Statistical Model for Predicting the Winner


  • @akreider2 yeah, that’s a good point.

    Separately I think it might be cool to get a place for people to report their win/losses with certain common strategies vs other common strategies. Might be cool to have a model for what to use versus what, optimize the game-play for those who play competitively online, although I suspect they’d already have a good idea of what works versus what.

  • 2007 AAR League

    It’d be fun to train an AI using pro games. I wrote an AI once, but it was for Connect Four (a very simple game) and many years ago.


  • @akreider2 yeah… that’s pretty out of my league. Id simply like to analyse some game data from many different players of different levels to develop a little cheat sheet that would tell you how to play against each strategy and skill level.


  • How does the model handle force projection?

    Example: USA has 50 tanks in Brazil and no transports
    Germany has 50 tanks in Germany and no transports

    Which nation will kill more units and take more IPC and take more VC?


  • Hhhmmmmm…

  • '21 '20

    the model doesn’t, i don’t think. What i was proposing was to make a variant model for G40 in which the user would manually account for things not represented by troops or IPCs.


  • @Imperious-Leader The proposed model wouldn’t be that effective at handling stupid moves, and mistakes.

    You could partially control for mistakes with a variable that is based on the player’s league ranking (if they are in the league).

  • 2007 AAR League

    As a baseline test, let me know if you can find a game where the Axis player has equal or greater total unit value to the Allies and isn’t winning.

    Alternatively, can you find one where the Allies have a total unit value of 400 more than the Axis and they aren’t winning (this latter might be possible if the total unit value is very high)?

  • 2007 AAR League

    One of the big issues for pulling this off - does anyone know if you can extract data (notably total unit value and other statistics) from the Triple A save files?

    Is it in XML, CSV, JSON or some other easy to use format?


  • Barnee know any of that ?

  • 2023 '22 '21 '20 '19 '18

    @SS-GEN

    yea you would want to ask @redrum I know tuv in tripela isn’t represented 100% accurate although it’s pretty close. I believe they’ve made some changes recently that make it better.

    In global it seems the allies are overvalued a bit from what i remember

    Edit
    Actually I think it counts it right but it includes infrastructure. So for combat unit vs combat unit it’d b a bit different because of all the allied infrastructure.

  • 2007 AAR League

    I remember wondering whether I should include the infrastructure and forget what decision I made. The infrastructure is often useful and the model would be controlling for the discrepancy at the startup. So I think it’d mostly count new infrastructure which is typically important.


  • @akreider2
    Hi ak

    yea I was just thinking you need to allow for it somehow but allies have kind of an excess I guess you would say that doesn’t really get used. At start anyway

    idk redrum probably reply in a day or so. He’s usually pretty good at that

  • 2023 '22 '21 '20 '19 '18

    yea counting new might be a good way to go. Probably better than total. Or somehow how PU totals vs how much could be used for factories. Planes and ships for AB and NB.

    Idk that seems to be a lot of work for maybe a small reward. Even if u get close with just new it should be of some value I would think


  • @akreider2 said in Statistical Model for Predicting the Winner:

    As a baseline test, let me know if you can find a game where the Axis player has equal or greater total unit value to the Allies and isn’t winning.
    Alternatively, can you find one where the Allies have a total unit value of 400 more than the Axis and they aren’t winning (this latter might be possible if the total unit value is very high)?

    In 1942.2 i usually win from behind because im keen on build centers being in the direct area of fighting, which totally dictates what Clausewitz would characterize as mass, economy of force, and objective among his eight principles of war. A simplistic “counting” and judgement based on IPC values is too static. I was recently down by about 115 points and won as axis. ( Points being total aggregate attack factors and im sure IPC values aren’t far behind).

    Another point is this is a game of dice. In fact roll buckets of dice! The point is the dice server they use is slanted ( even if they deny it it is messed up)
    So… im fighting a slanted dice server and having to deal with a piss poor partner who loses his position right off the bat . He was Japan and I’m Germany.
    I kept USA/UK away till round 8 with subs, but I had the need for more ground pounders so i can finish off Russia, which happened. The collapse of them left a void where Japan could get some money and Germany was at 51IPC, but The Allies had ALOT more costly pieces, but they lost anyway.

    As far as the 400 number ( which is a huge advantage) However if the Axis have 2000 IPC worth of land units, and the Allies have 2400 in Naval units, your idea is shot down. Like i said your just comparing all units total value in IPC and not taking account of types and force projection, the study is faulty.


  • This post is deleted!

  • @Imperious-Leader Statistical models predict things to be more likely than not. They can still be wrong. They aren’t perfect. If some idiot builds only naval units, or if someone decides to throw the game - it won’t work. (It also won’t work if the sample is too biased, or too small or half a dozen other reasons).

    A classic example is that 538 model had Trump with a 20% of winning the US presidential election. Now I really don’t want to get into any political discussion, but it was completely possible that even though the 538 model favored Clinton, that Trump could still easily win. Any Axis and Allies player will be able to tell you how often unlikely events happen (so far my most unlikely dice roll was 1 in 100,000).

    My background: I’ve always been fascinated by statistics. At age 5 I won a contest estimating the number of smarties (M&Ms but it was in the UK) in a jar. I was reading the World Almanac at age 11. I wanted to be an economist at age 12. I ended up with a MA in Sociology. I’m a part-time professional gambler since 2016 on PredictIt (and have consistently made money for the past 5 years). Though I’m only familiar with linear and logistic regression. There might be better techniques to use (and if so I’d be happy to hear about them).

    My hypothesis is that among good players, positioning leads to IPC gains and can be modeled. I haven’t tested this hypothesis in Global 1940.


  • @akreider2 said in Statistical Model for Predicting the Winner:

    A classic example is that 538 model had Trump with a 20% of winning the US presidential election. Now I really don’t want to get into any political discussion, but it was completely possible that even though the 538 model favored Clinton, that Trump could still easily win. Any Axis and Allies player will be able to tell you how often unlikely events happen (so far my most unlikely dice roll was 1 in 100,000).

    Well Truth be known is Trump had 1.2 million votes more than clinton when he won the election. In California and New York where the voting booths closed late, another ~6 million votes came in democrat for states that everybody knew would go for clinton. So your example of “bad dice” really wasn’t that at all. It was poor polling and media promoting clinton and having their minds convinced of an easy clinton victory.

    My background: I’ve always been fascinated by statistics. At age 5 I won a contest estimating the number of smarties (M&Ms but it was in the UK) in a jar. I was reading the World Almanac at age 11. I wanted to be an economist at age 12. I ended up with a MA in Sociology. I’m a part-time professional gambler since 2016 on PredictIt (and have consistently made money for the past 5 years). Though I’m only familiar with linear and logistic regression. There might be better techniques to use (and if so I’d be happy to hear about them).
    My hypothesis is that among good players, positioning leads to IPC gains and can be modeled. I haven’t tested this hypothesis in Global 1940.

    It is necessary that UK/USA need to buy naval and air to shuck to Europe. The ratios of land/sea/air units are different and if Germany takes Moscow, but to take a VC you need land units and once the Axis take karelia, they need 2 more and these are attainable by land (Moscow and India and or possibly Hawaii)

    Based on force projection the Axis are better poised to do this because they start so many land units. In a way totals IPC totals might predict some direction of victory, but like in Poker you can be on the nub with a “chip and a chair” and still beat your opponent with a number of all in’s your chip stack has little to do with win %, except in a casual way.


  • You could try to measure supply lines to targets (eg. victory cities), but it’d be a challenge. You want to maximize your unit value within a short distance of a target and your resupply rate.


  • @akreider2 Well, at that point youd have to build it into something online, it would simply be too difficult or pointless to manually plug all of the supply lines and resupply rates

Suggested Topics

  • 9
  • 6
  • 5
  • 6
  • 16
  • 14
  • 127
  • 243
Axis & Allies Boardgaming Custom Painted Miniatures

27

Online

17.0k

Users

39.3k

Topics

1.7m

Posts