be fixed scenarios with fixed terrain(including shape) with fixed forces and looooooots of gametesting.
Stop. I think people are trying to dive “too deep” here and making to too complicated.
First off, a lot of people are saying you have to factor in terrain, scenario, mission, deployment, setup etc etc. You don’t need to factor in any of that in order to balance units against each other. Using the previously mentioned Revier Grappling hook as an example – it is an option you can spend points on currently. You spend these points prior to knowing the table setup (in theory) making them a strategic choice. They have nothing to do with the “balance” of the unit in terms of base costing.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Balance can be driven by setting a benchmark for various sets of things –
1. Basic stat line
2. Weapons
3. Abilities
For example, a Tactical Marine is the basic stat line for every infantry model in the Marine army. Once assigned an “initial” points value you can then move onto the next stage and determine weapon costs based on their stats and abilities etc etc. Obviously a lot of trial and error would need to be used to begin with. But, once you have your basic costing, you can then move onto another variant of Marine, a Terminator, for example. Using the standard Marine base cost, you can then add, or more points based on a perceived cost of additional stats. Maybe +1 wound is 10 points, +1 armour save is 5 points, and gaining an invuln save of x = 20 points. Of course these are just made up numbers right now, but it is the kind of basic logic you’d need to follow. All weapons would be pointed based off their stats alone, ignoring the “platform” they are on, like they are currently (i.e a lascannon has the same price regardless of whether it is a predator, dev unit, dreadnought or flyer using it).
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
It would be a massive amount of time and effort in order to point the first army, but then you'd have a model to quickly transpose to all the other armies. The issue then becomes tweaking the costs of each value to ensure everything is reasonable (i.e a Guard Infantry model isnt like 15 points).
Just to point something out about how many options there are an AM infantry squad has 1152 possible load outs. (2 vox options x 6 HW options x 6 Special weapon options x 4 sarge gun options x 4 sarge melee options), That is just a single unit. That means taking just 2 of those units gives you 1,327,104 possible different combinations.
So a single battalion with 2 company commanders(25 possible load outs) and 3 infantry squads has 955,514,880,000 possible combinations. So you need nearly 1 trillion games to even field all of the possible load outs, that is using 2 different units, and not taking opponents into account.
This is why I say that if you want to come close to balancing everything units need to have more limited options. Primaris Intercessors for example have 72 possible load outs (mostly squad size) (3 gun options x 6 squad size options x 2 sarge options x 2 grenade launcher options)
I had help with a doctorate in statistics and game theory who helped me write the formula we used for Azyr for AOS before the general's handbook killed the fan systems.
I will say with advanced mathematics that you can get pretty close to balancing a game, but not entirely.
Azyr used weighted averages. So the points cost of something was how well it performed against every possible combination and then averaged that number out.
The part where it failed were on the abstract synergies. Say dude gives +1 attack to every unit within 9" of him. You can't really account for this accurately because the actual mathematic increase will depend entirely on what units are within 9" of him.
If its a unit of scrubs with low grade attacks, the mathematic increase in performance will be X, but if you have a unit of high quality attacking models next to him, the performance will be X + Y. Coupled in that then you have to account for how many high grade units are in range at any given time.
Its impossible to do properly.
However you can get very close and then you have to have a fudge factor that covers the abstract abilities. Of course, no human beiing alive today could create a system that makes everyone happy. Someone will always gripe.
Even if a model was really worth 100 points, all it takes is for someone to play poorly, or get outplayed and that model gets tanked riight away and they are going to shoot your entire system down as being horrible and unbalanced because that 100 point model didn't return them 100 points liike they thought it should.
Kdash wrote: Stop. I think people are trying to dive “too deep” here and making to too complicated.
First off, a lot of people are saying you have to factor in terrain, scenario, mission, deployment, setup etc etc. You don’t need to factor in any of that in order to balance units against each other. Using the previously mentioned Revier Grappling hook as an example – it is an option you can spend points on currently. You spend these points prior to knowing the table setup (in theory) making them a strategic choice. They have nothing to do with the “balance” of the unit in terms of base costing.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Balance can be driven by setting a benchmark for various sets of things –
1. Basic stat line
2. Weapons
3. Abilities
For example, a Tactical Marine is the basic stat line for every infantry model in the Marine army. Once assigned an “initial” points value you can then move onto the next stage and determine weapon costs based on their stats and abilities etc etc. Obviously a lot of trial and error would need to be used to begin with. But, once you have your basic costing, you can then move onto another variant of Marine, a Terminator, for example. Using the standard Marine base cost, you can then add, or more points based on a perceived cost of additional stats. Maybe +1 wound is 10 points, +1 armour save is 5 points, and gaining an invuln save of x = 20 points. Of course these are just made up numbers right now, but it is the kind of basic logic you’d need to follow. All weapons would be pointed based off their stats alone, ignoring the “platform” they are on, like they are currently (i.e a lascannon has the same price regardless of whether it is a predator, dev unit, dreadnought or flyer using it).
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
It would be a massive amount of time and effort in order to point the first army, but then you'd have a model to quickly transpose to all the other armies. The issue then becomes tweaking the costs of each value to ensure everything is reasonable (i.e a Guard Infantry model isnt like 15 points).
Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
Automatically Appended Next Post:
auticus wrote: I had help with a doctorate in statistics and game theory who helped me write the formula we used for Azyr for AOS before the general's handbook killed the fan systems.
I will say with advanced mathematics that you can get pretty close to balancing a game, but not entirely.
Azyr used weighted averages. So the points cost of something was how well it performed against every possible combination and then averaged that number out.
The part where it failed were on the abstract synergies. Say dude gives +1 attack to every unit within 9" of him. You can't really account for this accurately because the actual mathematic increase will depend entirely on what units are within 9" of him.
If its a unit of scrubs with low grade attacks, the mathematic increase in performance will be X, but if you have a unit of high quality attacking models next to him, the performance will be X + Y. Coupled in that then you have to account for how many high grade units are in range at any given time.
Its impossible to do properly.
However you can get very close and then you have to have a fudge factor that covers the abstract abilities. Of course, no human beiing alive today could create a system that makes everyone happy. Someone will always gripe.
Even if a model was really worth 100 points, all it takes is for someone to play poorly, or get outplayed and that model gets tanked riight away and they are going to shoot your entire system down as being horrible and unbalanced because that 100 point model didn't return them 100 points liike they thought it should.
Correct me if I'm wrong but AOS seems to have fewer options for most units as well. Units are bought in blocks, and don't typically have much in the way of weapon options. In that case (except the mentioned synergy issue) it is much more attainable to work things out mathematically because there are fewer variables. IT seems like terrain would also have less of an impact because shooting is less of an issue.
Breng77 wrote: Just to point something out about how many options there are an AM infantry squad has 1152 possible load outs. (2 vox options x 6 HW options x 6 Special weapon options x 4 sarge gun options x 4 sarge melee options), That is just a single unit. That means taking just 2 of those units gives you 1,327,104 possible different combinations.
So a single battalion with 2 company commanders(25 possible load outs) and 3 infantry squads has 955,514,880,000 possible combinations. So you need nearly 1 trillion games to even field all of the possible load outs, that is using 2 different units, and not taking opponents into account.
This is why I say that if you want to come close to balancing everything units need to have more limited options. Primaris Intercessors for example have 72 possible load outs (mostly squad size) (3 gun options x 6 squad size options x 2 sarge options x 2 grenade launcher options)
Why are you trying to balance all the "combinations"? Why aren't you instead trying to balance the individual weapons, individual options and the cost of a basic model without any weapons included?? If a meltagun is balanced with a plasmagun in terms of points then the 1000's of combinations won't matter as everything is balanced vs everything else.
It might be impossible to balance entirely - there are limits to what you can do with a d6 system and discrete low point values.
But you could get it to a level where there were no objectively bad in all circumstance units and no objectively auto take in all circumstance units.
You could avoid the common issue of having two units in a roster that do the same thing, but one is clearly better in all situations than the other so one or both is never seen on tables.
Kdash wrote: Stop. I think people are trying to dive “too deep” here and making to too complicated.
First off, a lot of people are saying you have to factor in terrain, scenario, mission, deployment, setup etc etc. You don’t need to factor in any of that in order to balance units against each other. Using the previously mentioned Revier Grappling hook as an example – it is an option you can spend points on currently. You spend these points prior to knowing the table setup (in theory) making them a strategic choice. They have nothing to do with the “balance” of the unit in terms of base costing.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Balance can be driven by setting a benchmark for various sets of things –
1. Basic stat line
2. Weapons
3. Abilities
For example, a Tactical Marine is the basic stat line for every infantry model in the Marine army. Once assigned an “initial” points value you can then move onto the next stage and determine weapon costs based on their stats and abilities etc etc. Obviously a lot of trial and error would need to be used to begin with. But, once you have your basic costing, you can then move onto another variant of Marine, a Terminator, for example. Using the standard Marine base cost, you can then add, or more points based on a perceived cost of additional stats. Maybe +1 wound is 10 points, +1 armour save is 5 points, and gaining an invuln save of x = 20 points. Of course these are just made up numbers right now, but it is the kind of basic logic you’d need to follow. All weapons would be pointed based off their stats alone, ignoring the “platform” they are on, like they are currently (i.e a lascannon has the same price regardless of whether it is a predator, dev unit, dreadnought or flyer using it).
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
It would be a massive amount of time and effort in order to point the first army, but then you'd have a model to quickly transpose to all the other armies. The issue then becomes tweaking the costs of each value to ensure everything is reasonable (i.e a Guard Infantry model isnt like 15 points).
Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
Of course the “in-game” value of a lascannon is more if it can see everything, than if it can’t – but that ISNT a case of weapon and model balance, that is a case of strategy and game setup. You can’t attempt to “price in” a ruin vs a mountain vs nothing.
If a lascannon shoots a target it will do x amount of damage. If a lascannon doesn’t shoot because of a wall, it’s “average” damage per shot output doesn’t change. Saying that you need to “price in” players good, or bad positioning is stupid. That is an aspect of strategy and generalmanship, not model balance.
If a weapon has the “ability” to ignore LoS then obviously you would then point that weapon accordingly for that ability – not change the points of a weapon that cannot ignore LoS…
Winning a game is all about correct placement, correct target prioritisation and correct decision making in relation to the game at hand.
If a unit has a great stat line but works out badly in one given game setup, is it a case that the “unit is bad” or “the unit was USED badly” in the given game?
For example, a Shadowsword is FANTASTIC at deleting a tank a turn with it’s main weapon. But, if it gets killed first turn because you deploy wrong, or the other player got 1st turn and had 15 lascannons so it didn’t get to shoot at all, does that then make the Shadowsword a “bad unit” and therefore needs a points decrease? Or, is it just that the Shadowsword was played incorrectly? Or was it that the opponents list just happened to counter the Shadowsword?
Those instances aren’t a “balance” issue, they are a list building, deployment or genuinely unlucky issue. (the next game your opponent might have nothing to deal with the Shadowsword, so it completely destroyed their army… Suddenly it’s a good unit?)
Correct me if I'm wrong but AOS seems to have fewer options for most units as well. Units are bought in blocks, and don't typically have much in the way of weapon options. In that case (except the mentioned synergy issue) it is much more attainable to work things out mathematically because there are fewer variables. IT seems like terrain would also have less of an impact because shooting is less of an issue.
Shooting dominated the meta for most of AOS lifetime. The big things to min/max in AOS are mortal wounds (same as 40k) and high volume of ranged attacks. This is followed by high rending and high damage melee attacks.
Some of the units have just as many options as 40k units.
Same concept appliies. You have to get the unit baseline and then figure out how much a weapon increases that baseline and come to the weighted average against every type of defensive stat in the game.
Blacksails wrote: I'm sure that technically you could mathematically balance 40k, but it would take way more time and effort than the classic methods.
I helped test and work on a game a friend was making that was heavily math based in the balancing department. He'd constantly track how many turns the game would run, when the first shots would fly, how much damage was done each round by groups of weapons, so on and so forth. The result was an oddly balanced point system, but it required an excel spreadsheet programmed with the formulas. Weapons would cost differently based on what arcs you put them in, the speed of the ship, and other secondary subsystems. It was interesting, but generally not feasible for mass consumption.
Better to just eyeball balance with 40k at first, then refine through playtesting and basic mathhammering.
Now I 've run the same model on 40k and will say that the point values in 40k are closer to the median than AOS is, but 40k is roughly 13% unbalanced in a skew kind of way. (that is to say 13% of the game falls outside of a median and either underperforms or overperforms)
Can you guess what most players see in all their games
For example: conscripts when they were released with their busted rules had a 27% skew. Thats huge (what the median is will always be debated but if I'm balancing a system I try to keep the skew between 5 and 8%). The last time there was a skew like that that I saw was 5th edition draigo and paladin buddies (roughly a 21% skew). So 8th edition has released probably the most busted unit of all time in conscripts before their nerf. (I haven't run all the data for most of the editions, but my eldar starcannon spam tournament 3rd ed army was 17% over performing for their points cost)
Blacksails wrote: I'm sure that technically you could mathematically balance 40k, but it would take way more time and effort than the classic methods.
I helped test and work on a game a friend was making that was heavily math based in the balancing department. He'd constantly track how many turns the game would run, when the first shots would fly, how much damage was done each round by groups of weapons, so on and so forth. The result was an oddly balanced point system, but it required an excel spreadsheet programmed with the formulas. Weapons would cost differently based on what arcs you put them in, the speed of the ship, and other secondary subsystems. It was interesting, but generally not feasible for mass consumption.
Better to just eyeball balance with 40k at first, then refine through playtesting and basic mathhammering.
Pretty sure there were no auras in that game.
Kind of? They weren't auras with straight 're-roll dice!' or 'add +1 to everything', but you needed sensors to successfully target ships and you could design some ships to be 'command' ships which could do the bulk of the targeting for other ships.
Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points. But yes that was something I heard a lot of with Azyr. That it was boring because everything was too balanced and the same. I find people don't really want balance. They want structure and then they want to be able to effectively min/max within that structure and get their 2000 points to operate like its 3000 points.
Kdash wrote: Stop. I think people are trying to dive “too deep” here and making to too complicated.
First off, a lot of people are saying you have to factor in terrain, scenario, mission, deployment, setup etc etc. You don’t need to factor in any of that in order to balance units against each other. Using the previously mentioned Revier Grappling hook as an example – it is an option you can spend points on currently. You spend these points prior to knowing the table setup (in theory) making them a strategic choice. They have nothing to do with the “balance” of the unit in terms of base costing.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Balance can be driven by setting a benchmark for various sets of things –
1. Basic stat line
2. Weapons
3. Abilities
For example, a Tactical Marine is the basic stat line for every infantry model in the Marine army. Once assigned an “initial” points value you can then move onto the next stage and determine weapon costs based on their stats and abilities etc etc. Obviously a lot of trial and error would need to be used to begin with. But, once you have your basic costing, you can then move onto another variant of Marine, a Terminator, for example. Using the standard Marine base cost, you can then add, or more points based on a perceived cost of additional stats. Maybe +1 wound is 10 points, +1 armour save is 5 points, and gaining an invuln save of x = 20 points. Of course these are just made up numbers right now, but it is the kind of basic logic you’d need to follow. All weapons would be pointed based off their stats alone, ignoring the “platform” they are on, like they are currently (i.e a lascannon has the same price regardless of whether it is a predator, dev unit, dreadnought or flyer using it).
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
It would be a massive amount of time and effort in order to point the first army, but then you'd have a model to quickly transpose to all the other armies. The issue then becomes tweaking the costs of each value to ensure everything is reasonable (i.e a Guard Infantry model isnt like 15 points).
Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
Of course the “in-game” value of a lascannon is more if it can see everything, than if it can’t – but that ISNT a case of weapon and model balance, that is a case of strategy and game setup. You can’t attempt to “price in” a ruin vs a mountain vs nothing.
If a lascannon shoots a target it will do x amount of damage. If a lascannon doesn’t shoot because of a wall, it’s “average” damage per shot output doesn’t change. Saying that you need to “price in” players good, or bad positioning is stupid. That is an aspect of strategy and generalmanship, not model balance.
If a weapon has the “ability” to ignore LoS then obviously you would then point that weapon accordingly for that ability – not change the points of a weapon that cannot ignore LoS…
Winning a game is all about correct placement, correct target prioritisation and correct decision making in relation to the game at hand.
If a unit has a great stat line but works out badly in one given game setup, is it a case that the “unit is bad” or “the unit was USED badly” in the given game?
What if it is more than one game setup? And you keep saying player decision. Unless the player is setting up the terrain, it may not be their decision but instead the terrain that dictates the effectiveness for the weapon, which has nothing to do with generalship or strategy. If you are playing on a table where you can never use the maximum range of the weapon no matter what decision you make that matters, just as a table with more LOS blocking terrain favors units that ignore LOS versus one with little to no LOS blocking terrain. So how should that be costed if we don't know what the terrain will be? Do we cost it assuming it will be able to hide in a good position for its range, or do we assume that it will be a situational ability?
Same with the lascannon, if the table set up or mission dicates that it won't be able to fire at peak efficiency for say 3 turns of the game, why is it still worth the same as it would be in a game where it shoots 5 times?
Your method is the following - cost everything as if we are playing in its optimal condition, then if those conditions are uncommon those units now suck, and those that frequently have optimal conditions are great. Which is what we have right now.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
Breng77 wrote: Just to point something out about how many options there are an AM infantry squad has 1152 possible load outs. (2 vox options x 6 HW options x 6 Special weapon options x 4 sarge gun options x 4 sarge melee options), That is just a single unit. That means taking just 2 of those units gives you 1,327,104 possible different combinations.
So a single battalion with 2 company commanders(25 possible load outs) and 3 infantry squads has 955,514,880,000 possible combinations. So you need nearly 1 trillion games to even field all of the possible load outs, that is using 2 different units, and not taking opponents into account.
This is why I say that if you want to come close to balancing everything units need to have more limited options. Primaris Intercessors for example have 72 possible load outs (mostly squad size) (3 gun options x 6 squad size options x 2 sarge options x 2 grenade launcher options)
Why are you trying to balance all the "combinations"? Why aren't you instead trying to balance the individual weapons, individual options and the cost of a basic model without any weapons included?? If a meltagun is balanced with a plasmagun in terms of points then the 1000's of combinations won't matter as everything is balanced vs everything else.
Sure they do, because those things cannot be balanced in all situations unless they are identical. If they do different things, regardless of points costs then the combination of Melta + combi-melta + lascannon, has a different value than melta + combi-flamer + autocannon. Either all of these options are valid and balanced or, there is an optimal load out. Doing it by model ignores the impact of the unit as a whole. A single special weapon in a squad may have different value points wise to multiple special weapons.
Lets put it this way, doing it your way is going to end up as good enough, because you are never going to balance every possible option available.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
Which only works if there are limits on list building to avoid skew lists.
Not while maintaining unit differentiation and customizable armies. You're still going to run into the issue of Player A bringing only light machineguns while Player B brings only mainline tanks, or "All Snipers: The Army" having differing performance on Planet Bowling Ball vs. The Maze of Arbitrarily Tall Hedges.
In the context of a single product - e.g. the Death Guard half of Dark Imperium vs. the Space Marine half it should be achievable.
In a game where winning the single die roll for first turn can give a +40% win rate, good enough balance is probably fine. It's psychologically useful for players to be able to blame poor performance on luck or game design.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
And here is where part of the "omg fix this mess - the game is not balanced!1" is coming from. Tacticals may be well balanced mathematically but if you're only facing tanks, it doesn't seem so to you.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
And here is where part of the "omg fix this mess - the game is not balanced!1" is coming from. Tacticals may be well balanced mathematically but if you're only facing tanks, it doesn't seem so to you.
Yup, I really wish matched play had stricter army building rules that forced at least somewhat varied army construction. I'm not sure how exactly that would happen, but allowing people to take basically whatever they want means balanced play can never really happen. I'm not sure we are super far from decent balance in the game if people are taking varied lists. But with spam and skew lists we are anything but balanced.
Kdash wrote: Stop. I think people are trying to dive “too deep” here and making to too complicated.
First off, a lot of people are saying you have to factor in terrain, scenario, mission, deployment, setup etc etc. You don’t need to factor in any of that in order to balance units against each other. Using the previously mentioned Revier Grappling hook as an example – it is an option you can spend points on currently. You spend these points prior to knowing the table setup (in theory) making them a strategic choice. They have nothing to do with the “balance” of the unit in terms of base costing.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Balance can be driven by setting a benchmark for various sets of things –
1. Basic stat line
2. Weapons
3. Abilities
For example, a Tactical Marine is the basic stat line for every infantry model in the Marine army. Once assigned an “initial” points value you can then move onto the next stage and determine weapon costs based on their stats and abilities etc etc. Obviously a lot of trial and error would need to be used to begin with. But, once you have your basic costing, you can then move onto another variant of Marine, a Terminator, for example. Using the standard Marine base cost, you can then add, or more points based on a perceived cost of additional stats. Maybe +1 wound is 10 points, +1 armour save is 5 points, and gaining an invuln save of x = 20 points. Of course these are just made up numbers right now, but it is the kind of basic logic you’d need to follow. All weapons would be pointed based off their stats alone, ignoring the “platform” they are on, like they are currently (i.e a lascannon has the same price regardless of whether it is a predator, dev unit, dreadnought or flyer using it).
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
It would be a massive amount of time and effort in order to point the first army, but then you'd have a model to quickly transpose to all the other armies. The issue then becomes tweaking the costs of each value to ensure everything is reasonable (i.e a Guard Infantry model isnt like 15 points).
Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
Of course the “in-game” value of a lascannon is more if it can see everything, than if it can’t – but that ISNT a case of weapon and model balance, that is a case of strategy and game setup. You can’t attempt to “price in” a ruin vs a mountain vs nothing.
If a lascannon shoots a target it will do x amount of damage. If a lascannon doesn’t shoot because of a wall, it’s “average” damage per shot output doesn’t change. Saying that you need to “price in” players good, or bad positioning is stupid. That is an aspect of strategy and generalmanship, not model balance.
If a weapon has the “ability” to ignore LoS then obviously you would then point that weapon accordingly for that ability – not change the points of a weapon that cannot ignore LoS…
Winning a game is all about correct placement, correct target prioritisation and correct decision making in relation to the game at hand.
If a unit has a great stat line but works out badly in one given game setup, is it a case that the “unit is bad” or “the unit was USED badly” in the given game?
What if it is more than one game setup? And you keep saying player decision. Unless the player is setting up the terrain, it may not be their decision but instead the terrain that dictates the effectiveness for the weapon, which has nothing to do with generalship or strategy. If you are playing on a table where you can never use the maximum range of the weapon no matter what decision you make that matters, just as a table with more LOS blocking terrain favors units that ignore LOS versus one with little to no LOS blocking terrain. So how should that be costed if we don't know what the terrain will be? Do we cost it assuming it will be able to hide in a good position for its range, or do we assume that it will be a situational ability?
Same with the lascannon, if the table set up or mission dicates that it won't be able to fire at peak efficiency for say 3 turns of the game, why is it still worth the same as it would be in a game where it shoots 5 times?
Your method is the following - cost everything as if we are playing in its optimal condition, then if those conditions are uncommon those units now suck, and those that frequently have optimal conditions are great. Which is what we have right now.
Player decision should be everything.
Terrain does have an impact on a weapons ability to reach its “max game effectiveness”, but, that then comes down to, as a player – “do I chose to take a turn or 2 of less effective firepower in order to survive/get into position, or do I play more aggressively in order to attempt to obtain maximum effectiveness?” That is all about strategy and generalship.
If you go to a tournament (competitive or not) you won’t be setting up the terrain. This means you need to plan accordingly at the point of list building and decide, based on personal preference what to take. Maybe you take a lascannon on a devastator, or maybe you take a multi melta on a landspeeder. That is the point of balance, where each unit has a role and performs on par with the other options and it is the players decision on what to take and how to use it. Just because something might not be 100% effective in every single game, does not suddenly make it’s value less.
If you personally set up the table in advance you can then list tailor to suit that particular table, but then, I’d say that is pretty unfair and unsportsmanlike (unless your opponent has the same opportunity), which then, again, leads to it being a players choice on what to bring, not “well in this situation x does better than y, so therefore y should be less points than x”.
40k is a giant game of rock-paper-sissors. Some table setups favour one type of army, another setup completely nerfs it. This has an impact on each game, of course, but it does NOT have an impact on the balance of a unit. If we play on a heavy terrain table and I just take Basilisks and you only take slow moving long ranged heavy weapons, of course I’ll destroy you. But, if you take loads of deep-striking assault units you’d destroy me. But then, if I take all the infantry spam and heavy weapons I’d then destroy your assault spam. Just because 1 unit or 1 setup counters another unit, it does not mean that all the units are “unbalanced”.
It would be the same price… because at point of list building it DOES HAVE the same worth, regardless of whether it can then only fire 3 times or 5 times….
The “optimal” condition would be the same for every single unit – all shooting and assault is possible. That is the “standard” of balancing things.
If you, as a player then put a unit into a situation that is not optimal, that has nothing to do with how balanced a unit is or not. Sure, it affects the units ability to act at 100%, but it doesn’t make a unit “unbalanced”. If I use a unit of jump pack units and keep hiding behind terrain so you can’t shoot me, that doesn’t suddenly make Assault Marines “OMG SO BROKEN NERF PLEASE”, it just means I’m playing the terrain better than you and you’d need to adapt to it. Likewise, if you setup a non LoS gunline behind terrain, I’d have to find the best way to use my units to counter you.
What we have right now, is rock-paper-scissors, however, what we also have is a large amount of units that are massively out of balance when compared to everything else when everything is in “optimal” conditions.
Mathammer is all about “averages”. Averages are ALWAYS based on optimal conditions. Which is why, right now we are able to identify all the outliers and see people winning tournaments with them.
Automatically Appended Next Post:
Asmodai wrote: Not while maintaining unit differentiation and customizable armies. You're still going to run into the issue of Player A bringing only light machineguns while Player B brings only mainline tanks, or "All Snipers: The Army" having differing performance on Planet Bowling Ball vs. The Maze of Arbitrarily Tall Hedges.
Totally, but, that isn't an issue of balance, that is an issue of list building.
Now, in a narrative sense, these games can totally happen all the time, but then, as players you would agree the game setup, mission and points. It is also the case of - "well, tanks should always beat infantry with only machine guns anyway".
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
And here is where part of the "omg fix this mess - the game is not balanced!1" is coming from. Tacticals may be well balanced mathematically but if you're only facing tanks, it doesn't seem so to you.
Totally, 100%. Player perception of balance is massively different to actual balance. But you can't have both. And, at that point, you unfortunately need to ask the question of where it is the games fault, or both players fault.
koooaei wrote: Another problem with mathematically-based point costs is that eventually you end up with a game of potatoes. No matter what you take -everything's gona be equally effective. You could take an army of potatoes and they'd be as good as a space ship cause the game's perfectly balanced.
If you are seeking a balanced game then that's kind of the point. 2000 points is the same as your 2000 points.
I agree, that is the point of balance.
However, due to different weapon options, units would still be different in terms of their efficiency vs different units. For example, a Space Marine with a bolter might be as “balanced” as a Devestator with a lascannon, but the lascannon will do better vs tanks, while the bolter will do better vs chaff. That is then where strategy and the table come into play, as it’d force you to attempt to correctly use the units, or suffer an impact on overall efficiency.
And here is where part of the "omg fix this mess - the game is not balanced!1" is coming from. Tacticals may be well balanced mathematically but if you're only facing tanks, it doesn't seem so to you.
Yup, I really wish matched play had stricter army building rules that forced at least somewhat varied army construction. I'm not sure how exactly that would happen, but allowing people to take basically whatever they want means balanced play can never really happen. I'm not sure we are super far from decent balance in the game if people are taking varied lists. But with spam and skew lists we are anything but balanced.
I agree, that I feel that matched play should have stricter list building restrictions, but this does not affect overall balance of the game. If all the units were balanced and had identifiable roles, you should be able to take whatever you want. It then becomes a challenge of “can your opponent build a list to match, or can you play your army in all situations.”
@Kdash - Sorry but I still disagree, unit balance does not exist in a vacuum, terrain does matter to how effective certain units are. If every game is city fight a lascannon devastator squad is not worth its points compared to a more mobile options. Thus is not properly balanced. If terrain is typically a more balanced set up then it comes down to generalship but the value of unit abilities absolutely is dependent on what type of game you are playing, and unless you know this ahead of time the game is not balanced, because you cannot reasonably make informed choices about which units will be effective. By your argument range should not be factored into weapon cost because it does not matter how many times you are likely to fire said weapon when thinking about its cost. Which I'm sure you don't agree with. My point is that unless terrain is standardized to some degree it is not possible to accurately value units.
The same is true with list building restrictions. It is not possible for the game to be balanced. You say if units had identifiable roles than you should be able to take whatever you want. This simply isn't true. If I have 3 anti-infantry units and 3 anti-tank units and you bring 10 tanks, the game ceases to be balanced. Only when you can have some expectation of units in a variety of roles can you have balance. Unless missions are designed to punish skew builds in some way, but as long as tabling is a win condition this is unlikely.
For instance a Leman russ could be balanced, but 10 Leman russes is a very different balance consideration as it has the ability to render a good portion of your opponents units sub-optimal. The opposite is also true, if you bring 10 Russes and I bring 50 Lascannons, the game will not be balanced.
The meta may find some semblance of balance in this RPS environment but any individual game has a chance of being poorly balanced.
The value of units changes based on their strategic situation.
The ideal "balance" I think would be 'expensive, elite armies get beaten in headon fights with big, lumbering, slow armies. BUT, being small and elite, should have much better force concentration, and then with good play be able to take advantage of the terrain and setup to put pressure on only a single portion of the larger, slower force at a time, chewing it apart with superior 1v1 engagements over and over again, rather than trying to 1v30 all at once.'
I hope that makes sense.
Anyways, you can see how unit value changes. If you give the large lumbering horde army a bunch of ignores LOS shooting, then you've essentially removed the elite army's ability to use terrain to help with their force concentration and to chew up the enemy force. However, if there is too little terrain, then the elite army could shoot up the ignore LOSers, and since they cost points, the horde army paid an opportunity cost by not bringing tougher vehicles that could have endured the shooting.
Conversely the elite army could also be immobile in many ways (tied to Guilliman for example). This decreases their effectiveness, as the enemy, despite being a larger, slower, harder-to-deploy force, can control the engagement. It is a question of whether or not staying in place (in Guilliman's aura) increases efficiency enough to compensate for the loss of the strategic advantage of force concentration.
Unfortunately for the Space Marines in our example, they lack to the tools to concentrate force easily, so "trading away mobility to be in Guilliman's aura" isn't a meaningful trade because they didn't have that much better mobility to begin with. But I digress.
The value of units changes based on their strategic situation.
The ideal "balance" I think would be 'expensive, elite armies get beaten in headon fights with big, lumbering, slow armies. BUT, being small and elite, should have much better force concentration, and then with good play be able to take advantage of the terrain and setup to put pressure on only a single portion of the larger, slower force at a time, chewing it apart with superior 1v1 engagements over and over again, rather than trying to 1v30 all at once.'
I hope that makes sense.
Anyways, you can see how unit value changes. If you give the large lumbering horde army a bunch of ignores LOS shooting, then you've essentially removed the elite army's ability to use terrain to help with their force concentration and to chew up the enemy force. However, if there is too little terrain, then the elite army could shoot up the ignore LOSers, and since they cost points, the horde army paid an opportunity cost by not bringing tougher vehicles that could have endured the shooting.
Conversely the elite army could also be immobile in many ways (tied to Guilliman for example). This decreases their effectiveness, as the enemy, despite being a larger, slower, harder-to-deploy force, can control the engagement. It is a question of whether or not staying in place (in Guilliman's aura) increases efficiency enough to compensate for the loss of the strategic advantage of force concentration.
Unfortunately for the Space Marines in our example, they lack to the tools to concentrate force easily, so "trading away mobility to be in Guilliman's aura" isn't a meaningful trade because they didn't have that much better mobility to begin with. But I digress.
That would work if games were always on terrain heavy boards, and there was a distinct difference in mobility between those elite armies. Further the elite armies would need to be considerably stronger on a unit per unit basis and quickly be able to deal with the isolated forces. Asymmetric objectives could also aid in this vision. If my elite army is on an infiltrate and disrupt mission and thus have a different set of objectives this could make for more balanced games. When it largely comes down to just destroy the opponent certain unit roles become under valued.
Yes, that's true. I think the problems with elite armies currently are:
1) Not enough terrain - this is on the players 2) Not enough mobility - kind of. Some Elite armies actually do (e.g. Eldar), but some don't. Generally I put this on GW. It's all on GW if they say somewhere a 6x4 is the only way to play, because table space is the other issue that devalues (or overvalues, depending) mobility. 3) Relative strength: I think this is generally achieved. I think 1 model vs 1 model, a Terminator, Custodes, or Grey Knight is markedly, drastically superior to a Guardsman, Termagaunt, or Ork. 4) Asymmetric objectives would absolutely help, but are very difficult to do in a wargame; if it's just a PUG and not planned ahead, then this could go wrong very quickly, with people gaming the system to get whatever specific set of objectives they wanted. 5) Destroy the opponent is a bad mission type for elites for sure - in general, I don't think the U.S. would send Delta Force out there to destroy a whole Russian Federation tank company in a straight fight, even if their training and equipment cost the same (e.g. "points cost").
That's part of the problem I see, actually. People bring e.g. 10 terminators and a bunch of vanguard vets to fight an Imperial Guard infantry company supported by two tank platoons and some scout recon elements in an open-field battle, and then wonder why they lost...
Yes, that's true. I think the problems with elite armies currently are:
1) Not enough terrain - this is on the players
2) Not enough mobility - kind of. Some Elite armies actually do (e.g. Eldar), but some don't. Generally I put this on GW. It's all on GW if they say somewhere a 6x4 is the only way to play, because table space is the other issue that devalues (or overvalues, depending) mobility.
3) Relative strength: I think this is generally achieved. I think 1 model vs 1 model, a Terminator, Custodes, or Grey Knight is markedly, drastically superior to a Guardsman, Termagaunt, or Ork.
4) Asymmetric objectives would absolutely help, but are very difficult to do in a wargame; if it's just a PUG and not planned ahead, then this could go wrong very quickly, with people gaming the system to get whatever specific set of objectives they wanted.
5) Destroy the opponent is a bad mission type for elites for sure - in general, I don't think the U.S. would send Delta Force out there to destroy a whole Russian Federation tank company in a straight fight, even if their training and equipment cost the same (e.g. "points cost").
That's part of the problem I see, actually. People bring e.g. 10 terminators and a bunch of vanguard vets to fight an Imperial Guard infantry company supported by two tank platoons and some scout recon elements in an open-field battle, and then wonder why they lost...
Before I started messing with actual rules, my 7th ed games of Eldar vs Tyranids were ballanced exactly in such way - I designed games to be such asymmetric encounters with objectives and terrain supporting/hindering imballances. But this is good solution for close friends/people who have same approach to 40K (and understand the system well enough, as you must manage all aspects of the game: force selection, terrain layout, game lenght, winning conditions, everything) but it's completely unsuited for PUG (unless you happen to stumble upon a like-minded person) or tournament environment...
I’m not arguing that terrain doesn’t play a part in how effective a unit is in a game – I agree with you on that. What I am saying, is that the terrains effects shouldn’t have an impact on cost. If we start going down that route you end up with a dozen different points cost for the same unit, all depending on the specific terrain setup of a given table. Terrain should impact strategy, not define balance. This is simply because every table is different and the level of terrain is different. A common ground has to be found as a starting point. If a devastator unit gets 1 turn of shooting game 1, 3 turns game 2 and 5 turns game 5, how do you define its effectiveness? Base everything off only getting 3 turns a game? Or do you assign different costs for each of the 3 games? Of which, you won’t fully know until after the game. You also have to take into account a unit’s “threat” zone, even if it isn’t firing. In your cities table example, sure, the lascannons might not be able to shoot every turn, but how do you put a cost on their ability to control an entire section of the table, by denying it to your opponent if he wants his tanks to survive? You might not have shot that turn, but, suddenly you’ve made your opponent spend another turn of moving and not shooting in an attempt to survive.
Also, I’ve not said anything about a weapon’s range not being factored into its cost. Range is one of the factors that would be used to determine a weapons cost. Inferno Pistol, Meltagun and Multi-melta are all examples of this, where they have the same damage output, but would be different costs due to range.
Unit balance would start to impact on skew builds though. Currently 3 anti-infantry and 3 anti-tank vs 10 tanks is skewed, I agree, but, is that because anti-tank is too expensive right now, or tanks too cheap? Let’s say you can now suddenly get 1.5 infantry anti-tank units per 1 tank, while still having the same number of units of infantry anti-infantry units due to the points changes. You’d then be looking at 15 anti-tank units to the 10 tanks. There wouldn’t then be a problem in regards to being able to take whatever you want. Unless everyone out there builds TAC lists every single game, you will always, in your view, have an “unbalanced game”, even if, when you look at the bigger picture, the game is balanced. You can’t micro manage every game.
@Unit I know and accept that a unit’s value is different in every single game, and often does fluctuate within the game as well. This is down to matchup though, not overall balance. If I come up against a horde army, but have 70% of my army setup to take out tanks and flyers, 70% of my army will find its value decrease, and the other 30% potentially increase. Likewise, if I come up against a load of tanks with only a small amount of infantry, the 70%’s value increases, up until the point where all the enemy’s tanks have been destroyed. At that point, 70% of my army instantly loses its value. Again, I would argue that this is more to do with list building than trying to balance the game overall. Taking 70% anti-tank SHOULD be able to be countered by hordes of infantry, rather than have the cost of the anti-tank adjusted just because x% of the time they are in a battle where they have nothing “efficient” to shoot at.
Choosing a set playstyle when you build your list is entirely your choice – you should always go into list building with a plan/idea. But, expecting the cost of units to change because of that choice, or because of terrain or because of matchup is just wrong. Everything is risk and reward. It would be like only taking a set list to a local store and then only picking 1 table to play on because it suits you, then only accept games from people who had lists you considered “balanced” for your army.
In my view, this is the arena of Narrative and Open play, not matched play (which should be the driver for balance). Narrative and Open would allow you to “change the rules” as such for those games where you want to field 3k points of infantry vs 2k points of tanks for example as attempts to have -fun- games with super skewed lists. It would be like me deciding Baneblades had to go up in cost so you could no longer get 3 in a 2k list. Now, this would make Unit sad, but would make lots of people in his club happy, but is not the answer to balance.
Balance must be looked at outside of individual games, and instead across the entire picture. If someone wants to skew their list completely to 1 style, then that is fine, but they should also understand the potential issues they will come up against by doing so.
So how is the value being different in every single game allow for mathematical balance?
You say thats "matchup" not "balance" but I would say that "unbalanced matchups" are undesirable, and if they exist, then the system is no different than it is now, because people would still run into unbalanced lists and have no fun.
I'm not really sure how you can divorce "matchup" and "balance" from eachother - could you explain that to me?
Which I think is one reason they went down the rabbit hole of making terrain relatively useless. It doesn't have a large impact in games, therefore the balancing mechanism that they chose to use wouldn't be as impacted by it.
auticus wrote: Which I think is one reason they went down the rabbit hole of making terrain relatively useless. It doesn't have a large impact in games, therefore the balancing mechanism that they chose to use wouldn't be as impacted by it.
I think the problem though is they didn't touch the biggest terrain issue - LOS blockers. Regardless of the actual "cover" mechanics, requiring LOS for shooting attacks means that a battlefield that looks like a Zone Mortalis board is fundamentally going to hurt high-powered shooting units more than a board that looks like this.
In the former, charging units and shooting units without LOS are going to reign supreme, while regular shooting units like predator annihilators or leman russ tanks are just waiting, unable to shoot, until they're touched in combat and suddenly also unable to shoot.
In the latter, ground-pounding assault units and units that don't require LOS are really bad compared to Predator Annihilators and Leman Russes, because they lack durability and generally firepower compared to the other two - and there will ALWAYS be LOS.
You can totally make a formula to make unit characteristic balanced compared to their durability and ability to do damage. Obviously things like toughness wounds and saves scale differently as they combine together - this is easy to determine in regards to math. Things like mobility and weapon range are harder to balance - That kind of stuff would need to be play tested to determine the value of the mobility.
Yes, that's true. I think the problems with elite armies currently are:
1) Not enough terrain - this is on the players
2) Not enough mobility - kind of. Some Elite armies actually do (e.g. Eldar), but some don't. Generally I put this on GW. It's all on GW if they say somewhere a 6x4 is the only way to play, because table space is the other issue that devalues (or overvalues, depending) mobility.
3) Relative strength: I think this is generally achieved. I think 1 model vs 1 model, a Terminator, Custodes, or Grey Knight is markedly, drastically superior to a Guardsman, Termagaunt, or Ork.
4) Asymmetric objectives would absolutely help, but are very difficult to do in a wargame; if it's just a PUG and not planned ahead, then this could go wrong very quickly, with people gaming the system to get whatever specific set of objectives they wanted.
5) Destroy the opponent is a bad mission type for elites for sure - in general, I don't think the U.S. would send Delta Force out there to destroy a whole Russian Federation tank company in a straight fight, even if their training and equipment cost the same (e.g. "points cost").
That's part of the problem I see, actually. People bring e.g. 10 terminators and a bunch of vanguard vets to fight an Imperial Guard infantry company supported by two tank platoons and some scout recon elements in an open-field battle, and then wonder why they lost...
1.) Somewhat on the players, but if GW gave some guidance as to what was expected it would be good. They used to at least have the 25% guideline. But even just some pictures of example tables saying, " this is the kind of terrain we tested the game on" would be nice.
2.) Yup table space is an issue as points costs have gone down, mobility has become less useful.
3.) I think 1 model vs 1 model doesn't work so well, when say a terminator costs as much as 10 guardsman. So a terminator should be able to deal with say 2 or 3 guardsman easily if he gets the drop on them.
4.) Other games manage it. malifaux for instance. If you randomly generate the objectives and have say 4 or 5 to choose from it works ok. If you look at NOVA or ITC secondaries this is a good example of how some of this could be done.
You think terrain is useless? Okay. It has tremendous impact on how the game plays out.
No not really. Not unless you custom build line of sight blocking terrain, to which you risk your community squaking that you are modeling for advantage and that you should only use GW kits for terrain so that they can continue having full line of sight everywhere.
Terrain doesn't do much of anything except serve up as decoration the vast majority of the time.
Its role is largely ornamental with the occasional mechanical benefit.
Mobility is largely a waste these days. You don't need it when you can just alpha puke onto your opponent.
I've played through dozens of games with standard GW terrain as the rule, and line of sight is largely not inhibited anywhere. It doesn't slow you down. It doesn't cause deep strike mishaps anymore. GW kits don't block line of sight without heavy modifications or houserules. It doesn't influence the game largely in any direction and you could get similar results from most of the games I've seen with no terrain at all on a flat table.
If you could give a list of benefits of what terrain offers besides being ornamental, I'd great love to hear what you have to offer, because right now when compared with proper wargames, the terrain in 40k doesn't do much of anything and I think that is very much intentional.
1) I suppose. The problem is that tournaments will throw all of it out and do their own thing anyways, and casual players don't have access to a deep terrain pool, so they'll see "NEAT!" and do their own thing anyways. Plus, it seems pedantic, but you could have the exact same ruins in the exact same place, but if one has a window it's immediately wayyy different in a very important way.
2) Yes. I think table space is actually the largest problem with 40k right now, but that's my opinion and I don't really have any data to back it up.
3) I think a Terminator can, rather easily, deal with 2 or 3 guardsmen in the current standing. Storm bolter + powerfist basic terminator deep striking will kill 1.19 guardsmen in shooting, and then (if he gets the charge) .83 in combat with his power fist. So that's just over 2 guardsmen right there (with the difference between it being 2 and 3 guardsmen being one 4+ to hit roll on a powerfist strike).
4) Malifaux is infinitely smaller. Part of the issue with 40k is scale - it is actually possible in 40k to (forgive the bad analogy but it's along the lines of the sheer diversity 40k has) end up with Delta Force (500 points of adeptus custodes) fighting an open-field battle against a Typhoon-class submarine (a Marauder Destroyer) in the middle of Tenochtitlan (Feudal Word city set on a lake) in 1503 (arbitrary year, sorry). It's ... a bit hard to just come up with asymmetric objectives.
Kdash wrote: I’m not arguing that terrain doesn’t play a part in how effective a unit is in a game – I agree with you on that. What I am saying, is that the terrains effects shouldn’t have an impact on cost. If we start going down that route you end up with a dozen different points cost for the same unit, all depending on the specific terrain setup of a given table. Terrain should impact strategy, not define balance. This is simply because every table is different and the level of terrain is different. A common ground has to be found as a starting point. If a devastator unit gets 1 turn of shooting game 1, 3 turns game 2 and 5 turns game 5, how do you define its effectiveness? Base everything off only getting 3 turns a game? Or do you assign different costs for each of the 3 games? Of which, you won’t fully know until after the game. You also have to take into account a unit’s “threat” zone, even if it isn’t firing. In your cities table example, sure, the lascannons might not be able to shoot every turn, but how do you put a cost on their ability to control an entire section of the table, by denying it to your opponent if he wants his tanks to survive? You might not have shot that turn, but, suddenly you’ve made your opponent spend another turn of moving and not shooting in an attempt to survive.
Also, I’ve not said anything about a weapon’s range not being factored into its cost. Range is one of the factors that would be used to determine a weapons cost. Inferno Pistol, Meltagun and Multi-melta are all examples of this, where they have the same damage output, but would be different costs due to range.
Unit balance would start to impact on skew builds though. Currently 3 anti-infantry and 3 anti-tank vs 10 tanks is skewed, I agree, but, is that because anti-tank is too expensive right now, or tanks too cheap? Let’s say you can now suddenly get 1.5 infantry anti-tank units per 1 tank, while still having the same number of units of infantry anti-infantry units due to the points changes. You’d then be looking at 15 anti-tank units to the 10 tanks. There wouldn’t then be a problem in regards to being able to take whatever you want. Unless everyone out there builds TAC lists every single game, you will always, in your view, have an “unbalanced game”, even if, when you look at the bigger picture, the game is balanced. You can’t micro manage every game.
The issue is then tanks are a poor choice if your opponent gets 15 presumably effective anti-tank units to 10 tanks, and get other units, while the tank player only has tanks. Why take tanks in this case. The point of restricting army composition as part of balance is that it to an extent forces TAC lists because that is where the game can be balanced.
On the terrain angle all I'm saying is that GW needs to provide guidelines around what type of terrain was used for testing/balancing units. You cannot design units that are both worth while in dense terrain, and equally worth while in open terrain unless all units function the same way. IF there is no guideline then tournament standards decide what the effective units are by dictating what standard terrain set ups look like. To me this is the difference between matched play and narrative play for balance. GW should do something like, "for a balanced game we suggest that you play 1500 points on a 6 x4 table with ~33% terrain coverage, at least 1/3rd of this terrain should block LOS." Then provide some examples of balanced terrain layouts. This gives players an idea of what type of game was used to balance the units.
To the range question, you said it doesn't matter how many times a weapon shoots during the game. If that is true range is largely irrelevant other than as a defensive measure.
Regarding the initial question: There are just too many non-linear equations in several varialbes to balance it!
Even finding the right cost of a model only by its stats is hard, but maybe possible. But as soon as you consider category (with limits to detachments), special (army) rules or possible synergies with other units, it really gets sophisticated.
The problem is usally that you really have to price the best combination with other army elements in which a unit can appear. (WH40k is not really good in solving this problem...) It's often about saturation of certain elements that grant non-linear increasing strength: while most armies can handle a few tough units and a few horde units, they often fail if some saturation is met. But that's where (local) meta often gets into play. But this requires all armies to have certain tools to deal with stuff... Really not easy.
I've been thinking on this a little bit in terms of the T9A-project (i.e. the most balanced version of fantasy battles)...
Conclusion: It's better to make good guesses and do price adjustments by extensive playtesting... and be restrictive with options!
Unit1126PLL wrote: 1) I suppose. The problem is that tournaments will throw all of it out and do their own thing anyways, and casual players don't have access to a deep terrain pool, so they'll see "NEAT!" and do their own thing anyways. Plus, it seems pedantic, but you could have the exact same ruins in the exact same place, but if one has a window it's immediately wayyy different in a very important way.
2) Yes. I think table space is actually the largest problem with 40k right now, but that's my opinion and I don't really have any data to back it up.
3) I think a Terminator can, rather easily, deal with 2 or 3 guardsmen in the current standing. Storm bolter + powerfist basic terminator deep striking will kill 1.19 guardsmen in shooting, and then (if he gets the charge) .83 in combat with his power fist. So that's just over 2 guardsmen right there (with the difference between it being 2 and 3 guardsmen being one 4+ to hit roll on a powerfist strike).
4) Malifaux is infinitely smaller. Part of the issue with 40k is scale - it is actually possible in 40k to (forgive the bad analogy but it's along the lines of the sheer diversity 40k has) end up with Delta Force (500 points of adeptus custodes) fighting an open-field battle against a Typhoon-class submarine (a Marauder Destroyer) in the middle of Tenochtitlan (Feudal Word city set on a lake) in 1503 (arbitrary year, sorry). It's ... a bit hard to just come up with asymmetric objectives.
1.) Guidance for balance would still be good, and tournaments (if the game was actually balanced) would be likely to make use of such guidance. Right now because the game is not balanced tournaments do their own thing.
3.) That does not sound all that easy, since he is not very likely to make the charge.
4.) Not really, you just need to provide options at the outset for objectives so players can make a choice about which they think they can achieve. Tournaments are doing this right now.
1) Fair enough, though I still think making hard & fast "rules" for terrain placement will be a bad idea, while making "suggestions" will end up like most of the "suggestions" in the current rule-book i.e. ignored.
3) Depends on if he is a Black Templar or not! There are ways to ensure he gets the charge - Chapter Tactics & Command Re-Rolls come to mind, at least in the case of loyalist terminators - but now we have to expand out to force multiplication and stuff like that. Stuff that has costs (CPs or Points or even Opportunity Costs) - but generally, I do think most elite units can 1v1 or even 1v2 or 1v3 most horde units, model per model.
4) This is true, but still kind of gamey. Like the difference between "First Blood" and "First Strike" at NOVA was no contest at all - "first strike" was always the better choice. It's 'false choice' in many ways. But yes, having this in the game would improve it dramatically I think.
For my money I think you need hard and fast rules for terrain. Back in the olden days of 40k and whfb we had a terrain chart and you rolled D3 pieces of terrain per table quarter. Thats how everyone did their tables and built armies accordingly.
"Suggestions" are always ignored. People don't want suggestions. They want hard and fast rules.
Without hard and fast rules you get arguments and salty butthurt if mr shooty player plays mr assault player who drops down customized large builddings that block LOS because no one can claim the rules disallow or allow it.
auticus wrote: For my money I think you need hard and fast rules for terrain. Back in the olden days of 40k and whfb we had a terrain chart and you rolled D3 pieces of terrain per table quarter. Thats how everyone did their tables and built armies accordingly.
"Suggestions" are always ignored. People don't want suggestions. They want hard and fast rules.
Without hard and fast rules you get arguments and salty butthurt if mr shooty player plays mr assault player who drops down customized large builddings that block LOS because no one can claim the rules disallow or allow it.
d3 pieces of terrain per table quarter doesn't help anything. I'm fairly sure the "bowling ball planet" table I linked above had like 1-3 terrain pieces per quarter.
I think you'd essentially have to pull a Starcraft (where's martel?) and pre-built boards for other people to play on, and then balance the armies around the selection of available "maps" like Starcraft does (e.g. this "map" is better for Zerg while this other is better for Protoss). But that takes the sheen off of playing such a neato wargame as Warhammer in general, and I am strongly against the idea for perhaps obvious reasons.
Starcraft puts a lot of time into building boards. The dreaded lost temple board was pulled from the rotation and redesigned to not be so imba for terran.
Martel732 wrote: Starcraft puts a lot of time into building boards. The dreaded lost temple board was pulled from the rotation and redesigned to not be so imba for terran.
Yay there he is!
And yeah that's my point. The developers themselves built them, they still sometimes make mistakes, and there's no real "fun" in having a varied board - not so much experimentation or creative use. The chokepoints are always in the same place, the destructible rocks are always in the same place, the watchtowers are always in the same place, etc. etc.
In some tournaments though there are neato words on the maps.
You could balance the game mathematically before special abilities and rules come in, but players wouldn't like it.
And I'm not talking about the "people secretly want to be OP" junk.
I mean writing an army list would be insane. It would be like "ok this guy has 1 attack with w battlefield role, and x strength and y survivability and z movement and I want to give him a power weapon, so it costs (some equation incorporating wxy and z.)"
You could even fit in racial alterations easily by saying "eldar reduce their movement cost multiplier 10%, orks decrease their strength multiplier 10%, marines decrease their multipliers by 2.5% each." Or something of the sort to give each race their niche.
But the entire process would be NASA levels of complicated.
Unit1126PLL wrote: 1) Fair enough, though I still think making hard & fast "rules" for terrain placement will be a bad idea, while making "suggestions" will end up like most of the "suggestions" in the current rule-book i.e. ignored.
3) Depends on if he is a Black Templar or not! There are ways to ensure he gets the charge - Chapter Tactics & Command Re-Rolls come to mind, at least in the case of loyalist terminators - but now we have to expand out to force multiplication and stuff like that. Stuff that has costs (CPs or Points or even Opportunity Costs) - but generally, I do think most elite units can 1v1 or even 1v2 or 1v3 most horde units, model per model.
4) This is true, but still kind of gamey. Like the difference between "First Blood" and "First Strike" at NOVA was no contest at all - "first strike" was always the better choice. It's 'false choice' in many ways. But yes, having this in the game would improve it dramatically I think.
1.) They may be but I get the feeling if they were cased in "hey we playtested and balanced the game assuming there was x amount of terrain, and x amount of LOS blocking terrain etc." people would use it. I remember a lot of use of the 25% guideline back in the day. It would not be perfect because it isn't fixed, but it is better if people use say 33% of the table being terrain 1/3rd of which is LOS blocking if that is what testers used vs 0 LOS blocking, or 33% of all LOS blockers etc. For instance while not terrain Malifaux came out and said 50 points is where we balanced the game. You aren't required to play it, but it is what most people play because that is where the game works best.
3.) Even re-roll charges is only a 50-50, so not really ensure getting the charged. Ravenguard is really the only near guaranteed set up charge (because you can move prior to charging)
4.) Not saying they are all good, but you don't have to take first strike at all, so you might decide to take "kill the most expensive unit" or whatever.
1) True, there should be /some/ guidelines. That'd help a bit.
2) Okay, Raven Guard then. I still think we're losing the forest for the trees, and that point is that (dice permitting, it being a dice game & all), elite armies can certainly 1v1 horde armies, and in many cases, 1v2 or 1v3 them, or even worse for certain units. The problem is they have to set the conditions for this engagement to occur 1v1 instead of 1v30 (as I mentioned) so they're harder to play on open boards.
4) Right, like I said, something like this would be good for the game - though perhaps GW was trying not to implement it, as it encourages skew even more: "Oh did I bring all tanks? That's okay, I'll choose my own objectives that suit that army just fine."
D3 per table quarter was the 3rd edition standard. I was just giving a reference to a time when there were codified rules for terrain placement and what terrain would be on the table.
I Idon't think *every* game should have a ton of terrain either. I think that each game should be different. Some with none. Some with a lot. Some in the middle.
Unit1126PLL wrote: 1) True, there should be /some/ guidelines. That'd help a bit.
2) Okay, Raven Guard then. I still think we're losing the forest for the trees, and that point is that (dice permitting, it being a dice game & all), elite armies can certainly 1v1 horde armies, and in many cases, 1v2 or 1v3 them, or even worse for certain units. The problem is they have to set the conditions for this engagement to occur 1v1 instead of 1v30 (as I mentioned) so they're harder to play on open boards.
4) Right, like I said, something like this would be good for the game - though perhaps GW was trying not to implement it, as it encourages skew even more: "Oh did I bring all tanks? That's okay, I'll choose my own objectives that suit that army just fine."
4.) True, but the opposite is also true depending on the objectives available to the opponent.
4) Yes... which is why it encourages skew. Lists will be like "I brought 600 conscripts, but don't worry, if I fight tanks I'll use this set of objectives which makes the tanks irrelevant." and the tanks will say "600 conscripts? I'll take this set of objectives that makes the conscripts irrelevant." At that point, the armies aren't even interacting with eachother - the conscripts are pursuing their objectives, the tanks are doing theirs, and any shooting/assaulting that happens may very well turn out to be incidental and irrelevant to the outcome.
An example might be:
The conscript player picks 3 objectives: 1) Have more models alive at the end of the game than your opponent- +3 VP 2) Having more than 3 units in your deployment zone gives you +1 VP for each turn after the first. 3) If no enemy enters your deployment zone before the end of the game, +3 VP.
The tank company player picks 3 Objecitves: 1) Having 0 units in your deployment zone gives you +1 VP for each turn after the first 2) Kill more models than your opponent: +3VP 3) Assaulting an enemy unit during a turn gives you +1 VP.
You can see why the game would be purely incidental, and that neither army can meaningfully stop the other from achieving its objectives.
EDIT: The example is pretty bad, but the point is that people will game the hell out of the system, and more stuff that is "Gamey" and less "Engaging & interesting" is really dumb.
You could reasonably balance the *potential* of a unit in 40k, using a mathematical model. Balancing a player's ability to take advantage of that potential is entirely another thing.
You could have a model with 1" move, 1" melee range, 1 wound, 1 toughness, no save, but an attack that hits automatically and does 100 mortal wounds. The defensive potential is terrible, the movement potential is terrible, we'll assume there's no psychic phase potential and no shooting but the Close combat potential is amazing!
So, *potentially* you could destroy anything that you could dupe your opponent into moving next to. And you likely have other shenanigans, like a psychic power available that would let you move that unit to within striking range.
The value of such a unit is next to nothing, because on it's own, it has no likely means to achieve it's objective of destroying a high-value target.
What it does, is skew the potential of other unit capable of delivering it. The potential of a Chimera goes WAAAAAAY up if you add this Mini-nuke to the Astra Militarum, for example. The Chimera's potential value goes up so much, it would be worthless to put anything else inside of it by comparison.
Which is part of the problem when balancing 40k. You can have balls-out potential on some models, but it's own limiting factors bring the points down. Smite is a limiting power, on a Primaris, but you can put 12 PP's in a Chimera and improve their movement / defensive capability, and the damage potential is increased. Not because the PP's got better, but because the Chimera increased their potential. So the CHIMERA that delivers the PP's needs to go up in price, because the CHIMERA creates the potential.
Which makes the Chimera over costed when carrying an Infantry squad, but that's because the CHIMERA's potential is wasted, by not filling it with Primaris Psykers.
That's the trouble. People think that PP's are undercosted. But the truth is that the Chimera / whatever delivery method chosen is undercosted, when carrying Primaris Psykers.
Foot Slogging Primaris Psykers have precious little choice in target. The opponent chooses what to place close to the Primaris cluster, and then the Primaris wipe it out. By giving them improved movement / deployment options, that element increases the potential of the PP, and thus needs to have a cost related to the potential it adds to the army.
An alternative would be to base the cost for a Chimera on the unit it carries. Transports could only carry specified units / models. A Chimera that transports unit A costs B points. A Chimera that transports C unit costs D points. A Chimera costs X points, plus Y points for every Z model it carries.
Such balance could be achieved, though the benefit of such an investment is unlikely to be worth it. Heuristically, you're better off playing games, and adjusting points as you go. You could have a simple formula to give you a "base cost model" for models, and then adjust after a bit of play testing.
Something like a vote system, while sure to be abused, would say every 6 months, players vote for a +5% to -5% cost on units. As new items are released, and old units become worth more or less, you have adjustment on the fly. Points become fluid with the game.
Easily abused, and not likely real-world practical. In a perfectly honest and informed universe that would be an easy solution.
You could... but it would probably be a very different type of game. You would either have fixed armies with no options, (Checkers/Chess) or everything balanced by points... and then you end up with 1 Space Marine vs. 10 Cultists. You could eliminate the dice altogether, but that element of chance is what appeals to a lot of people.
There are a lot of different ways you could "balance" 40K, but it would either totally remake the game into something unrecognizable as 40K, or eliminate all those elements that make the game attractive to play.
There should exist some form of basic formula that you use to design units for every faction (Every faction should have a different formula, because good meele should be cheaper for Khorne than for Tau). But then, you need to playtest the gak out of it, to encounter the appropiate point costs for things that are very hard to balance by pure math.
It depends on what you mean by balance. Is it possible to make things more or less even in all cases? Sure? Do people want a game where their choices are irrelevant to the point where 50 points of bolters do the exact amount of damage to a Land Raider as 50 points of LasCannons do while LasCannons are also equally effective at clearly hordes of Boyz? Probably not?
Ultimately, the game needs math. It needs to know what kind of output you get out of each of its options and what it should be charging for them. At the same time, balance needs to come from other places. There's value in the game asking a diverse set of questions and forcing players to make choices accordingly, but to achieve that, the game needs to require and reward that kind of diversity. It needs to require diverse battlefield roles to succeed more than it needs everything to be "equal".
Players also need to realize that the more options they have, the more things just aren't going to be optimal. There's probably a perfect game possible that individually costs each model for each wargear combination correctly, but ultimately that guy with the poor accuracy might just be more efficient with a high volume weapon than something with a single important shot. It's not perfect, but sometimes perfect means cutting things that aren't going to work regardless of points. Probably what's more important for players though is that a game that rewards less spam makes it easier to recover from bad options. It's one thing to have that one weak unit; its another when you've got 6 of them.
There are a lot of points that have to be looked at:
-Make every faction unique by defining army strenght and weaknesses, and stick to them.
-Make every unit work, give all of them a roll, remove fake options.
-Reduce Rock/Paper/Scissor from matchups.
After creating units, you can adjust with point costs so that every unit is viable. Restrictions are another important part.
You need internal AND external balance.
In the end x points of army A (the most viable builds) have to get roughly 50% (more like 45-55%) chance of winning against x points of army B (also the most viable builds).
But you can't model this and compute solutions with mathematics. As I said, there are too many non-linear equations in several variables to solve. Better way would be to teach a computer to play and use machine learning. But it's hard to teach a computer how to play a complicated game like this. (Go and Warhammer are both complex, but Warhammer is complicated compared to Go...)
Formula? probably not in the way you would usually think of it - you can however simulate it reasonably well (if a pretty complex model).
Wouldn't be a simple simulation as you have to model an army to consider a units place within it, then consider a multitude of armies to see how a unit varies.
It would be possible, but the result is only ever going to be as good as the assumptions.
I did something at an exceedingly crude level for warhammer 7th when looking at the various upgrades for orc boys - running them against a basket of enemy units one on one, but taking into account various buffs etc they could have.
Wind it up, set it to run say 1,000 times and you can have an average result for the outcome but also critically the variance in that output - I weighted a more predictable outcome slightly higher than one with more potential but a higher variance.
Keep in mind its possible to simulate highly complex systems so simulating warhammer is certainly possible, especially as unlike many "real" systems the rules are all known in advance.
The question is more "is there any actual benefit to doing so?"
Given it will take quite a while and wouldn't change much.
What would be worthwhile is GW finding one of their bean counters who is good with numbers, one of their creative types and someone who can programme, and locking them in a room for a while as GW could benefit from a programme that is able to play the game to a reasonable level to use as a testing tool.
What you end up with is basically not a formula, but a computer game, able to play itself and tell you an outcome, not so much in "unit X should cost Y points" but in relative terms so you can pick your own baseline.
It will also then show you what changing the scenario does to the mix, so you can simulate across many and adapt based on the types of games actually played.
In the end though given GW don't use fractional point values you can science the poo out of it all you like, but a Grot is still going to come in at 3 points most likely, because 2 is too little and 4 is too much - what you will "know" is say that having five units of 30 statistically results in a Grot that should cost 4 points or whatever.
I doubt I'll ever actually be bored enough to write such a simulation though, nearly as much as I doubt the rules will sit still long enough to benefit from it
I mean the basic of basic game before all the special synergies and special rules could be fully balanced out based on mathematics, and the statistical probability of dice rolls
the problem will always come from various unit buffs and some special rules.
that and non tangable abilities like deep strike or things like being able to leave combat and still shoot or charge.
Unit1126PLL wrote: 4) Yes... which is why it encourages skew. Lists will be like "I brought 600 conscripts, but don't worry, if I fight tanks I'll use this set of objectives which makes the tanks irrelevant." and the tanks will say "600 conscripts? I'll take this set of objectives that makes the conscripts irrelevant." At that point, the armies aren't even interacting with eachother - the conscripts are pursuing their objectives, the tanks are doing theirs, and any shooting/assaulting that happens may very well turn out to be incidental and irrelevant to the outcome.
An example might be:
The conscript player picks 3 objectives:
1) Have more models alive at the end of the game than your opponent- +3 VP 2) Having more than 3 units in your deployment zone gives you +1 VP for each turn after the first.
3) If no enemy enters your deployment zone before the end of the game, +3 VP.
The tank company player picks 3 Objecitves:
1) Having 0 units in your deployment zone gives you +1 VP for each turn after the first
2) Kill more models than your opponent: +3VP
3) Assaulting an enemy unit during a turn gives you +1 VP.
You can see why the game would be purely incidental, and that neither army can meaningfully stop the other from achieving its objectives.
EDIT:
The example is pretty bad, but the point is that people will game the hell out of the system, and more stuff that is "Gamey" and less "Engaging & interesting" is really dumb.
That is up to writing objectives that force interaction between players but can be done
Better by some builds than others.
Breng77 wrote: Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
But it is true, it's called using the correct tool for the job and should be considered by the player when loading out his troops. What balances the fact that the lascannon is mounted on a flyer instead of a dev or tac unit is that the flyer costs a lot more than the single model that's carrying it on the ground.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Sure if you don't want situation where every unit is equally good option then sure. Of course then one player will have more than 50% chance of winning before first dice is rolled.
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
Why you are worried about ability of rerolls affecting offensive output when you have ALREADY DECIDED TO IGNORE LOTS OF THINGS that affect offensive output?
You don't care about all the factors so how you arbitarily decide what factors you care about?
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Sure if you don't want situation where every unit is equally good option then sure. Of course then one player will have more than 50% chance of winning before first dice is rolled.
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
Why you are worried about ability of rerolls affecting offensive output when you have ALREADY DECIDED TO IGNORE LOTS OF THINGS that affect offensive output?
You don't care about all the factors so how you arbitarily decide what factors you care about?
So, what you’re saying with your first comment, is that you want strategy to be worthless along with player skill and understanding of the game?
I think you’ve mis-understood what I was saying in regards to the pricing of abilities.
Abilities would be priced into the cost of the units that provide them (i.e the re-rolls 1 buff would be priced into the cost of the model providing the buff, not costed into the standard unit benefitting from it).
Likewise, defensively, a 4++ ability would be costed into the unit that has it/provides it and so on.
Terrain is different and not a constant like abilities are. Your statement there, seems to imply that you want to “add in” the cost of potentially getting a +1 save due to being in cover to all units, whilst also then potentially reducing the cost of everything just because, in some instances they might not have LoS.
This is why I believe terrain should be kept separate from any attempts to “balance” all the units, weapons and options in the game. If one player plays the table well, and the other player does not, the player doing well should not be penalised (as the other player would be getting in-built buffs) because of it.
Going back to what others said about there always being unbalance within games due to matchups, and therefore it implies the game overall is unbalanced, I can’t agree with. It is literally impossible to predict every single list building choice every player might make, and then somehow “balance” that against another random list created in the same way. However, it is possible to balance the game as a whole.
Again, with the conscript spam vs tanks as an example. What matters more for the game?
1. Conscripts and Tanks are balanced overall.
2. Conscripts are balanced against Tanks specifically just in case 2 players decide they want to run those 2 lists.
Should it matter than Elite style armies are balanced across the game, or do they need to be balanced specifically against smite spam?
Should LoS ignoring weapons be balanced across the game, or should they occasionally be nerfed into the ground each time you’re playing on a heavy LoS blocking terrain table?
This game is, and always will be, a mini game of rock-paper-scissors, inside of a bigger game. The bigger game being the entire range of units in 40k and the mini game being the specific game played at whatever points level you choose. If I take a Tau gunline and you take nothing but footslogging melee units in a 2k game how would you balance that, whilst retaining overall balance for the gunline vs other styles of list and ensuring Tau as a faction remaining balanced in the entirety of the game?
People need to accept, that the biggest cause of “inbalance” within games, has nothing to do with the balance of the overall game. It biggest factor, in your standard games, is people and their choices.
The question therefore is, how do you balance peoples choices and ensure that, no matter what, the game remains something like a 45/45/10 (10% chance of a draw)? In my mind, that is practically impossible, because an extremely high percentage of what determines the outcome of a standard game of 40k, is down to the player and not the units themselves.
You can have the most broken, OP netlist going right now and still lose games. Of course, you have an advantage, but how much of an advantage depends on other factors outside of the game.
The only way you would be able to prevent skew lists from causing a problem, is with list building restrictions. I.e. limit the amount of times you can take each unit in your overall list. All that does though, is reduce the potential swing of a one-sided rock-paper-scissors matchup. Balance would come from the units themselves.
Breng77 wrote: Sorry this is not true, the value of a lascannon is much higher if it can see every point within 48" than if it has a hard time drawing LOS more than 24" in most places, immobile artillery has a greater value in missions that don't involve claiming objectives or needing to move. Something can be great statwise, if it is bad at winning the game it is a bad unit. You mention abilities, the ability to ignore LOS for shooting has a different value on planet bowling ball than in city fight.
But it is true, it's called using the correct tool for the job and should be considered by the player when loading out his troops. What balances the fact that the lascannon is mounted on a flyer instead of a dev or tac unit is that the flyer costs a lot more than the single model that's carrying it on the ground.
The issue with that is that abilities of units are hard to just put into the points of units because the equipment on a unit can vastly change its effectiveness, so if it has options a flyer with a lascannon might be worth quite a bit more than the same one with a lasgun, and the difference in effectiveness might be more than a static difference in weapon cost can represent. Which means you never see a unit with certain weapons. This is especially problematic when other models get the same weapon at the same price, because a different weapon might be of different value on that unit. Take the IG plasma changes, that is gw acknowledging that the cost difference between a BS 3+ and 4+ Lasgun, is different than a bs 3+ and 4+ plasma gun.
Ok I skimmed through the topic and I didn't see anyone try to formalize the problem mathematically (maybe I missed it), which is the first step toward solving it.
First let's assume that there is only one possible terrain, because else it's way too complex. Alternatively you should give extremely explicit rules for how the terrain is going to be build.
What we want, for the game to have external balance, is to make sure that for each possible matchup, there is a Nash equilibrium where the win expectancy for each player is 50%.
What we want, for the game to have internal balance, is to make sure that each unit is present in the army selection in some Nash equilibrium.
If we manage to prove this, as we will have many very precise information about the Nash equilibrium, it would mean that we likely know all of it. In other words, it means that we would literally know the exact best possible way to play the game. Which will make the game... boring. Becomes literally the same as Rock Papper Scissors. But thanksfully we won't be able to compute this. We can't even find a Nash equilibrium for freaking CHESS! And chess is both deterministic (no random rolls) and discrete (chess board with pieces being on specific squares versus game table where a model can be anywhere).
We still can solve the problem with some very easy solutions like everyone has the same profile, or we have just three profiles which works as rock paper scissors, but those are just boring.
Terrain and mission etc are all part of the strategic side of the game, the part completely in the hands of the players. This should not have an impact on the points costs of units. A predator is still a predator regardless of whether it has a clear los across the table or whether there are los blocking items in front of it. It doesn’t affect the “mathematical efficiency” of the unit, but rather the “strategic worth” of the unit in that given situation. It’s offensive and defensive power in the situations where it arises are what should drive balance calculations – not the “oh, for some reason player 1 put his tank behind a wall and now can’t shoot at anything without moving”.
Sure if you don't want situation where every unit is equally good option then sure. Of course then one player will have more than 50% chance of winning before first dice is rolled.
Abilities will be more tricky to nail down. Things like re-roll 1’s is easy enough – it provides a x% increase in a models offensive output. This can then be based off assumptions that it is used in conjunction with 2 5 man units, buffing the effectiveness of 11 models in total. However, abilities like “character”, “fly”, “psyker” etc would all need to be played with to find a starting trial value.
Why you are worried about ability of rerolls affecting offensive output when you have ALREADY DECIDED TO IGNORE LOTS OF THINGS that affect offensive output?
You don't care about all the factors so how you arbitarily decide what factors you care about?
So, what you’re saying with your first comment, is that you want strategy to be worthless along with player skill and understanding of the game?
I think you’ve mis-understood what I was saying in regards to the pricing of abilities.
Abilities would be priced into the cost of the units that provide them (i.e the re-rolls 1 buff would be priced into the cost of the model providing the buff, not costed into the standard unit benefitting from it).
Likewise, defensively, a 4++ ability would be costed into the unit that has it/provides it and so on.
Terrain is different and not a constant like abilities are. Your statement there, seems to imply that you want to “add in” the cost of potentially getting a +1 save due to being in cover to all units, whilst also then potentially reducing the cost of everything just because, in some instances they might not have LoS.
This is why I believe terrain should be kept separate from any attempts to “balance” all the units, weapons and options in the game. If one player plays the table well, and the other player does not, the player doing well should not be penalised (as the other player would be getting in-built buffs) because of it.
Going back to what others said about there always being unbalance within games due to matchups, and therefore it implies the game overall is unbalanced, I can’t agree with. It is literally impossible to predict every single list building choice every player might make, and then somehow “balance” that against another random list created in the same way. However, it is possible to balance the game as a whole.
Again, with the conscript spam vs tanks as an example. What matters more for the game?
1. Conscripts and Tanks are balanced overall.
2. Conscripts are balanced against Tanks specifically just in case 2 players decide they want to run those 2 lists.
Should it matter than Elite style armies are balanced across the game, or do they need to be balanced specifically against smite spam?
Should LoS ignoring weapons be balanced across the game, or should they occasionally be nerfed into the ground each time you’re playing on a heavy LoS blocking terrain table?
This game is, and always will be, a mini game of rock-paper-scissors, inside of a bigger game. The bigger game being the entire range of units in 40k and the mini game being the specific game played at whatever points level you choose. If I take a Tau gunline and you take nothing but footslogging melee units in a 2k game how would you balance that, whilst retaining overall balance for the gunline vs other styles of list and ensuring Tau as a faction remaining balanced in the entirety of the game?
People need to accept, that the biggest cause of “inbalance” within games, has nothing to do with the balance of the overall game. It biggest factor, in your standard games, is people and their choices.
The question therefore is, how do you balance peoples choices and ensure that, no matter what, the game remains something like a 45/45/10 (10% chance of a draw)? In my mind, that is practically impossible, because an extremely high percentage of what determines the outcome of a standard game of 40k, is down to the player and not the units themselves.
You can have the most broken, OP netlist going right now and still lose games. Of course, you have an advantage, but how much of an advantage depends on other factors outside of the game.
The only way you would be able to prevent skew lists from causing a problem, is with list building restrictions. I.e. limit the amount of times you can take each unit in your overall list. All that does though, is reduce the potential swing of a one-sided rock-paper-scissors matchup. Balance would come from the units themselves.
Which is why I think you need restrictions, because one sided match-ups is what you are trying to avoid. It is literally impossible to balance units that have specific roles if cases occur where those roles have no value. It isn't strategic it is unbalanced game design. Now I do think it should be possible to make bad lists. Even with restrictions it should be possible for someone to take all anti-infantry weapons and have no answers to the tanks that do appear. What we are trying to avoid is the scenario where that matchup is all tanks vs all anti-infantry or the opposite. The hardest part of this to avoid is super horde spam making anti-tank irrelevant. Mostly because many cheap infantry units are stock troops for certain armies. That said it is doable with limits within particular factions.
I'm still on the "this is possible" train. Let's be clear, you will never achieve a perfect 50/50 balance in every context. Why would you want to? Some units are strong against specific opposing units or in a specific context by design. To "fix" this would kill much of the flexibility of the game.
In 2006, Counter Strike: Source introduced Dynamic Weapon Pricing. The mechanism was that the total in-game dollars (i.e. points) spent on each weapon was calculated on a weekly basis, and the cost of each weapon dynamically adjusted every Monday. Thus, if everyone uses an M4 rifle in the first week, the price increased. If no one uses the MP5 during the same week, the price decreased. Iterating on that every week would have allowed an equilibrium to eventually occur. It was disliked by the community and discontinued.
How would this look in 40k? Well, if everyone goes to a tournament with Guilliman and no one goes with Assault Marines, I think we can agree that the indication is that Guilliman is strong for his points and Assault Marines are weak for their points. Thus you would adjust Guilliman up in points and assault marines down.
This is not a mathematical formula, which was the original request. This is a mechanism through which one could balance the game through iteration. Simply rinse and repeat until you find equilibrium.
You use the pricing model I previously proposed to set a starting value on new units, then you use this mechanism to adjust according to some metric on effectiveness.
A fool's errand, if there ever was one. I am truly sorry if this statement makes you feel bad.
Let me elaborate: this game isn't Warcraft II. There are more than two sides and no sides have equal forces.
And this game shouldn't be like the abstract nonsense that is the chess. That game is balanced but it's a horrible, horrible strategy game because of that very fact.
This game isn't supposed to be balanced and it's by design.
RedCommander wrote: A fool's errand, if there ever was one. I am truly sorry if this statement makes you feel bad.
Let me elaborate: this game isn't Warcraft II. There are more than two sides and no sides have equal forces.
And this game shouldn't be like the abstract nonsense that is the chess. That game is balanced but it's a horrible, horrible strategy game because of that very fact.
This game isn't supposed to be balanced and it's by design.
RedCommander wrote: And this game shouldn't be like the abstract nonsense that is the chess. That game is balanced but it's a horrible, horrible strategy game because of that very fact.
This game isn't supposed to be balanced and it's by design.
I’m not sure if you’re trolling. By nearly any metric, chess is a great strategy game. It’s rules have remained unchanged for hundreds of years, it’s played globally, and we are still seeing new developments in strategy.
Your argument is that balance is the thing that makes chess a bad strategy game. Every indication is that chess is actually a good strategy game, I’d go so far as to say that it’s superior to 40k from a Puritan game-design perspective. And my argument would be that the balance strongly supports the “goodness” in chess.
Using maths for balance is basically what GW has done with the previous editions and I don't think it can ever work with a complicated game like this.
The only way to get decent balance, in my opinion, is to use analytics data from sales, battle successes and frequency on the table and this appears to be what they are doing.
To clarify, people will buy and use good models and won't as much with bad ones. Just seeing units on the table is not such a bad indicator of whether they are OP. If they adjust the points too much and it stops appearing, they they know they went too far. It takes time though and obviously isn't perfect as it doesn't take into account rule of cool and unpopular models for other reasons.
The more mature 8th becomes the closer and closer we should get to reasonable balance.
Sorcererbob wrote: I'm still on the "this is possible" train. Let's be clear, you will never achieve a perfect 50/50 balance in every context. Why would you want to? Some units are strong against specific opposing units or in a specific context by design. To "fix" this would kill much of the flexibility of the game.
In 2006, Counter Strike: Source introduced Dynamic Weapon Pricing. The mechanism was that the total in-game dollars (i.e. points) spent on each weapon was calculated on a weekly basis, and the cost of each weapon dynamically adjusted every Monday. Thus, if everyone uses an M4 rifle in the first week, the price increased. If no one uses the MP5 during the same week, the price decreased. Iterating on that every week would have allowed an equilibrium to eventually occur. It was disliked by the community and discontinued.
How would this look in 40k? Well, if everyone goes to a tournament with Guilliman and no one goes with Assault Marines, I think we can agree that the indication is that Guilliman is strong for his points and Assault Marines are weak for their points. Thus you would adjust Guilliman up in points and assault marines down.
This is not a mathematical formula, which was the original request. This is a mechanism through which one could balance the game through iteration. Simply rinse and repeat until you find equilibrium.
You use the pricing model I previously proposed to set a starting value on new units, then you use this mechanism to adjust according to some metric on effectiveness.
That is not balance at all, that is just shifting the imbalance from one unit to the other. Like trying to balance weighting scales by constantly moving all weight from one scale to the other. You will never get balance in that way.
Sorcererbob wrote: I'm still on the "this is possible" train. Let's be clear, you will never achieve a perfect 50/50 balance in every context. Why would you want to? Some units are strong against specific opposing units or in a specific context by design. To "fix" this would kill much of the flexibility of the game.
In 2006, Counter Strike: Source introduced Dynamic Weapon Pricing. The mechanism was that the total in-game dollars (i.e. points) spent on each weapon was calculated on a weekly basis, and the cost of each weapon dynamically adjusted every Monday. Thus, if everyone uses an M4 rifle in the first week, the price increased. If no one uses the MP5 during the same week, the price decreased. Iterating on that every week would have allowed an equilibrium to eventually occur. It was disliked by the community and discontinued.
How would this look in 40k? Well, if everyone goes to a tournament with Guilliman and no one goes with Assault Marines, I think we can agree that the indication is that Guilliman is strong for his points and Assault Marines are weak for their points. Thus you would adjust Guilliman up in points and assault marines down.
This is not a mathematical formula, which was the original request. This is a mechanism through which one could balance the game through iteration. Simply rinse and repeat until you find equilibrium.
You use the pricing model I previously proposed to set a starting value on new units, then you use this mechanism to adjust according to some metric on effectiveness.
That is not balance at all, that is just shifting the imbalance from one unit to the other. Like trying to balance weighting scales by constantly moving all weight from one scale to the other. You will never get balance in that way.
Eventually you end up with a balanced system though - if you use math to figure out how best to make the tweeks over time. In time you'd have a system where no one was picking 1 thing over another because it was cheaper - at that point they would only chose off preference.