Switch Theme:

Share on facebook Share on Twitter Submit to Reddit  [RSS] 

In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 11:43:54


Post by: momerathe


Different people have different approaches to the game - some people enjoy cutthroat competition, some are into it for beer and pretzels (soft drinks for minors ). I'm imagining that hardcore competitive gamers are more near the top, and more casual gamers are further down the scale, but I'd be interested to see what the range of opinions are.

(obviously it depends on the size of the skill difference, but for a baseline consider the people who show up at your local FLGS)

I was particularly thinking about this while listening to a tournament podcast - the ITC rules seem pretty efficiently designed as a sorting algorithm for selecting tournament winners (allowing for matchup etc.); much shade is thrown on maelstrom for being too random (although not just for that), but maybe it's just a question of horses for courses.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 11:48:25


Post by: G00fySmiley


somewhere between 80-90%. it is a dice game, but skill still matters above all else. The only chance the lower skill should have is if they built thier list and its a counter to the more skilled player. IE skilled player came armed for imperial knights and leman russes and the opponent brought 260 grots, 3 wierdboys, ghaz, badruk and mad dock... liek what are you going to do.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 11:48:51


Post by: Overread


Honestly any wargame that doesn't allow the higher skilled player to win most of the time is likely a game running so much random that players are left with far fewer actual input and choices in the game, esp in critical moments of the game.

In theory if the armies are balanced points wise then a higher skilled player will typically bring a better designed army, deploy it better, play with it better and generally win most games.

In contrast a less skilled player is going to be worse in all those areas and likely lose most of the time.


For a wargame where part of the game is making choices this is how it should be. Because its putting the weight of the game in the choices you make. In contrast a game like Snakes and Ladders is purely one of chance. There's no control or thinking involved; you roll the dice and see where you land.
Could a wargame work like that, eh possibly but it might come off as feeling rather bland if not down right boring or just lacking any form of engagement.


A game of pure chance isn't bad; a game of pure skill isn't bad; but in a wargame I think that skill is the greater of the two. Chance is part of it, but if you put your archers in the front line and the enemy charges them with heavy cavalry then those archers should be dead. Straight up dead - they might get lucky and take down a rider or two; and there is where the chance comes into play, but overall they should be cut to bits.

Otherwise the game starts to lose meaning and connection to what its trying to represent.




Also note "high" and "low" skill are terribly vague terms for wargames since there is no standard testing or metric by which gamers can measure themselves. So some of the variation in answers is going to be influenced by how people interpret those terms. Is that low skill a beginner or a gamer who has been gaming for years, but never worked on improving their game. Is that high skill player just someone who knows the rules well or someone who wins multiple major tournaments


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 11:53:48


Post by: auticus


I voted 90%. The game should mostly be about skill on the table IMO.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:07:21


Post by: momerathe


 Overread wrote:
In contrast a game like Snakes and Ladders is purely one of chance. There's no control or thinking involved; you roll the dice and see where you land.


Funnily enough, I was going to put "snakes and ladders" in as the 50% option.

Also note "high" and "low" skill are terribly vague terms for wargames since there is no standard testing or metric by which gamers can measure themselves. So some of the variation in answers is going to be influenced by how people interpret those terms. Is that low skill a beginner or a gamer who has been gaming for years, but never worked on improving their game. Is that high skill player just someone who knows the rules well or someone who wins multiple major tournaments


Yes.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:21:05


Post by: p5freak


Its still a rock paper scissors game. If your list is the paper to the other players scissors your higher skill level doesnt really matter much. And the dice gods also have a word in this. I would say 60-70%.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:30:14


Post by: Fan67


Lower skill players are generally lose not because of dice, but because of lack of knowledge. They forget threat zones, act without knowing opponent's abilities, open up for tri-pointing, make wrong range between screening units and valuable ones, concentrate more on wrong/less valuable/risky mission goals.

On the other hand skilful players mostly lose because of dice (theirs' or their opponents').

So I think 80-90% is a reasonable enough rate in ideal conditions. But rock-paper-scisors nature of armies bring actual rate closer to 60% i guess.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:42:06


Post by: timetowaste85


Equally good armies that are tuned against each other, the higher skill player should win almost every time. Like 80% of the time. Random luck can happen, but if everything is equal except the skill, higher skill should win.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:49:58


Post by: auticus


 timetowaste85 wrote:
Equally good armies that are tuned against each other, the higher skill player should win almost every time. Like 80% of the time. Random luck can happen, but if everything is equal except the skill, higher skill should win.


That is true. Equal armies do show you who is the better player, not the player who can build (or copy) the most efficient mathematical list.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:55:26


Post by: momerathe


So the reason I was thinking about this is the idea of having different game types for different purposes, without getting into the whole "ways to play" thing with Narrative.

If you're running a tournament, or the ITC, then you want your game to be as close to a pure contest of skill as you can get them - the whole point is to find the "best" player (or the best on the day at least).

If OTOH you're just playing for "fun", you might think it's a feature that anyone has a chance of winning every now and then, so they don't get discouraged, and get their occasional day in the sun. (I scare-quoted "fun" because some people really get a kick out of competitive play, and I don't want to imply that it's not fun )


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:55:32


Post by: Haanz


I voted 90%, but 80% wouldn't offend me either.

Honestly though, it's less about how random the game is for me about how much army composition matters that I wish would change. I feel like Army Composition and Matchup is probably the biggest determining factor in games rather than actual player agency in a lot of circumstances.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 12:59:10


Post by: Wayniac


80-90% IMHO, voted 80. Most of the time, but there should always be those unexpected luck/lucky maneuver type things where a lower skill player should win.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 13:38:07


Post by: Peregrine


Less than 50%, preferably less than 25%. WAAC TFGs should be punished.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 16:03:20


Post by: G00fySmiley


 p5freak wrote:
Its still a rock paper scissors game. If your list is the paper to the other players scissors your higher skill level doesnt really matter much. And the dice gods also have a word in this. I would say 60-70%.


I don't think even with a list designed to beat Nick N or Juice (among others)'s genesteelers they would still probably find a way to beat me. I regard myself as a above average (but barely probably top 60ish%) player but watching their games while i am looking a turn ahead on predictions/averages they seem to be looking all the way to the end game beyond what I pull off.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 16:12:45


Post by: Inquisitor Lord Katherine


That depends on the level of skill difference. There should be good enough odds that the "underdog" can win if they take advantage of key opportunities.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 16:54:58


Post by: Elbows


That depends on how you define wargame.

40K is barely a wargame if we decided a wargame should represent actual combat (even in a fictional setting). If we use the term wargame loosely (as we most often do) what we're really discussing is an advanced form of a chess. A balanced, even game based purely on skill. But then, Chess doesn't really represent war or combat at all.

40K has only the slimmest trappings of "chaos" that would be present in a real battlefield, just enough to allow the occasional mistake or lucky win by the underskilled player. Sure, it's a dice game but you can reduce that impact pretty strongly. So, as a non-real-wargame I'd say 80% of the time, giving the 20% to crazy dice swings.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 17:10:14


Post by: Stux


It's funny, because while if you poll the community you'll get a strong showing for less randomness - in practice strategy games with a decent amount of randomness generally have broader appeal.

A big part of this is because people generally overestimate their ability to play games. Randomness allows you to be worse than you think you are, but still win games a reasonable amount and have something external to blame even when it actually was your fault you lost.

It's a situation where what people think they want doesn't really bare out in the data of what people collectively do.

Pretty much every gaming group has that person who is probably a nice guy but just isn't very good at playing. They'll more than everyone else, but they'll pull out a win every so often, and be really happy about it! Imagine cutting their wins in half, or even less. They probably would stop playing fairly quickly.

All this isn't to say that the game couldn't be improved with regards to randomness. It certainly could. But there's more you can do than simply adjust the winrate. Some types of randomness feel worse than others, so the key is leaning more on the feel good randomness - and a lot of it is down to how overt the randomness is as opposed to being a bit more obfuscated.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 17:17:17


Post by: Desubot


Define wargame and define skill.

Realistically an "Ideal" game the only thing that should determines the win should be the choices in game you make. aka terrain, and target or objective priorities.

But since dice rolls the stats and chances should generally even out so that your choices are the only things that make victory.
imho.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 17:34:35


Post by: Stux


I voted 70% for what it's worth. The worst player in the shop winning only 1/10 games is unacceptable in my opinion. 1/5 still not great. 3/10 seems ok.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 17:59:31


Post by: auticus


Man I'm reminded daily why I don't get along well with modern game design lol.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 19:00:56


Post by: BrianDavion


how MUCH better are we talking? are we talking a little better, a lot better? In my opinion ideally the better player should win almost every time but only if he's much much better


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 19:31:19


Post by: Elbows


As Desubot said and I referenced in an earlier response...define "Wargame".

Wargame is a somewhat confusing sobriquet. Is it a recreation of a conflict or a normal game using war as the genre/setting? Warhammer 40K is obviously the latter. You could replace all of the weapons with paintball guns and replace all the terrain with inflatable bunkers and the end result would more or less be the same. While there are nods to the historical wargaming genre, it's more competition/tournament/normal game based than anything else.

I would argue that an ideal game (not necessarily wargame) would of course be dictated by skill, or luck depending on whichever one is involved in the game design. Consider a normal boardgame. It's probably a mix of skill and luck (and very intentionally so). Consider chess, it's more skill than luck, etc.

A lot of competitive players decry the use of dice or "randomness", because that's more a function of a real wargame than chess. A narrative or historical gamer will appreciate the reality of dice/randomness more. However these two players have vastly different aims in most situations.

My ideal "wargame" is always Side A vs. Side B vs. Chaos (Chaos = weather, unforeseen events, random occurrences, things outside of the player's control etc.)

That means that my ideal wargame would never fit in a tournament or competitive setting because there would be too much teeth gnashing, and that's fine. So, as bizarre and fake as a game like 40K is (with matched armies played on little terrain with no environmental impacts, and a God-like mastery of your units) I'd still say an 80/20 is probably about right. Maybe as much as 90/10 if the rules/codices were better written.



In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 20:52:42


Post by: Xenomancers


LOL. What kind of skills are we talking about?
If it's something like kicking a ball or swining a bat. Something that is actually difficult and requires practice it should be about 100%. In a wargame though. There aren't real skills beyond the learning curve for the game.

In a game where you know the rules and your opponent doesn't you shouldn't even be trying to win that. You should be teaching them to play. If they know the rules but just suck at strategy in general you should win probable about 80%.

Outside of that anything more than a 60% WR should be unreplicatable in a game where everything else is equal but your "skill" if dice is involved.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 21:00:47


Post by: Bharring


Chess is a game that is actually difficult, and requires practice to play.

I know my win rate is over 0% versus people who are better than me, and below 100% versus people who are worse than me.

So, clearly, even in an extremely difficult and skill-intensive game, even one with no random factors, there's still an element of chance.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 21:02:50


Post by: NinthMusketeer


Higher skill doesn't mean you always play the better game; good players have bad games sometimes, and vice versa. And dice are important to mix different elements up so things don't become stale. I'd say 75% is a decent average.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 21:16:24


Post by: Desubot


 Xenomancers wrote:
LOL. What kind of skills are we talking about?
If it's something like kicking a ball or swining a bat. Something that is actually difficult and requires practice it should be about 100%. In a wargame though. There aren't real skills beyond the learning curve for the game.

In a game where you know the rules and your opponent doesn't you shouldn't even be trying to win that. You should be teaching them to play. If they know the rules but just suck at strategy in general you should win probable about 80%.

Outside of that anything more than a 60% WR should be unreplicatable in a game where everything else is equal but your "skill" if dice is involved.


There isnt that much "skill" involved in wargame. but there are things that you can do outside of list building to increase your odds.

proper positioning, use of terrain and cover, focusing priority targets (say the opponents main anti tank), planning out your moves to win the objectives and also quick math on odds to make better choices. non of these things besides maybe the last one is dice dependent.

its not a physical skill like boxing or juggling, but it is a mental skill that can be practices. not that any of these are particularly hard or exclusive to only the best of the best.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 21:42:43


Post by: JNAProductions


I voted 80%, but my real number is probably 85% or so.

Skill should matter a lot more than luck, but players make mistakes and dice can turn against you.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 22:02:00


Post by: puree


Level of skill difference is clearly important - a little bit better might mean that you are still likely to make almost as many mistakes that you can't recover from compared to the other person, a lot better means that the ratio of mistakes made will be heavily in your favour, and that it is very likely the other person will make the mistakes before you.

Dice may or may not be important, dice can be rolled quite a bit but the nature of the game means what you roll at any moment in time will not be that decisive for a good player. In others they may have a larger influence (or their may be random card draws etc which are in essence 'dice' in terms of a random element). In some games the key skill is not about having a plan (they are 10 a penny), but having a backup plan you can actually execute if it goes wrong and not just blaming the dice for your plan's failure.

Not all games have anything to do with list building, some may even have set scenarios (ASL) or a list building phase where there really isn't a huge difference beyond 'flavor'.

The only game I played enough to gauge win rate on a personal level was a certain tactical spaceship combat game based on star trek, which I played every week for some 5 years with a local group, plus online games/tourneys. Once past the newbie level luck didn't really play a large part in that game despite the dice, so any skill difference could be significant. In my local group my win rate was about 96-98% and a couple of those losses were uneven battles in campaigns, online tourneys probably about 75-80%? I wouldn't have said my local opponents were bad, and they certainly didn't lack experience, but it was very much a game that would see a quick ramp up of win ratio as the skill difference increased, not unlike chess.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 22:05:50


Post by: Peregrine


 Stux wrote:
I voted 70% for what it's worth. The worst player in the shop winning only 1/10 games is unacceptable in my opinion. 1/5 still not great. 3/10 seems ok.


Assuming that the worst player is genuinely a terrible player (and not just the guy who only finishes in the top 10 of major tournaments instead of winning them like the rest of the group) why shouldn't they lose 90% of the time or more? For a terrible player to win 30% of their games you have to have a game where RNG thoroughly invalidates player agency. And at that point why bother playing? Your decisions won't matter very much, so why not just roll a bunch of dice and see who rolls better?


Automatically Appended Next Post:
 Elbows wrote:
My ideal "wargame" is always Side A vs. Side B vs. Chaos (Chaos = weather, unforeseen events, random occurrences, things outside of the player's control etc.)


I will never understand the appeal of this. Sure, a heavy RNG element for environment/random chaos/etc may make a more accurate simulation of a battle, but it's a miserable experience as a game. When that kind of RNG is significant enough to bother with including you have a high chance of ending up with games where RNG determines a winner and playing out the game is pointless. So yeah, it might be an accurate simulation if half of your 40k army is consumed by a warp storm before the game begins, but then what is the point of playing 2000 points vs. 1000 points once RNG has decided that you lose? You might as well declare the winner at that point, set up another game, and hope that the RNG lets you actually play it.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 22:26:58


Post by: puree


I will never understand the appeal of this. Sure, a heavy RNG element for environment/random chaos/etc may make a more accurate simulation of a battle, but it's a miserable experience as a game. When that kind of RNG is significant enough to bother with including you have a high chance of ending up with games where RNG determines a winner and playing out the game is pointless. So yeah, it might be an accurate simulation if half of your 40k army is consumed by a warp storm before the game begins, but then what is the point of playing 2000 points vs. 1000 points once RNG has decided that you lose? You might as well declare the winner at that point, set up another game, and hope that the RNG lets you actually play it.


Undoubtedly you know that is pure hyperbole (as a response to what you replied to).

The first SPI game I got at the age of 12 was Rostov: The First Soviet Counter-Attack (german vs soviet 1941). It had random weather per day, random reinforcements, heck the soviet player didn't even know the value of his own units until they fought the first time - was the hastily thrown in infantry division going to melt away at first contact or be heroes of the soviet union. Plenty of random reflection of reality.

It was not in any way a miserable game, it was still a good game with plenty of tactics and strategy. The germans needing to race against an increasing soviet buildup and worsening weather starting from an overwhelming start position. The random elements reflected certain aspects of the operation - you knew the likely weather, but not what it would be day to day, you knew you were getting that armor division, but would it arrive tomorrow or in 2 days time, you knew another russian formation had been thrown in and the sort of range of values it might have but until first contact couldn't be sure what it would be.

These are issues which a good player can handle and plan for, just because you can't plan for a division arriving this turn doesn't mean you can't plan for it arriving at all. As noted above what can separate the good from bad player is the ability to have plans AND backup plans. If you didn't plan for the delay in extra troops you were fully aware was a 33% possibility that is your fault, if you were crippled by the snow that had a 50% probability today that was your fault.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 22:34:04


Post by: ERJAK


 Xenomancers wrote:
LOL. What kind of skills are we talking about?
If it's something like kicking a ball or swining a bat. Something that is actually difficult and requires practice it should be about 100%. In a wargame though. There aren't real skills beyond the learning curve for the game.

In a game where you know the rules and your opponent doesn't you shouldn't even be trying to win that. You should be teaching them to play. If they know the rules but just suck at strategy in general you should win probable about 80%.

Outside of that anything more than a 60% WR should be unreplicatable in a game where everything else is equal but your "skill" if dice is involved.


I think this might take the cake for 'dumbest thing I've heard on dakka'. I might sceenshot this for the 'gak dakka says' podcast I've been thinking about doing for a while now.

Kicking a ball and swinging a bat are things children do on accident.

The rest of this honestly says more about you than it does about wargaming. It seems like you've convinced yourself that skill is unattainable in a clearly skill based game to justify your own lack of success. This suggests that you are not one for accepting fault for your own failures.

I could spend a while tearing apart the 'well if they suck, 80% ' line but lets just hit the high points. 'If they suck' you mean like lacking SKILL in an activity leading to defeat at the hand of a more SKILLED opponent? '80%' If you're only beating someone who sucks 80% of the time, it's because you suck only slightly less. A good player shouldn't be losing to a below average player that often, let alone someone who sucks.


Automatically Appended Next Post:
 auticus wrote:
 timetowaste85 wrote:
Equally good armies that are tuned against each other, the higher skill player should win almost every time. Like 80% of the time. Random luck can happen, but if everything is equal except the skill, higher skill should win.


That is true. Equal armies do show you who is the better player, not the player who can build (or copy) the most efficient mathematical list.


Except being able to build the better list is PART of the skill aspect of the game. Using the exact same list also over-emphasizes list familiarity, playstyle affinity, and ESPECIALLY first turn advantage.

People who are bad tend to overcredit list strength (as a way to excuse their own lack of success) but the truth is at a level of competition where the games actually matter (top 16 at major tourneys) all the lists are pretty much the same powerlevel anyway.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 23:06:48


Post by: Stux


 Peregrine wrote:
 Stux wrote:
I voted 70% for what it's worth. The worst player in the shop winning only 1/10 games is unacceptable in my opinion. 1/5 still not great. 3/10 seems ok.


Assuming that the worst player is genuinely a terrible player (and not just the guy who only finishes in the top 10 of major tournaments instead of winning them like the rest of the group) why shouldn't they lose 90% of the time or more? For a terrible player to win 30% of their games you have to have a game where RNG thoroughly invalidates player agency. And at that point why bother playing? Your decisions won't matter very much, so why not just roll a bunch of dice and see who rolls better?


Because no one who invests significant money in this hobby should lose too much.

Yeah, ok, if they are truly awful - like a 7 year old playing against veteran adults - sure then it should be a higher chance of a loss. But randomness necessarily should be a big part of deciding the winner. Less significant than skill, but it needs to be up there.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 23:17:37


Post by: puree


Because no one who invests significant money in this hobby should lose too much.


Not sure what that has to do with an Ideal wargame. An ideal wargame is one you invest no money in

If people are losing a lot and that is an issue for them because it will cost money then they stop collecting/playing (ergo invest very little). If they continue to pay and play then it probably isn't an issue at all. Not everyone worries about win/loss ratios, things like collecting/painting/making dioramas (and the odd game you always lose) and just playing friendly games for the social interaction with your friends are also a factor in why we spend money.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/24 23:24:33


Post by: carldooley


I keep thinking about games like Warmachine or Hordes. The skill level in that game is one where the only real issue is keeping a new player's interest while acquiring the same level of knowledge as the general player base. Not enough screening? Dead warnoun. Positioning bad? Dead warnoun. The biggest game determinant is player skill, as most units and models can be used depending on the list. 40k on the other hand? Balanced by throwing lots of dice.

Dead warnoun? endgame.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:05:00


Post by: puree


 carldooley wrote:
I keep thinking about games like Warmachine or Hordes. The skill level in that game is one where the only real issue is keeping a new player's interest while acquiring the same level of knowledge as the general player base. Not enough screening? Dead warnoun. Positioning bad? Dead warnoun. The biggest game determinant is player skill, as most units and models can be used depending on the list. 40k on the other hand? Balanced by throwing lots of dice.

Dead warnoun? endgame.


At the other end, Frostgrave. A single dice (D20) roll off for most things with small modifiers = high variability and a game where skill seems to hardly matter. That has its upsides, you are seldom out of a game until the end and beginners have a decent chance fairly quickly. The downside is outwitting the opponent only to watch that D20 roll off wipe you out. Whilst I'm not so much a mini player (more board wargames) so haven't played a lot of mini games, Frostgrave is probably the most RNG over skill 'wargame' I've met. Although it is still quite fun.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:12:25


Post by: HoundsofDemos


I miss some of the more random elements from previous editions like scattering, terrain effects, vehicles getting stuck or models slowed down by terrain. How you react and recover from something not going quite as planned is a sign of skill and also leads to some memorable moments.



In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:25:33


Post by: Overread


 Peregrine wrote:



Automatically Appended Next Post:
 Elbows wrote:
My ideal "wargame" is always Side A vs. Side B vs. Chaos (Chaos = weather, unforeseen events, random occurrences, things outside of the player's control etc.)


I will never understand the appeal of this. Sure, a heavy RNG element for environment/random chaos/etc may make a more accurate simulation of a battle, but it's a miserable experience as a game. When that kind of RNG is significant enough to bother with including you have a high chance of ending up with games where RNG determines a winner and playing out the game is pointless. So yeah, it might be an accurate simulation if half of your 40k army is consumed by a warp storm before the game begins, but then what is the point of playing 2000 points vs. 1000 points once RNG has decided that you lose? You might as well declare the winner at that point, set up another game, and hope that the RNG lets you actually play it.[/quote

I think the best is to have a game where the core balance is Army A VS Army B and then you can add on top weather or wandering monsters or spells or such for the random chaos on top. Ergo make the core game balanced and then you can introduce fun other stuff on top as a bolt on extra. Sometimes its nice to spice things up having a wandering dragon in the middle of the battlefield; or have weather effects; or do a running chase etc..../


*Running chase - where every end of turn all the terrain moves X inches in the same direction to reflect the army doing a filght through a scene. So it might be bikes and such in 40K (old white dwarf had it with jetbikes years ago); or cavalry in AoS]


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:34:31


Post by: Eonfuzz


ERJAK wrote:
 Xenomancers wrote:
LOL. What kind of skills are we talking about?
If it's something like kicking a ball or swining a bat. Something that is actually difficult and requires practice it should be about 100%. In a wargame though. There aren't real skills beyond the learning curve for the game.

In a game where you know the rules and your opponent doesn't you shouldn't even be trying to win that. You should be teaching them to play. If they know the rules but just suck at strategy in general you should win probable about 80%.

Outside of that anything more than a 60% WR should be unreplicatable in a game where everything else is equal but your "skill" if dice is involved.


I think this might take the cake for 'dumbest thing I've heard on dakka'. I might sceenshot this for the 'gak dakka says' podcast I've been thinking about doing for a while now.

Kicking a ball and swinging a bat are things children do on accident.

The rest of this honestly says more about you than it does about wargaming. It seems like you've convinced yourself that skill is unattainable in a clearly skill based game to justify your own lack of success. This suggests that you are not one for accepting fault for your own failures.

I could spend a while tearing apart the 'well if they suck, 80% ' line but lets just hit the high points. 'If they suck' you mean like lacking SKILL in an activity leading to defeat at the hand of a more SKILLED opponent? '80%' If you're only beating someone who sucks 80% of the time, it's because you suck only slightly less. A good player shouldn't be losing to a below average player that often, let alone someone who sucks\


I actually almost agree with Xeno here, the amount of "skill" involved in most wargames is equivalent to the amount of skill in card games, ie, hearthstone.
There really isn't much, and as long as you understand the rules and rule interactions you can netlist and netgame the rest.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:36:35


Post by: JNAProductions


 Eonfuzz wrote:
ERJAK wrote:
 Xenomancers wrote:
LOL. What kind of skills are we talking about?
If it's something like kicking a ball or swining a bat. Something that is actually difficult and requires practice it should be about 100%. In a wargame though. There aren't real skills beyond the learning curve for the game.

In a game where you know the rules and your opponent doesn't you shouldn't even be trying to win that. You should be teaching them to play. If they know the rules but just suck at strategy in general you should win probable about 80%.

Outside of that anything more than a 60% WR should be unreplicatable in a game where everything else is equal but your "skill" if dice is involved.


I think this might take the cake for 'dumbest thing I've heard on dakka'. I might sceenshot this for the 'gak dakka says' podcast I've been thinking about doing for a while now.

Kicking a ball and swinging a bat are things children do on accident.

The rest of this honestly says more about you than it does about wargaming. It seems like you've convinced yourself that skill is unattainable in a clearly skill based game to justify your own lack of success. This suggests that you are not one for accepting fault for your own failures.

I could spend a while tearing apart the 'well if they suck, 80% ' line but lets just hit the high points. 'If they suck' you mean like lacking SKILL in an activity leading to defeat at the hand of a more SKILLED opponent? '80%' If you're only beating someone who sucks 80% of the time, it's because you suck only slightly less. A good player shouldn't be losing to a below average player that often, let alone someone who sucks\


I actually almost agree with Xeno here, the amount of "skill" involved in most wargames is equivalent to the amount of skill in card games, ie, hearthstone.
There really isn't much, and as long as you understand the rules and rule interactions you can netlist and netgame the rest.
Okay. But this isn't about 40k as it stands now, this is about a hypothetical IDEAL 40k, or other wargame.

Chess is, for instance, a wargame. Would you say there's no skill involved in chess?


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:40:38


Post by: Eonfuzz


 JNAProductions wrote:

Okay. But this isn't about 40k as it stands now, this is about a hypothetical IDEAL 40k, or other wargame.

Chess is, for instance, a wargame. Would you say there's no skill involved in chess?


Chess is less of a wargame and more of a logic game. But in any case I'd say yes, there is skill involved in chess.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:43:01


Post by: JNAProductions


So, in a hypothetical version of 40k that's as close to the perfect wargame as possible, should player skill be a factor beyond just basic rules knowledge?


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 00:47:19


Post by: Eonfuzz


 JNAProductions wrote:
So, in a hypothetical version of 40k that's as close to the perfect wargame as possible, should player skill be a factor beyond just basic rules knowledge?


See, I'm not sure. I've played a LOT of wargame tabletops and most of the "skill" revolves around building the right list and knowing the rule minutiae.
(And no, list building isn't a showcase of skill)


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 01:28:12


Post by: auticus


People who are bad tend to overcredit list strength (as a way to excuse their own lack of success)


Listbuilding is by far not a display of skill to me. Its a display of understanding middle school math and knowing that the guy with 5 attacks at S5 and costs 80 points is superior to the guy with 5 attacks at S4 but who costs 75 points. Or in GW's case, the guy with 3 attacks at S4 who for whatever reason costs 120 points next to the 80 point guy with 5 S5 attacks. Because he has some super cool cinematic power that works in jungle terrain that is likely to never be encountered.

Taken to a final conclusion with a spreadsheet. Anyone that has any small amount of Algebra experience can macro code an excel spreadsheet to break those numbers down into a power coefficient.

I know. I did that for ten years. And placed high at the GT level. So I know at least for me, I'm not downplaying listbuilding because "i'm bad at the game and need an excuse". I played at the GT level, had several high places, and the lists were crutches for pretty much everyone in attendance.

My first GT experience was Baltimore in 1999 and I had been playing 40k a year and I placed near the top 10 with eldar. Because my list was busted, and I copy/pasted it off of Portent.net and got lucky I was fed space marine opponents the entire event. I thought I was pretty hot **** back then though and would have jumped on anyone saying listbuilding was not a skill. But really it wasn't, and when you took my busted lists away from me I am a 50/50 player that wins as much as they lose.

And since 2000 or so lists are as easy as copy/paste. As far as LISTBUILDING is concerned, that makes it a non-skill to me.

Now for anyone who would challenge others by saying listbuilding is a skill and only people that are bad would say its not, I challenge them to hit up adepticon or the LVO with a middle of the road list or a close to garbage list and then go on and win or place very high in the event, and at that point I'll believe that listbuilding isn't that big of a deal in its weight and impact on the game. There have been a small handful of people that have done that, and those players are really good. They are a very tiny percentage of the tournament population as well.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 01:29:42


Post by: Jerram


I initially said 80% but out of curiosity I went and looked at some professional sports leagues and i could only find one where the team with the best season record won 80% of their games (NFL) and it also happened to be one of the smaller number of games played at 16, the two leagues with 82 games played were both mid 70% and the league with over 160 games played had the bext record at 67% winning record so maybe 70% is a better number. Of course those are team sports, would be interesting to see something similar for something like tennis.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 03:15:00


Post by: Peregrine


Jerram wrote:
I initially said 80% but out of curiosity I went and looked at some professional sports leagues and i could only find one where the team with the best season record won 80% of their games (NFL) and it also happened to be one of the smaller number of games played at 16, the two leagues with 82 games played were both mid 70% and the league with over 160 games played had the bext record at 67% winning record so maybe 70% is a better number. Of course those are team sports, would be interesting to see something similar for something like tennis.


That's not a very good comparison because professional leagues only have the absolute best players with a very small skill gap. Even the worst professional team is still the elite of the elite in the sport as a whole. A more relevant win rate would be, say, having an NBA team play 82 games of basketball against a random high school team. The NBA team is almost certainly going to win every single game by massive margins, and the lesser team being able to score at all is probably going to require some fluke luck. And god help the amateurs who have to play against an NFL team, as the entire team will probably end up in the hospital.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 04:20:13


Post by: argonak


so I said 70%, but if the game was better designed I'd think 100% was fair, as long as GW included some sort of built in handicap option for players to use when going up against someone they know is worse off than them.

When I was a kid, if my dad beat me in a game of chess he's spot me a piece for the next game. If he beat me again, he'd spot me another piece, and so on and so on until I won. Then it'd go the other way. It kept the game fun and a challenge for both of us, rather than being boring for one and miserable for the other.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 05:30:43


Post by: Apple fox


puree wrote:
 carldooley wrote:
I keep thinking about games like Warmachine or Hordes. The skill level in that game is one where the only real issue is keeping a new player's interest while acquiring the same level of knowledge as the general player base. Not enough screening? Dead warnoun. Positioning bad? Dead warnoun. The biggest game determinant is player skill, as most units and models can be used depending on the list. 40k on the other hand? Balanced by throwing lots of dice.

Dead warnoun? endgame.


At the other end, Frostgrave. A single dice (D20) roll off for most things with small modifiers = high variability and a game where skill seems to hardly matter. That has its upsides, you are seldom out of a game until the end and beginners have a decent chance fairly quickly. The downside is outwitting the opponent only to watch that D20 roll off wipe you out. Whilst I'm not so much a mini player (more board wargames) so haven't played a lot of mini games, Frostgrave is probably the most RNG over skill 'wargame' I've met. Although it is still quite fun.


We have end up returning to other games over frostgrave. It was a long paranoid mess of a game, Where even fun mechanics could be such a huge detriment we stopped using them.
You can also due to the way some things work, Completely break the game.
It has a huge skill gap, with the RNG mostly only matter at the low end. But with some things really easy to abuse if done right. I think it needs a second go at the rules as much as i think 40k does D:


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 06:23:36


Post by: skchsan


In an ideal 40k set up, meaning the game hasn't reduced itself to "cookiecutter powergaming netlist where it doesnt matter who you kill first since 70% of army is getting wiped first turn anyways" situation, it comes down to LOS denial, funneling, and target selection as far as "skills" go. Common "skills" some people refer to such as screening, reserves, and DS denial happens at list building - so if you count that as "skills," so be it.

Terrain is largely underrated (rightfully so as 40k terrain ruleset is wonky at best, counter productive at worst) in a game of 40k. With proper terrain set up, you will be able to curtail your enemies' shooting capacity by a large degree, forcing them out of position just to be able to shoot at you. Forcing them OOP can turn the tides for you by creating space for DS or fast units to squeeze into, and even force them off objectives.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 06:45:47


Post by: stratigo


puree wrote:
I will never understand the appeal of this. Sure, a heavy RNG element for environment/random chaos/etc may make a more accurate simulation of a battle, but it's a miserable experience as a game. When that kind of RNG is significant enough to bother with including you have a high chance of ending up with games where RNG determines a winner and playing out the game is pointless. So yeah, it might be an accurate simulation if half of your 40k army is consumed by a warp storm before the game begins, but then what is the point of playing 2000 points vs. 1000 points once RNG has decided that you lose? You might as well declare the winner at that point, set up another game, and hope that the RNG lets you actually play it.


Undoubtedly you know that is pure hyperbole (as a response to what you replied to).

The first SPI game I got at the age of 12 was Rostov: The First Soviet Counter-Attack (german vs soviet 1941). It had random weather per day, random reinforcements, heck the soviet player didn't even know the value of his own units until they fought the first time - was the hastily thrown in infantry division going to melt away at first contact or be heroes of the soviet union. Plenty of random reflection of reality.

It was not in any way a miserable game, it was still a good game with plenty of tactics and strategy. The germans needing to race against an increasing soviet buildup and worsening weather starting from an overwhelming start position. The random elements reflected certain aspects of the operation - you knew the likely weather, but not what it would be day to day, you knew you were getting that armor division, but would it arrive tomorrow or in 2 days time, you knew another russian formation had been thrown in and the sort of range of values it might have but until first contact couldn't be sure what it would be.

These are issues which a good player can handle and plan for, just because you can't plan for a division arriving this turn doesn't mean you can't plan for it arriving at all. As noted above what can separate the good from bad player is the ability to have plans AND backup plans. If you didn't plan for the delay in extra troops you were fully aware was a 33% possibility that is your fault, if you were crippled by the snow that had a 50% probability today that was your fault.
\

This is what we'd call a narrative campaign in 40k


Automatically Appended Next Post:
 auticus wrote:
People who are bad tend to overcredit list strength (as a way to excuse their own lack of success)


Listbuilding is by far not a display of skill to me. Its a display of understanding middle school math and knowing that the guy with 5 attacks at S5 and costs 80 points is superior to the guy with 5 attacks at S4 but who costs 75 points. Or in GW's case, the guy with 3 attacks at S4 who for whatever reason costs 120 points next to the 80 point guy with 5 S5 attacks. Because he has some super cool cinematic power that works in jungle terrain that is likely to never be encountered.

Taken to a final conclusion with a spreadsheet. Anyone that has any small amount of Algebra experience can macro code an excel spreadsheet to break those numbers down into a power coefficient.

I know. I did that for ten years. And placed high at the GT level. So I know at least for me, I'm not downplaying listbuilding because "i'm bad at the game and need an excuse". I played at the GT level, had several high places, and the lists were crutches for pretty much everyone in attendance.

My first GT experience was Baltimore in 1999 and I had been playing 40k a year and I placed near the top 10 with eldar. Because my list was busted, and I copy/pasted it off of Portent.net and got lucky I was fed space marine opponents the entire event. I thought I was pretty hot **** back then though and would have jumped on anyone saying listbuilding was not a skill. But really it wasn't, and when you took my busted lists away from me I am a 50/50 player that wins as much as they lose.

And since 2000 or so lists are as easy as copy/paste. As far as LISTBUILDING is concerned, that makes it a non-skill to me.

Now for anyone who would challenge others by saying listbuilding is a skill and only people that are bad would say its not, I challenge them to hit up adepticon or the LVO with a middle of the road list or a close to garbage list and then go on and win or place very high in the event, and at that point I'll believe that listbuilding isn't that big of a deal in its weight and impact on the game. There have been a small handful of people that have done that, and those players are really good. They are a very tiny percentage of the tournament population as well.


I mean, there's some skill in throwing up that excel sheet. It'd take me a while to do it since I'm bad at excel and algebra.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 07:12:02


Post by: BrianDavion


I do feel obliged to point out that the more granularity you add to the game that make skill matter, the harder it is for the game to be played at a large scale. There are games out there with a much higher degree of tactics then 40k absolutely, but trying to play them at the scale 40k is played at would be... difficult. a game involving a small number of units can make for a much greater amount of depth, with things like range envolvopes, speed, terrain, angle of fire etc all being more possiable.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 07:28:19


Post by: Nibbler


As a fun-player, who likes to laugh and joke with the other people around the table (I don't like the word opponent in that context), I love random events.
There was this game (I think it was Judge Dredd) were every roll of six caused a random event and it was nuts...
You get to see absurd actions, heroic deeds and crazy kills - that's what I'm up to.
Such a game messes with things like skill and win-greed and therefor it can be quite frustrating for someone who takes it to serious.

On the other hand, I understand (and sometimes also like) some kind of "professional" comparing of tactics (don't nail me down on this word please) and skills - in these games there is no room for those big random incidents, because they anticipate almost every seroius competition...

So, I find it to be a difficult question without marking the context any further.
Perfect wargame for funny and crazy evenings / weekends / whatever? For me it's big bunch of random events you can't controll completely

Perfect wargame to simulate a serious battle between equal (on the powerlevel side) forces? Tone down the randomness to a minimum and let the player controll what happens on the battlefield


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 10:27:32


Post by: Breng77


I need an it depends on the difference in skill level. I think there are some players who will swap wins most of the time because they are of similar skill, and others matchups where one player nearly always wins.

As to those saying there is no skill in wargames, that is 100% false, there are plenty of games where model positioning matters, quick assessment of chance of success matters, strategy with regards to the mission matters, etc

How much those things matter depends on the rules of the game, in games with constant win conditions (like tabling) the skill is lower, in those with more objective based win conditions it is higher. Any game with more need for movement will have a higher skill level.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 11:11:14


Post by: auticus


I mean, there's some skill in throwing up that excel sheet. It'd take me a while to do it since I'm bad at excel and algebra.


Fair enough, but if you look hard enough you can also find tools online that help identify power coefficients or find someone that is good at writing tools and math to make one for you. Essentially listbuilding as a "skill" is laughable. If when I was an athlete I could go take a pill and instantly get several years worth of weight training and endurance without having to do the work, that would have gone a long way for me. I could have then just focused on "the skill" of the game.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 11:26:25


Post by: Nithaniel


Breng77 wrote:


As to those saying there is no skill in wargames, that is 100% false, there are plenty of games where model positioning matters, quick assessment of chance of success matters, strategy with regards to the mission matters, etc


We are definitely missing an agreement on what skill is. Its all of the above and then some. In high level games the variability in lists should be minimal like tournament final tables so what is the difference in skill? Part of this is making mistakes. I've played a lot of chess and as the game develops the end winner is the one who didn't make as many or any mistakes. A perfect example of this is the LVO 2018 final between Nick and Tony which was iirc a mirror match. In this game Tony made a mistake in order of processing stratagems with movement and lost him a game that he looked like he was going to win. At the time Nick was touted as the best player in the world (debatable) but he was losing before the mistake.

Assuming lists are not skill, what separates the higher level players from the intermediate is their abilities to make mistakes under pressure.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 11:47:24


Post by: small_gods


 Nithaniel wrote:
Breng77 wrote:


As to those saying there is no skill in wargames, that is 100% false, there are plenty of games where model positioning matters, quick assessment of chance of success matters, strategy with regards to the mission matters, etc


We are definitely missing an agreement on what skill is. Its all of the above and then some. In high level games the variability in lists should be minimal like tournament final tables so what is the difference in skill? Part of this is making mistakes. I've played a lot of chess and as the game develops the end winner is the one who didn't make as many or any mistakes. A perfect example of this is the LVO 2018 final between Nick and Tony which was iirc a mirror match. In this game Tony made a mistake in order of processing stratagems with movement and lost him a game that he looked like he was going to win. At the time Nick was touted as the best player in the world (debatable) but he was losing before the mistake.

Assuming lists are not skill, what separates the higher level players from the intermediate is their abilities to make mistakes under pressure.


I find this to be a vastly underestimated part of the game. A lot of people at tournaments have a reasonably high skill level but it is the ability to keep all that in your head whilst trying to think through then next steps you need to make where the real skill lies.

A lot of people can screen well, play the objectives well, have decent target prority and know how to limit your opponets options. But to be able to do all that simultaneously, under a time pressure, after playing over long periods is where the real skill lies.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 11:56:39


Post by: Breng77


 Nithaniel wrote:
Breng77 wrote:


As to those saying there is no skill in wargames, that is 100% false, there are plenty of games where model positioning matters, quick assessment of chance of success matters, strategy with regards to the mission matters, etc


We are definitely missing an agreement on what skill is. Its all of the above and then some. In high level games the variability in lists should be minimal like tournament final tables so what is the difference in skill? Part of this is making mistakes. I've played a lot of chess and as the game develops the end winner is the one who didn't make as many or any mistakes. A perfect example of this is the LVO 2018 final between Nick and Tony which was iirc a mirror match. In this game Tony made a mistake in order of processing stratagems with movement and lost him a game that he looked like he was going to win. At the time Nick was touted as the best player in the world (debatable) but he was losing before the mistake.

Assuming lists are not skill, what separates the higher level players from the intermediate is their abilities to make mistakes under pressure.


yes that is also a skill, being able to make good decisions under fatigue and pressure. Part of not making mistakes is also understanding what your opponents options are and weighing how your moves will interact with those options. With a time limit (especially if the game is clocked in some way) your ability to rapidly make good decisions is also a skill. There is also practiced skill, if you think about sports and why some guys have say a really good jump shot it is part natural ability and huge part repetition in practice. Top players in general play a lot more games than others. At my competitive peak (generally winning more than losing at big GTs, never near the top) I was playing about one game a week, and usually against the same small group of players. SO I was not always familiar with how every opposing army played, so it wasn't second nature to make certain moves.

My larger point was that those that believe that no skill is involved in wargaming are delusional, or are players who themselves are not skilled but don't want to admit that better players are better than them for objective reasons like skill.



In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 13:28:16


Post by: Bharring


I think this thread has warped my opinion a bit.

I think what the ideal wargame is varies between people.

On one end of the spectrum, people want a narrative spectacle to see their minis have fun. The absolutest take on that would be a 50% win rate, regardless of skill (beyond basic game knowledge). The idea being, the rules are there to provide a spectacle. It's not a challenge - it's an experience.

On the other end of the spectrum, people want a competition. They want a forum for proving superiority. They want a game that is tactically or strategically deep (or both), so they can measure their ability against an opponent. The absolutest take on this would be a 100% win rate for any difference in skill. The idea being, we're here to measure ability. And the further from 100% we get, the less meaningful the results are.

Now, the first group will call the second group WAACs. And the second group will call the first group CAACs. Neither is accurate. The first group is less competitive, and more into wargames as a venue for expression. The second group is more competitive, and more into wargames as a way to prove ability. So neither side is "right" or "wrong" - just different concepts.

To make matters more confusing, most people aren't 100% in either category. Most people sit along the spectrum. Most people see value in both propositions - only the most extreme only care about one set (and I'd argue those people are better off in other hobbies, but that's another subject).

My ideal would be 65% when playing casually, or 75% when playing competitively; so there's a notable difference in outcome based on skill, but the game can take you into weird places you didn't expect based on dice.

I like that random chance can mean I wind up in strange situations like a bunch of Fire Warriors charging Marines (done that successfully before), or a single Chapter Master holding up a Wraithlord, Avatar, Phoenix Lord, and Banshee squad for most of the game (also happened to me). In both cases, these made for much more interesting games. But, on the other hand, I've loved games where I've been able to perfectly pick apart the opponent, such that I'm only really facing a quarter of their army at a time, and absolutely rock. In other words, 40k as-is provides a venue for both interests. I'm not saying it's perfect, but I am saying it's good.

So my answer is "Nowhere close to 100% or 50%".


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 13:49:52


Post by: Stux


Nail on the head for me.

Basically there is no single ideal wargame, because people want different things (and often don't actually know what they want).


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 14:03:13


Post by: small_gods


Yeah second that from me. I've always enjoyed the challange of playing competitively but understand that's not what everyone wants.

I have a friend who will only play single codex and no named characters as he finds it unrealistic that Ahriman would be fighting small skirmishes with just a handful of thousand sons and some plaguebearers. And this will ruin the enjoyment of the game for him.

The benifit of 40k is it allows for both those perspectives if both people playing are on the same page.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/25 19:11:55


Post by: Insectum7


 small_gods wrote:

The benifit of 40k is it allows for both those perspectives if both people playing are on the same page.

^ Very much so. It's fantastic in that regard.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 02:08:50


Post by: skchsan


If a proper wargame requires 0 skills, then I'd argue that poker isn't a game of skills but that of pure luck - which if you know even a little bit about poker you'd know that this isn't true.

Playing the odds and manipulating the battlefield to lever your chances is what differentiates between a novice and experienced players.

The problem is skills don't matter as much anymore in 40k because there are certain delete buttons in the game that outright ignore all tactics.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 05:49:52


Post by: Breton


momerathe wrote:
Different people have different approaches to the game - some people enjoy cutthroat competition, some are into it for beer and pretzels (soft drinks for minors ). I'm imagining that hardcore competitive gamers are more near the top, and more casual gamers are further down the scale, but I'd be interested to see what the range of opinions are.

(obviously it depends on the size of the skill difference, but for a baseline consider the people who show up at your local FLGS)

I was particularly thinking about this while listening to a tournament podcast - the ITC rules seem pretty efficiently designed as a sorting algorithm for selecting tournament winners (allowing for matchup etc.); much shade is thrown on maelstrom for being too random (although not just for that), but maybe it's just a question of horses for courses.


There isn't a good answer. Was Henry the better skilled general in the Battle of Agincourt, or were the French just unlucky with the weather? Was Ike better than von Rundstedt or hampered by less command authority? In theory an ideal wargame would incorporation these sort of random obstacles, and provide a way to judge each army's ability to overcome or abuse these advantages to compare if General A better abused the advantage, or General B better resisted it regardless of who won or lost "the battle".


Automatically Appended Next Post:
 skchsan wrote:
If a proper wargame requires 0 skills, then I'd argue that poker isn't a game of skills but that of pure luck - which if you know even a little bit about poker you'd know that this isn't true.

Playing the odds and manipulating the battlefield to lever your chances is what differentiates between a novice and experienced players.

The problem is skills don't matter as much anymore in 40k because there are certain delete buttons in the game that outright ignore all tactics.


The difference between skill and luck is usually the person writing the After Action Report.

Taffy 3 skillfully ran into a storm squall to avoid enemy fire.

The Japanese were unable to fire on the American carriers after they were unluckily cursed with bad weather and low visibility

Daniel Negraneau skillfully baited Phill Helmuth to all-in call his straight, before filling out his nut-flush draw on the river.

Phill Helmuth called Daniel Negraneau's flush draw before a catastrophic turn of Lady Luck's hand filled in Daniel's flush.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 07:05:04


Post by: Moriarty


Skill is calling the shot . . . and making it.

Luck is making the shot _without_ calling it.


Automatically Appended Next Post:
Skill will only get you so far as a general. Once the die are cast, you have no input to the result.

So saying, a skilled general can reduce the chances of bad luck being fatal.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 08:11:14


Post by: Jidmah


 Stux wrote:
I voted 70% for what it's worth. The worst player in the shop winning only 1/10 games is unacceptable in my opinion. 1/5 still not great. 3/10 seems ok.


Agree with this. Studies have shown that one of the major reasons why MtG has stayed afloat while most other TCGs died at some point is that even with a high difference in skill-levels any player wins the game sometimes. This is what makes Beer&Prezel gaming possible, because some people just have more innate talent for strategy than others.

There is no beer&prezel chess.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 08:31:12


Post by: Overread


 Jidmah wrote:
 Stux wrote:
I voted 70% for what it's worth. The worst player in the shop winning only 1/10 games is unacceptable in my opinion. 1/5 still not great. 3/10 seems ok.


Agree with this. Studies have shown that one of the major reasons why MtG has stayed afloat while most other TCGs died at some point is that even with a high difference in skill-levels any player wins the game sometimes. This is what makes Beer&Prezel gaming possible, because some people just have more innate talent for strategy than others.

There is no beer&prezel chess.


This is true, but at the same time, from my own experiences, whilst everyone might get a chance at winning in MTG matches; when its the beginner against an experienced player the beginner tends to only win if the experienced player gets shafted by their shuffling. Ergo you play the game and they get 1 land card the whole match; or they keep drawing all their expensive stuff or none of their combo cards etc... Ergo the less experienced player can see that the experienced player is losing not because of skill, but because the luck of the draw is purely screwing them over for one match.

So in my view its a victory, but its a hollow victory. Otherwise the general state of affairs is the better built better played deck will win a vast majority of the time. One bonus though is that matches are faster; much faster. So whilst there's the same skill difference it can get lost because, provided the group is large enough, players will settle with those of a similar skill level and should get enough matches of variety to win some and lose some.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/26 08:49:12


Post by: Jidmah


Lots of people are content with winning through luck and many inexperienced players will not be able to tell that they just won because of luck.
The kind of people you are referring to are the competitive guys (MtG calls them "Spikes"), those tend to take losses as a challenge and will get better and start winning games eventually.

In a beer&prezel setting winning is not that important, but I have not seen a single person stay in the game after losing 20 games in a row. Even the most casual fluff-bunny will complain about not having won a game in a long time eventually, people just work that way.

Deck building is an issue obviously, but so is army building in 40k. I've had so many games where Mortarion hulk-smashed the entire enemy army because they thought four lascannons and a combi-melta were enough for anti-tank.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 10:48:21


Post by: Nithaniel


I like the idea of this poll in the context of tournament gaming. With BCP and other data sources and 40kstats.com and chapter tactics reviewing the details of tournament stats it raises the idea of probabilities of victory between 2 competent players based on their win/loss records.

We are not far away from having statistical assessments of players in big matchups. The e-sports thing could be quite entertaining. Obviously sample sizes aren't big enough yet.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 13:35:53


Post by: Voss


60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 14:38:05


Post by: Overread


Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Yes and no - in an ideal world a highly experienced player playing against a beginner will:

a) Play with a handicap (eg less points) to level the playing field somewhat.

b) Be playing introduction games where who wins doesn't matter and the armies might not even be proper armies as they are showing mechanics, tactics, theories,

c) Play but not play at their best- they might use a weaker list for example.

d) Both go into it but the beginner knows they are going to lose, but is able to view how their opponent plays and use it as a teaching mechanic - seeing how they lose and what works and what doesn't


In theory and in an ideal world a pro and beginner shouldn't be playing each other regularly and playing their best game without any handicap in place. Otherwise I agree a pro would nearly always win against the beginner and that will put the beginner off.
However if the game is based on skill more than luck then the beginner has a more even chance to improve their game over time and rise up. If its far more luck than skill then the beginner will find it a LOT harder to improve their game because the random swing of luck will be a far bigger element. This might mean that good choices and bad chioces get confused with bad and good luck swings.


Eg: Archers VS Cavalry in a close combat situation

Luck based game : The archers having a 60% chance to win (ergo nearly 50-50) means that the bad choice of allowing archers to be charged in close combat might not always appear so bad if the archers get enough wins. With a high random bar a player might not even see the archers ever lose for several games in a row. It's still a bad tactical idea, but because the random is so random it makes it hard for them to learn

Skill based game : The archers having a 90% chance to lose means that the beginner loses their archers nearly every single time. However this makes the lesson really clear and easy for them - archers don't work against cavalry in close combat. Now they can use that along with other things they learn to improve their game. For example they might now learn to screen their archers with some infantry - and in theory will learn to use spear infantry (anti-cavalry) over using, say, swordsmen.


As you can see, the skill based approach makes it a lot easier to see patterns in play and evolve and advance one's own game. With a heavy luck based system the win Vs loss might even out a bit more; but at the same time the player can't so easily learn because good and bad choices are hidden behind a lot more random rolling.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 15:46:10


Post by: Peregrine


Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 17:08:38


Post by: Wayniac


 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?
well i mean we have Warhammer....

But this would be preferable to the "you're going to lose 30 games before you can maybe git gud" like warmahordes has. Talk about a good way to make a newbie give up.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/27 17:11:13


Post by: Peregrine


Wayniac wrote:
But this would be preferable to the "you're going to lose 30 games before you can maybe git gud" like warmahordes has.


Not really. A brutal learning curve sucks for new players but at least there's the potential of a good game once you overcome it. A game that is 95% RNG with minimal player agency is never going to be anything but passively rolling dice and waiting to see what RNG decides.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 00:46:34


Post by: ERJAK


Wayniac wrote:
 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?
well i mean we have Warhammer....

But this would be preferable to the "you're going to lose 30 games before you can maybe git gud" like warmahordes has. Talk about a good way to make a newbie give up.


Warhammer is not at all that. There's a reason the same 20 people podium basically every major event despite half the field having basically the same type of list.

And losing 30 games before you can maybe git gud is what happens in EVERY game and sport where you play a better player. How many older brothers have kicked their younger brother's into the dirt for YEARS on the basketball court or w/e?

Losing when you're not as good as someone at something is just what happens in reality.

I also really don't like that you're basically assuming every loss to be worthless. You can absolutely have fun in a game you lose and even if you don;t have 'fun' you can still gain a lot in terms of knowledge and experience.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 07:10:41


Post by: Apple fox


Wayniac wrote:
 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?
well i mean we have Warhammer....

But this would be preferable to the "you're going to lose 30 games before you can maybe git gud" like warmahordes has. Talk about a good way to make a newbie give up.


40k has loads of this in it, good players win more often. and in this its not that much different to warmahordes at all. one of the big difference is the reach of 40k. making the shooting in 40k far less reliant on movement. It kinda just means in the end that close combat will never really be good outside of OP options. But the more RNG just means less player thought, and more boring games. But 40k, even in its worst state has never really hit that point.
But as you try and remove the skill advantage, what tends to happen is better players still win. But the game becomes far less interactive, Even something like target priority can make a huge difference in Completely mirror army.
And i not sure many players would play a game where the first turn decides every game on the single roll without insanely bad luck.

Its rare in Warmahordes for the first turn to wipe out even entire units. But in 40k that first turn advantage can be devastating enough already.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 07:57:02


Post by: Wibe


Depends on how much higher skill we are talking about.
If it's a high-end competitiv player trying to win VS a new, or completely useless player, then it should be close to 99.99%+ (it's still a dice game, so evertrying "can" happen). But it should be some variations to the win rate if the skill levels are more equal.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 16:57:36


Post by: solkan


Try presenting an mechanism for determining which player has the higher skill level, and then defending that definition against review.

You can even suppose that you're given the typical asymmetry in the game system:
- Archers and the ranged units are good at shooting things, and bad at melee
- Cavalry are good at melee and good at crushing archers when they get there
- Melee screening units defend archers against cavalry
- In order to prevent things becoming an archery competition, the army has to move forward to claim objectives in the middle of the table

So eventually everyone agrees on the optimal mix of those units to use to achieve battlefield objectives. So the natural progression is that people determine what the counter-choices against the optimal mix are, and realize that the counter choices aren't the same as the optimal mix. Now two players sit down and the outcome for each player is determined by a significant factor outside of their control--what their opponent chose to do.

What does "more skill" mean in that situation?

This message has been brought to you by the Committee of People Who Have Listened to People Argue About Skill in Rock Paper Scissors (a game, at best, about predicting arbitrary choices). With a letter of support from Tic-Tac-Toe (where player skill is sufficient to solve the game and determine its outcome).


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 19:08:41


Post by: Overread


ERJAK wrote:
Wayniac wrote:
 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?
well i mean we have Warhammer....

But this would be preferable to the "you're going to lose 30 games before you can maybe git gud" like warmahordes has. Talk about a good way to make a newbie give up.


Warhammer is not at all that. There's a reason the same 20 people podium basically every major event despite half the field having basically the same type of list.

And losing 30 games before you can maybe git gud is what happens in EVERY game and sport where you play a better player. How many older brothers have kicked their younger brother's into the dirt for YEARS on the basketball court or w/e?

Losing when you're not as good as someone at something is just what happens in reality.

I also really don't like that you're basically assuming every loss to be worthless. You can absolutely have fun in a game you lose and even if you don;t have 'fun' you can still gain a lot in terms of knowledge and experience.


Also not every game will be against he pro playing their best with evenly pointed teams.
You might well play other less experienced players; the pro might play with a handicap at times to balance out the skill difference; the pro might throw a game or two here or there casually (not always the best policy though). Etc....
As Peri and others have said, in most games if you're new and inexperienced you are going to lose to the more experienced most times. Games are just the same and by and large most people can learn to overcome the barriers. The issues are more social ones where pros refuse to teach newbies or where the game gets an experience gap - ergo lots of pros and then a huge gap until you hit beginners resulting in a huge skill jump and often to few beginners lasting to intermediate.


One might argue that Warmachine's issue is that the system doesn't lend itself well to an intermediate ranking; or that the difference between new and intermediate is too great a skill gap. Resulting in a larger gap for the beginners to jump which means if they don't join en-mass they can find it hard to get into the game.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/28 23:07:02


Post by: liquidjoshi


Can't remember the exact source, but I seem to remember reading a study that said if you want to keep someone invested in a new game or activity, then they need to succeed (on average) around 33% of the time. Take it much lower than that and they'll tend to lose interest. With that in mind, 70% seems to be the optimum balance. Player choice is still relevant, but random chance can swing the balance as well. Neither is totally dominant, which seems to be the key for many players.

How you convert that into an actual gaming system, however, is another mystery entirely.

As a side note, if GW had kept Fantasy on and made its ruleset competitive focussed (80-90% player skill) while introducing AoS as a peer and pretzels (50-60% player skill), they could have been onto a winning formula. Trying to cram both a casual and competitive focus into one system simultaneously, IMO, was never going to work well, and Dakka has the threads to prove it.

TL;DR: 70% because science.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 03:41:43


Post by: w1zard


 liquidjoshi wrote:
Can't remember the exact source, but I seem to remember reading a study that said if you want to keep someone invested in a new game or activity, then they need to succeed (on average) around 33% of the time. Take it much lower than that and they'll tend to lose interest. With that in mind, 70% seems to be the optimum balance. Player choice is still relevant, but random chance can swing the balance as well. Neither is totally dominant, which seems to be the key for many players.

How you convert that into an actual gaming system, however, is another mystery entirely.

As a side note, if GW had kept Fantasy on and made its ruleset competitive focussed (80-90% player skill) while introducing AoS as a peer and pretzels (50-60% player skill), they could have been onto a winning formula. Trying to cram both a casual and competitive focus into one system simultaneously, IMO, was never going to work well, and Dakka has the threads to prove it.

TL;DR: 70% because science.

80/20 split has science behind it as well. Check out the Pareto principle. It was originally concocted as a way to explain stuff in business but has since been used and expanded by statisticians and mathematicians because it fits so well in describing many other natural phenomena.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 20:50:32


Post by: bort


Definitely depends on game. 0 luck would seem terrible in a war game as there should be some chaos or fake fog of war. If I were sitting down to a board game chit type war game I’d probably want 80-90% skill. For Warhammer though, part of the hobby is also the minis and higher luck means a bit more flexibility in taking models you like and avoiding some rock-paper-scissors certainty. I’d probably pick 70%-75% for 40K.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 22:23:17


Post by: Voss


 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?


I didn't say that, or anything remotely like that.
But the 80/90% takers seem to want exactly that (barring the RNG)- the game has an almost predefined outcome, and rolling the dice will almost never matter, and the less 'skilled' player won't ever make a game changing decision. That's horrible and dull, and definitely undercuts playing and participation.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 22:38:39


Post by: Peregrine


Voss wrote:
I didn't say that, or anything remotely like that.


That's exactly what you said, you just apparently didn't understand the implications of what you were advocating. A game where you can't get above a 60% win rate by superior skill is essentially pure RNG with very little player agency.


But the 80/90% takers seem to want exactly that (barring the RNG)- the game has an almost predefined outcome, and rolling the dice will almost never matter, and the less 'skilled' player won't ever make a game changing decision. That's horrible and dull, and definitely undercuts playing and participation.


Better that than a game where neither player will make a game changing decision because RNG controls the game. At least with the low-RNG game even the weaker player has a chance to improve their skills and make great decisions, even if doing so regularly may be somewhere in the future.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 23:21:48


Post by: Overread


Voss wrote:
 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?


I didn't say that, or anything remotely like that.
But the 80/90% takers seem to want exactly that (barring the RNG)- the game has an almost predefined outcome, and rolling the dice will almost never matter, and the less 'skilled' player won't ever make a game changing decision. That's horrible and dull, and definitely undercuts playing and participation.


Remember its only a "predefined outcome" when you're putting the very experienced against the very inexperienced - the best VS the worst. That's an extreme match up which in general is most times going to go toward the better player in almost all things in life. Unless the better player takes a handicap to level the playing field then they are probably going to win.

The thing is this kind of system presents a logical series of choices and actions which can result in that win. A new player can observe those, can be taught what to look for and can improve their game in logical steps toward improving their performance. They can learn, improve and develop.

When you present them with a game with 50% or near enough random luck on the win/loss then its far harder to teach them how to improve their game. IF at any point you can randomly lose because of the dice then that can be quite a bitter experience because now the dice hold sway. In general games like that tend to be pretty quick and short. The fact that you win/lose on a dice roll tends to work better when the expeirence is over fast because then volume of games tends to even out the win/loss experience.


Warhammer games can't work like that, they are long experiences and might last several hours. The last thing I think most players want is to feel that all those hours of play mean nothing next to the dice winning/losing the game for them.



Of course early on, in an ideal world, beginners will play other beginners so their win/loss rate shouldn't be too bad. They'll also play more guided games where winning/losing aren't important and what they are learning is the game mechanics. A good teacher might well setup the game for the player to win easily to help reinforce the lesson; but the win/loss wouldn't matter. Many times they might not even be playing a full turn sequence or game.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/07/31 23:45:40


Post by: bort


I should think 60% is the minimum to even consider. Any lower and you aren’t playing a game, you’re waiting 3 hours to see what winner the dice pick. And you’d never have pros as they would still be coin flip against a first timer.

Card games and poker work at like 60% because the skilled player comes out ahead in hundreds of quick plays. 60% might be fine for home Warhammer, but you can’t easily build up a competitive scene and rankings with it. Games take too long, you can’t play enough to reduce luck variance in results.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 00:31:27


Post by: Voss


 Overread wrote:
Voss wrote:
 Peregrine wrote:
Voss wrote:
60, tops. Elsewise being a new player is a curse they'll never surpass, and won't want to continue on with the game. You're basically shooting 'growing the audience' in the head.

At least if I grasp what is thread seems to think 'skill' is.


Are you kidding? You honestly want a game that is almost entirely RNG, where player decisions are barely relevant and you just passively roll dice to see who wins?


I didn't say that, or anything remotely like that.
But the 80/90% takers seem to want exactly that (barring the RNG)- the game has an almost predefined outcome, and rolling the dice will almost never matter, and the less 'skilled' player won't ever make a game changing decision. That's horrible and dull, and definitely undercuts playing and participation.


Remember its only a "predefined outcome" when you're putting the very experienced against the very inexperienced - the best VS the worst. That's an extreme match up which in general is most times going to go toward the better player in almost all things in life. Unless the better player takes a handicap to level the playing field then they are probably going to win..


That isn't the premise I'm seeing in this thread. That's simply that more 'skilled' (NOT experienced or inexperienced) should just win the overwhelming majority of games. That isn't going to teach anyone anything either

The thing is this kind of system presents a logical series of choices and actions which can result in that win. A new player can observe those, can be taught what to look for and can improve their game in logical steps toward improving their performance. They can learn, improve and develop

But they'll still be behind the more 'skilled' player, and therefore keep consistently losing, unless for some reason the 'skilled' player isn't also learning and improving. What's being presented here as a good thing is a closed system were veterans always win, and newbies are just victims until even newer folk join in (which if they're sensible they never will).

IF at any point you can randomly lose because of the dice then that can be quite a bitter experience because now the dice hold sway

I haven't seen anyone postulate that any single moment causes random loss. That's a completely different argument.
When people complain about the 'dice' losing games for them, they're generally talking about long streaks of poor dice rolls- usually engaging in weird superstitions about dice, confirmation bias, or are making excuses for poor decisions.
The last time I saw people complaining about a die roll winning or losing games was back when random game lengths were a thing and someone did or didn't get an extra turn were they would have just 'wrecked face' and turned the whole thing around.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 00:32:33


Post by: JNAProductions


Voss wrote:
That isn't the premise I'm seeing in this thread. That's simply that more 'skilled' (NOT experienced or inexperienced) should just win the overwhelming majority of games.
How does one become skilled?


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 00:37:46


Post by: Elbows


It requires the kind patience and caring of an elderly Asian man...and focus.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 00:46:29


Post by: Voss


 Peregrine wrote:
Voss wrote:
I didn't say that, or anything remotely like that.


That's exactly what you said, you just apparently didn't understand the implications of what you were advocating. A game where you can't get above a 60% win rate by superior skill is essentially pure RNG with very little player agency.

By definition it isn't pure RNG, and the argument doesn't address player agency at all.

Player agency _does not_ mean 'my decisions always win,' that isn't even functional in competitive game. It simply means that players have decisions that lead to a variety of consequences- not necessarily the end game. Being able to choose when to charge is an example of player agency. An ability that denies a player the ability to declare a charge is something that denies agency. It isn't some grand conceit that boils down to winning or losing. Its about options turn by turn within the game, that affect the game as the turns unfold.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 01:00:05


Post by: JNAProductions


Voss wrote:
 Peregrine wrote:
Voss wrote:
I didn't say that, or anything remotely like that.


That's exactly what you said, you just apparently didn't understand the implications of what you were advocating. A game where you can't get above a 60% win rate by superior skill is essentially pure RNG with very little player agency.

By definition it isn't pure RNG, and the argument doesn't address player agency at all.

Player agency _does not_ mean 'my decisions always win,' that isn't even functional in competitive game. It simply means that players have decisions that lead to a variety of consequences- not necessarily the end game. Being able to choose when to charge is an example of player agency. An ability that denies a player the ability to declare a charge is something that denies agency. It isn't some grand conceit that boils down to winning or losing. Its about options turn by turn within the game, that affect the game as the turns unfold.
Perhaps "deny" is the wrong word.

Would "Makes irrelevant" sound better? Because, if I can win with no skill 50% of the time against someone who's much better at the game simply due to RNG, I might technically HAVE agency, but it sure as hell doesn't matter much.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 02:33:02


Post by: Peregrine


Voss wrote:
By definition it isn't pure RNG


I said "essentially pure RNG", not "literally pure RNG". To get a win rate no higher than 60% even with an extreme skill disparity you have to have a game where the outcome is determined almost entirely by RNG. It won't be literally pure RNG, but from a player experience point of view it will be so close that it will be a terrible "game".

and the argument doesn't address player agency at all.


It does if you're able to draw conclusions from information. If the outcome of a game is determined almost entirely by RNG then there can be minimal player agency. Player decisions can not have any meaningful effect on the game, otherwise skill would be a factor and potential win rates would go above 60%. You have to have a game where player decisions are so irrelevant that you can make a perfectly correct decision at every opportunity while your opponent makes the wrong choice and still have barely higher than a coin flip chance of winning.

Being able to choose when to charge is an example of player agency.


Not if choosing to charge has no meaningful effect on the game. If whether you decide to charge or not is irrelevant in the outcome then the choice is merely an illusion, no more an example of player agency than the ability to pick which color pieces to use in a board game. Pick whichever option you want, or flip a coin and let RNG decide. The outcome is the same either way. And if the choice of when to charge has a meaningful effect on the game then you can't have a maximum 60% win rate. If there is a meaningful difference between a right choice and a wrong choice then good players will make right choices more often than weaker players and will accumulate advantages until they win the game at a much higher rate.


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 07:25:36


Post by: IronSlug


Depends of which part of the word WarGame gets the prevalence.

If we emphasis the "Game" part, something around 80-90%.

If its the "War" part, giving the randomness of a real battle, it is not rare to have the most competent general loose a battle because of exterior causes, so maybe 60-70%.

A very tight ruleset can be fair, but real life never is, so I guess it all depends if you are the "game as a sport/chess" type or "game as a simulation/narrative engine".


In an ideal wargame, the higher skill player should win what percent of the time? @ 2019/08/01 09:22:15


Post by: Overread


Voss wrote:


The thing is this kind of system presents a logical series of choices and actions which can result in that win. A new player can observe those, can be taught what to look for and can improve their game in logical steps toward improving their performance. They can learn, improve and develop

But they'll still be behind the more 'skilled' player, and therefore keep consistently losing, unless for some reason the 'skilled' player isn't also learning and improving. What's being presented here as a good thing is a closed system were veterans always win, and newbies are just victims until even newer folk join in (which if they're sensible they never will)..


You're still making the same mistake several are, which is to make the assumption that newbies will only ever play experienced people. You're also making the assumption that skill is like experience in a computer game and that if two people play the same number of games they will retain the same skill divide between them. In reality this is rarely the case. Often as not the experienced person will reach a capping point and the newer person will advance their game faster if they are given the proper support and education along the way.
Furthermore in any general club there will be a spectrum of player types. So you don't have to play the best in the club every single time. Furthermore there are means by which you can handicap the experienced person.

Also consider any other hobby - almost any hobby has a period of time where a new person is utterly rubbish. They will be worse than everyone else who has been doing it for a while and that is perfectly and utterly normal. Yes an increased amount of success in the early stages helps keep them around, but at the same time they've still got to learn. Plus don't forget there's building and painting which can easily take up hour upon hour of their time and be part of the hobby too. It's not just about winning/losing

IF at any point you can randomly lose because of the dice then that can be quite a bitter experience because now the dice hold sway

I haven't seen anyone postulate that any single moment causes random loss. That's a completely different argument.
When people complain about the 'dice' losing games for them, they're generally talking about long streaks of poor dice rolls- usually engaging in weird superstitions about dice, confirmation bias, or are making excuses for poor decisions.
The last time I saw people complaining about a die roll winning or losing games was back when random game lengths were a thing and someone did or didn't get an extra turn were they would have just 'wrecked face' and turned the whole thing around.



The point is if a player with more skill can only win a game 60% of the time against a player with significantly less to no skill in the game then it means that dice are running the game. I used one dice roll as an example; but be it one or a dozen or a hundred the point is that its the dice roll which is deciding if the game is won or lost more so than the players choice in models to bring and how they use them on the table. This makes for a frustrating game and a hard one to learn.
Imagine if a less skilled player charges their cavalry into their opponents spearmen:

Under a game with a 60% win/loss there's a nearly even chance that either side in the battle could win. The cavalry or the spearmen could win. This means it doesn't really matter if they charged with cavalry or infantry. It also means that even if the spearmen are technically the counter unit, the player might not learn it very well because the spearmen keep losing every so often. So they never get to advance their understanding of the game because the dice are swinging results far more so than the players choice.

Under a game with a closer to 80% or greater win/loss for the more experienced; the cavalry would be expected to lose against the spearmen because its a very poor match up and the spearmen are a direct counter unit to cavalry. This would be a pattern that would repeat over and over each time and would eventually teach the newer person that you don't charge spearmen with cavalry. They can learn that and next time use infantry or archers instead and get a reliable better result.