Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

 Peregrine wrote:

 junk wrote:
How do you all, individually, determine whether or not a unit is 'worth it's points'?

Trial and error. Looking at the rules and comparing it to similar units can give you a general idea of what has potential (for example, Vendettas are obviously good) but then you have to play games to know for sure. This is also how you balance things: you come up with what you think is an appropriate point cost, and then you playtest it to see what changes you need to make.

And of course now the problem is you're creating an extremely complex system, and getting rid of the only advantage (a degree of objectivity) that you used to have over just guessing at a point cost and playtesting.

So you have no measurable criteria other than "maybe this will work, I should try it"?

 junk wrote:
So you have no measurable criteria other than "maybe this will work, I should try it"?

Of course I have measurable criteria. But, as I've been saying, those measurable things aren't everything you need to consider. For example, I can look at how many MEQs a Vendetta kills per turn, how many damage results it inflicts on each AV, and how many shots it takes on average to kill it. I can then compare those numbers to other units and see how much the pure shooting is worth. But what I can't measure like that is how much the Vendetta's transport capacity is worth. I can make a guess at it based on my army's strategy and my personal preferences, but you could legitimately have a completely different value for that transport capacity. And neither of us would be basing our value on some objective measurement.

Yeah, I think you could actually.
Transport capacity is a measurable function of a vehicle. Every vehicle has a transport capacity, even though it most cases it's 0. Just because I don't yet know how to figure it out though doesn't mean it's not measurable.

I get what you're repeating about subjectivity, we'll work to minimize that with consensus, and then try to formulate those variables; that stuff can really only start to be measured once there is a core system in place. Right now, with a very flimsy sketch for a core system, yes, there is no way - I'd like to improve the system as much as possible.


Anyone else, what questions do you ask when you're trying to figure out if a unit is worth it's points?

The most important question I ask is "Does this unit, when used as competitively as possible, in a situation skewed towards it, make up its points cost over the course of the game?"

If the answer is yes, then "by how much?" is my next question. If not, my next question is "does it FEEL like it's made up its points?"

If yes, I look at what makes it feel that way. If not, I begin to look at redesigning the unit.

An important way to stop people from needlessly debuffing their units is to start with Racial stats, move to Archetypes, and then alter stats based on archetype points.

So, if making a completely new race, Racial Stats begin at 3's with 6+ save and Ld5 for 3 point per model, and follow a fairly normal cost-progression upwards.

At no point in any step can cost per model go below 1; If you lowered BS to 0 and Ld to 4, your model would still cost 1. If you had a Guardsman statline after Racial, and moved into a ranged-support Archetype, dropping Ballistics Skill to 0, Strength and Toughness, and Saves might make their cost drastically lower, but they would still be 1 point per model. After that, adding any sort of gear would be relatively 'expensive', on a model with nothing else going for it -- especially given some form of parity between model and wargear cost.

@Per again, objectivity is not necessary for players to observe a process and agree with it. The game is not based in any form of objective manner; all costs, rules, and statistics are as subjective as the idea of 'fairness' is. What you're arguing for seems to be points costing based on advanced mathematical formulas, in which case each player could determine the game's outcome from army composition alone.

 junk wrote:
Transport capacity is a measurable function of a vehicle. Every vehicle has a transport capacity, even though it most cases it's 0. Just because I don't yet know how to figure it out though doesn't mean it's not measurable.

The problem is that a given amount of transport capacity has a completely different value depending on the rest of the vehicle's rules. For example, consider the Vendetta's 12-model capacity vs. the Chimera's 12-model capacity. The Vendetta is faster and harder to kill, but has to wait a turn or three before it can deploy its passengers and the unit can't fire without disembarking. The Chimera is slower and less durable, but its passengers can start shooting right from turn one without ever leaving their mobile bunker. Now you have to decide whether fire points or flying is worth more of a multiplier on the base price of the unit, and there's no objective answer to that. A player who wants a unified wall of AV 12 tanks might leave their Vendettas entirely empty as dedicated AA units (transport capacity worth zero points), while a foot blob player with limited mobility might desperately need their Vendettas to deliver late-game scoring units anywhere on the table (transport capacity extremely valuable).

Now let's make it even more complicated and add 12 models worth of transport capacity to the Basilisk. How much is it worth? Very little since a Basilisk's primary role is to sit back out of LOS and shoot, which means it can't deliver its passengers anywhere. Now you need another arbitrary modifier to reduce the price of transport capacity if the vehicle has other roles that make it hard to use as a transport, and an arbitrary choice of which units qualify for the discount.

And all of this is just one army. Let's continue to make things worse and consider the 12-model capacity of the Devilfish. While the increase from 6 models to 12 is vital for IG players (fixed squad sizes of ten models) many Tau mech players take MSU squads in their Devilfish so the increase from 6 to 12 models is almost worthless. I guess we'll need another price modifier to reflect how often the army with the transport takes full-size units.

In the end you're going to have to special-case pretty much every transport, and whatever value you set for a unit of transport capacity is going to be irrelevant compared to the giant mess of arbitrary modifiers on that number. Your end result is going to be no less subjective than guessing "well, I think that's about 120 points worth".


Automatically Appended Next Post:

 chrisrawr wrote:
The most important question I ask is "Does this unit, when used as competitively as possible, in a situation skewed towards it, make up its points cost over the course of the game?"

Which only works in the very limited case of a pure killing unit that does nothing but remove enemy models. For example, how many points is sitting on an objective (without firing a single shot) all game and giving you 3 VP worth?

@Per again, objectivity is not necessary for players to observe a process and agree with it.

Then what's the point of this system? If you agree that it's just a bunch of arbitrary values assigned to certain rules then what are you gaining compared to designing the entire unit, making a rough guess at its point cost, and fine-tuning the rules/cost through playtesting? All you're doing is making a needlessly complicated formula that accomplishes the exact same objective.

I don't think it's all that subjective. I think most reasonable people who have played a few games against a variety of lists will basically agree about appropriate point costs. However, it's a very hard thing to come up with an algorithm for. Even ignoring concerns about how units work with other units or whether they're scoring or which slot they're in, you basically have to independently price every possible loadout for a unit if you really want to do it right. There's absolutely no reason to think that the difference in price between 5 marines with 1 plasma gun and 6 marines and 1 plasma gun should be the same as the difference in price between 5 marines with no plasma gun and 6 marines with no plasma gun. GW's failure to do this sort of balancing is why some loadouts are clearly superior to others. Nobody's taking 8-man Jetbike squads or 19-man Guardian blobs.

Now to bring in those other complications:

When I'm looking at a unit in the context of an already nearly-full list, I'm asking "does this unit make my list better in a maximally cost-effective way?" Wave Serpent spam suffers from a lack of long-range anti-tank, somewhat lackluster anti-air, and few great uses for its mandatory HQ choices. It's immediately apparent that taking T3 foot models to address these issues is probably a bad idea - part of the point of Serpent spam is to deny the enemy targets for their anti-infantry weapons. In the context of a mechanized Eldar list, Wraithknights look rather appealing for their cost. Crimson Hunters probably even make sense, even though they are otherwise somewhat underwhelming, because they provide great AA and because the rest of the list is very good at making sure it's safe for them to arrive.

Balance is really something that you determine on a list by list, rather than unit by unit, basis. A codex is well-designed to the extent that it allows a good variety of lists that are competitive against the best lists from other codices. A unit in a codex is balanced relative to other choices if it is an appealing option for many of those competitive lists but not so appealing that it almost always gets taken.

So you're not going to be able to compute 100% reliable point values without taking into account, basically, everything about every other codex and everything in the current codex. A single new codex release can force the meta to change significantly. Plus other things like typical sizes and placements of terrain matter quite a bit, and probably some meta changes are irrational.

It seems to me that by far the best way to go about playing a balanced game using the 40k rules as a starting point is to just build and play a whole bunch of lists. Decide which lists you most want to have be more competitive and which ones you want to be less competitive. Start slowly tweaking the point costs of units/upgrades which appear in the lists which aren't competitive enough up while tweaking the costs of units/upgrades which appear in the lists which are too competitive down. Of course, you don't really have to do the playtesting to see that Helldrakes should be more expensive while a variety of other things in the Chaos codex should be cheaper.

An easy, if perhaps artificial-seeming, way to add variety to the set of competitive lists would just be to modify the cost of particular codex entries according to the total number of unmodified points a codex spends on that entry. For example, you might say that the actual cost of a player's selections of a particular codex entry is equal to the total listed price squared, divided by 100. If a list only had 6 Fire Warriors, it would pay only 29 points for them (~5 PPM). If a list had 11, it would pay 98 points for them all (essentially what they cost now). If a list had 30, it would pay 729 points for the lot (2.7 times what they regularly cost). One could play with the function being used (probably dividing by 200 makes more sense since that's closer to the average cost of a single choice). The point is that, no matter how good a unit is under the current rules, it's unlikely to be so good as to be worth taking 400 unmodified points in it. And no matter how bad a unit is, there's probably a lot of reason to take a small squad. If you do this then you'll get a variety of competitive lists from all of the codices if you only do a few across-the-board adjustments to particular codices that still struggle.

Overall codex design definitely needs to be a consideration, as far as enumerating that, a simpler way to do it would be to design a set of army wide traits or archetypes that carry a certain modifier to all units within the racial scheme. But we still need a consistent baseline for evaluation, and we clearly haven't come close to reaching a consensus for a system of measurement yet, but so far, many of these suggestions have been highly helpful.

Regarding Peregrine's Suggestion that units are not exclusively delivery devices and receptacles for wounds, because some of them are scoring units, just leads me to conclude that scoring unit yes/no can be another factor to help determine unit cost.

The suggestion that determining a unit's effectiveness by the measure of 'will it make it's points back" does fall in line with the idea that a unit is a delivery device for wounds, that can also be measured by it's capacity to withstand or absorb wounds. I'll suggest this list of measurement criteria again, and see if we can refine it. Again, leaving vehicles off the table for the moment.

1. Chance to inflict wound (BS, WS, Offensive Wargear, S, A)
2. Chance to withstand wound (T, Save)
3. Volume of wounds / Threshold for damage (W, Ld, Min/max squad size)
4. Effective range (wargear, movement)
5. Special considerations (Ability to distribute buffs / penalties / modify other units)

Hi again.
I feel the most important part of my previous post was missed by some.
So I will re state it.
The 40k rule set is VERY COMPLICATED ,(lots of unnecessary extra resolution methods and exceptions.)

As the game is all about UNIT interaction, it would be much more sensible, (and slightly easier) to cost the game more accurately AT the UNIT level.

NOT individual models and equipment.

As doing this factors in the basic unit synergy, and removes some of the complication.

Also having unit CLASSES , so the primary use of the UNIT can be compared to other units in that class might be helpful.
EG artillery, light infantry , tank, transport etc.

I'm with you Lanark- But that unit is still a collection of Volume of Attacks, and Volume of wounds; so a single model unit that has multiple attacks and wounds, can still be measured against a multiple model squad, which means that to find the LCD we still have to break things down to a per-model cost analysis. Although Squad size Min/Max can't be ignored as a factor.

All of the complicated rules and exceptions can be considered adjustments to the base cost, and these will probably result in fractional costs or cost multipliers, eventually, but we'd be distracted from the primary goal by considering them up front.

As far as classes - I'm in agreement, it's the archetype suggestion that has been talked about.

Can we clearly define a list of archetypes? I was trying to work on this on my own, but I came up with about 15 which means I'm not generalizing enough. Things like Versatility throw off my designations.


Automatically Appended Next Post:
Wargear Consideration:

I was thinking about everyones suggestions, and it actually makes sense that Wargear is a function of the Total unit rather than the individual model. Putting a Lascannon in a squad of 10 space marines actually increases the overall squads average wound output, so a price can only be calculated by an overall measurement. The cost should reflect the change in the units average damage output, so the cost would increase depending on the number of attacks/models that can operate at that level.

It makes sense that Wargear options are conditional to squad size specificially because it keeps those costs/calculations from being undercosted or prohibitively expensive, and allows scarcity to be a consideration of cost.

I'm going back to the drawing board on the wargear stuff, scarcity will provide a threshold that will counter min/max.

pantsonhead wrote:
For example, you might say that the actual cost of a player's selections of a particular codex entry is equal to the total listed price squared, divided by 100.

This is a really bad idea. It only "works" because you happened to find a test case where the math coincidentally gives you something vaguely sane. Consider instead a basic LRBT, which now costs 225 points, or a basic Land Raider which costs 625 points. Meanwhile armies with limited choices in certain FOC slots (Tau troops, for example) will be massively overpaying because they just don't have enough good options. You can't play Tau if taking more than one squad each of your two troops choices starts making their point costs go up to ridiculous levels.

 junk wrote:
Overall codex design definitely needs to be a consideration, as far as enumerating that, a simpler way to do it would be to design a set of army wide traits or archetypes that carry a certain modifier to all units within the racial scheme.

So now you have a different set of archetypes for each army. How is this really any better than just coming up with point costs for each new unit without a fixed formula? You're introducing so many special cases and arbitrary point adjustments that all you're really doing is spending a lot of time and math pretending that you aren't just guessing at point costs.

and we clearly haven't come close to reaching a consensus for a system of measurement yet

And you never will, because there is no single system of measurement that can handle everything. What you really need is dozens of independent systems of measurement that will increase the complexity of your idea way beyond anything that is practical while still being completely subjective and arbitrary.

Regarding Peregrine's Suggestion that units are not exclusively delivery devices and receptacles for wounds, because some of them are scoring units, just leads me to conclude that scoring unit yes/no can be another factor to help determine unit cost.

It can be a factor, but it can't be a single objective one. For example, IG melta vets being scoring isn't worth all that much because they're a unit that is often dead before objectives are counted. Scoring for a blob squad, on the other hand, is the unit's entire reason for existing. A single multiplier to the base point cost for being a scoring unit will either make the vets too expensive or the blob squad too cheap. And when you're making special-case adjustments for each individual unit your pretense of having an objective formula is gone.

1. Chance to inflict wound (BS, WS, Offensive Wargear, S, A)
2. Chance to withstand wound (T, Save)

You can't group these things together because all wounds are not equal. For example, a STR 10 weapon on a unit designed primarily to kill light infantry is not really worth more than a STR 6 weapon (it gives flexibility since you can shoot at other targets, but both of them wound your primary target on a 2+). A STR 10 weapon on a unit designed primarily to kill vehicles is worth much, much more than a STR 6 weapon. Yes you can add another weighting factor to make STR more valuable on units that expect to shoot at targets where STR 7+ is important, but that weighting factor is arbitrary and no better than just assigning a point cost to the unit without a formula.

Peregrine,

Again, I appreciate the ardor. I'm having a hard time with the 'you can't' and 'it's impossible' stuff.

Your Vets justification of 'being dead' before the end of game isn't really applicable, as if they were weighted as a scoring troop choice, they would be used differently, they would influence troop selection, and making determinations along those lines is ultimately subjective, as it reflects a player choice.

Implementing a set of modifiers that could be used as army wide conditions, or codex specific, is an overall consideration for codex design. Internal balance within a codex, and external balance with the game environment are also different. We could reverse engineer those parameters as a way to test or justify our own internal determinations. Agreed that discovering the relative value of abilities like synapse or astknf is a pretty difficult endeavor - but some degree of tolerance of army wide modifiers is still a measurable trait.

The tools of that measurement may not be immediately apparent, but I'm hoping that one or several can be determined.

Regarding wounds not being equal; they are a dependent value for sure. Right now, for this thought exercise, I'm leaving vehicles off the table. I know that it will impact the overall valuation, and yes, I realize that they are an important part of the game, but I'm definitely not there yet.

I also don't mean to be that guy who says he has all the answers and this is definitely the right way to do it; but I think that we can at least zero in on some kind of rough valuation scheme that accounts for a majority of mechanics, refine it, and build from there.

Archetypes: Sub-Archetypes --- Scoring points cost modifier beside Archetype on the right - otherwise, normal cost as listed after a discount based on archetype.

The first Archetype is free. Additional Subtypes increase cost by a cumulative x1.25. Additional Archetypes increase cost by a cumulative x1.5. Start at the top and add Archetypes and Subtypes in order for total cost.

Dedicated Assault - x1.0 - High # of Attacks or powerful Attacks

-Blitz - I4+, some form of fast movement.

-Survivable - Good save, High Toughness, or FNP of some sort.

-Horde - Discount formula of some sort to be worked out.


Dedicated Shooting - x1.2 - BS discount of some sort, other stats increase in cost.
-Glass Cannon - Saves and T must be low, gun discount of some sort.

-Survivable - Saves and/or T are high, some form of FNP?

-Horde - Horde discount of some sort.


Harrying/Fast Unit - x1.2 - jetbikes, beasts, cavalry, jump/jetpacks
-Powerful Melee - If Close Combat Archetype taken, choose this or blitz.

-Powerful Ranged - If Ranged archetype taken, choose this or glass cannon.

-Survivable - Cannot take Tank Archetype as well. Discount on T, Sv, misc survivables.

Tank - x1.5 - Does it have T5+, Sv2, or FNP and a bunch of wounds? It came from here!
-Many Wounds - Discount on wounding.
-High T, Sv, or both - Discount on saves, toughness, etc.
-Other - discount on other forms of survivability.

 Peregrine wrote:

This is a really bad idea. It only "works" because you happened to find a test case where the math coincidentally gives you something vaguely sane. Consider instead a basic LRBT, which now costs 225 points, or a basic Land Raider which costs 625 points. Meanwhile armies with limited choices in certain FOC slots (Tau troops, for example) will be massively overpaying because they just don't have enough good options. You can't play Tau if taking more than one squad each of your two troops choices starts making their point costs go up to ridiculous levels.
.

Well, I did later say that the denominator should probably be 200, exactly because of concerns about expensive single units like Land Raiders, so I don't think it's fair to use those as a criticism of the idea. Also, even using 100 as the denominator, Tau can take two squads of Fire Warriors and two squads of Kroot for basically what they pay now.

But, sure, perhaps the scaling is too strong. If you're really just wanting to give people a little encouragement to take 100 points or so of various units and to discourage 400+ points of a unit, that's pretty easy to do. Get modified points by taking unmodified points raised to the power 1.5, divided by 14. 50 unmodified points becomes 25, 100 becomes 70, 150 becomes 130, 200 remains 200, 300 becomes 370, 500 becomes 800. Force Org requirements are going to prevent people from going through their codex and taking the minimum of every entry.

Chrisrawr, a simpler template scheme would probably be better as a starting point.

Again, I still need to figure out the evaluation criteria, but if we were to take a stab using the one's I already suggested it might look like this -

Mob: Low output, high volume, low resistance, low threshold, low effective range, lowest cost category

Horde: Median Output, High Volume, Low Resistance, mid threshold, Low effective range, low cost category

General Squad: Median output, median volume, mid resistance, mid coherence, mid range, mid cost category

Focused Squad: High output, Low Volume, Median resistance, High threshold, mid range, upper cost category

Dense Squad: High output, mid volume, high resistance, high threshold, mid range, highest cost category

Every variation of the above can be considered a 'specialist'
Increasing or decreasing 1 criteria by 1 degree carries a cost multiplier

A 'scout' squad might be a general squad with a higher range with a lower resistance or threshold

A 'sniper' squad might be, as above, but with increased damage output as well

A 'heavy squad' might be, as above with higher resistance or threshold

A 'fast mob' could be, as a mob, but with higher effective range

A 'force multiplier' could be, a solo that increases the effectiveness of other squads by a measurable percentage, that is factored into cost multiplier. e.g. HQ with area buffs, IC that improves attached squads, etc...


Again, I'm not suggesting that this is a correct system, only asking if this could be a viable scheme for determining relative costs.