That’s exacty why I tied it to expected hits, not dice.
I had this battle with Veqryn in 2011; when he first introduced the tripleA dice log. He just said dice is dice the order doesn’t matter. It’s a wrong assertation.
Rolling a 6 on a bomber is not the same thing as rolling a 6 on an infantry. You’re correct, everyone misses the temporal component of the dice when they tried to do this.
That’s why this is expected hits vs actual.
A good example of this was my recent/epic G1 against Mallery29. His dice weren’t bad when presented as just dice. But when presented in a hit handicap system, Germany was +9 hits more than expected on G1, and the allies were -6. That’s a 15 unit differential on the first turn of the game. Brutal.
Further to the same point - look at this game with Farmboy
https://www.axisandallies.org/forums/index.php?topic=41042.195
If you just looked at dice averages, nothing stands out.
When you run it through a expected vs actual hits handicap this is the result
Axis +22
Allies -30
A 52 unit differential; based entirely on luck. When you consider that the game is 11 turns deep, you would expect that the trends would head to 0,0. But now you can actually see the pain! Tangibly touch it!
That’s 22 units the allies could have on the board, and 30 units the axis would remove from the board, if hit results had been more normal. It’s made for a hell of a game.
Adam - lets try it out on one of your games, any ones you can think of that you did exceptionally well? or poorly? or were really just equally average?
I’m just saying that you won’t be accounting for a majority of the luck in a match. For example, Russian sub vs DD blocker, the result of that battle is worth way more than 1 hit or miss statistic. Plus the difference in unit cost, for example navy battle vs land battle.
This game had exceptionally poor dice for me, but the game files seem to have been deleted:
https://www.axisandallies.org/forums/index.php?topic=37061.270
This one was up and down, your statistic will say that Allies had the luck advantage, while a few key battles went in favor of the Axis:
https://www.axisandallies.org/forums/index.php?topic=40627.120
I’m just saying that you won’t be accounting for a majority of the luck in a match. For example, Russian sub vs DD blocker, the result of that battle is worth way more than 1 hit or miss statistic. Plus the difference in unit cost, for example navy battle vs land battle.
This game had exceptionally poor dice for me, but the game files seem to have been deleted:
https://www.axisandallies.org/forums/index.php?topic=37061.270This one was up and down, your statistic will say that Allies had the luck advantage, while a few key battles went in favor of the Axis:
https://www.axisandallies.org/forums/index.php?topic=40627.120
Well of course, getting lucky (dice when you absolutely need em), and getting diced can be two totally different things. I’m just trying to get rid of the one argument.
And yes I understand that Naval battles and land battles have totally different economic consequences. Ultimately I am trying to establish, that everytime you clicked to roll a dice - this is how you really did in comparison to average. I really don’t like guys that say “well the only reason you won was dice”. Whilst that may be - I’d prefer to have some tangible result that shows how on or off the mark i was compared to my opponent.
Evolution is also a consideration here - I’m also trying to push the Dev team down further game reporting rabbit holes. For example, how awesome would it be to have a full game casualty report? Germany lost 340 inf, 120 art, 300 mech etc; allied losses as follows etc.
Or a better reporting system for listing peoples purchases, so on one screen or one piece of paper, I can see all game purchases by a nation in order - better studying myself or my opponent.
I also think others may build on some of these concepts - like a total report of battle TUV gains/losses, or a report on people losing battles that were supposed to be expected wins. Like you go all in on Egypt and LOSE when it was 80% for you. How many battles 75% or better did you not win? that kind of thing.
I’m just saying that you won’t be accounting for a majority of the luck in a match. For example, Russian sub vs DD blocker, the result of that battle is worth way more than 1 hit or miss statistic. Plus the difference in unit cost, for example navy battle vs land battle.
This game had exceptionally poor dice for me, but the game files seem to have been deleted:
https://www.axisandallies.org/forums/index.php?topic=37061.270This one was up and down, your statistic will say that Allies had the luck advantage, while a few key battles went in favor of the Axis:
https://www.axisandallies.org/forums/index.php?topic=40627.120Well of course, getting lucky (dice when you absolutely need em), and getting diced can be two totally different things. I’m just trying to get rid of the one argument.
And yes I understand that Naval battles and land battles have totally different economic consequences. Ultimately I am trying to establish, that everytime you clicked to roll a dice - this is how you really did in comparison to average. I really don’t like guys that say “well the only reason you won was dice”. Whilst that may be - I’d prefer to have some tangible result that shows how on or off the mark i was compared to my opponent.
Evolution is also a consideration here - I’m also trying to push the Dev team down further game reporting rabbit holes. For example, how awesome would it be to have a full game casualty report? Germany lost 340 inf, 120 art, 300 mech etc; allied losses as follows etc.
Or a better reporting system for listing peoples purchases, so on one screen or one piece of paper, I can see all game purchases by a nation in order - better studying myself or my opponent.
I also think others may build on some of these concepts - like a total report of battle TUV gains/losses, or a report on people losing battles that were supposed to be expected wins. Like you go all in on Egypt and LOSE when it was 80% for you. How many battles 75% or better did you not win? that kind of thing.
Great ideas, I’d be very interested to see that.
About your issue with “the only reason you won was because of dice”, just ask for proof! My worry is now that they have your proof, they’ll start claiming that with actual validity!
Expected TUV vs actual TUV swing of a battle is already better than expected hits vs actual hits, because it takes into account the economic part, and also you won’t have 100 tanks vs 1 inf skewing results. Nerquen and Axis-Dominion did this for one game, and even for that I had my reservations.
I like that idea ! I think I can get the script written to show predicted vs actual TUV swing; very cool concept.
Exactly the kind of conversation I was hoping for.
As for people having a valid argument that they got diced ; i’m OK with it ; at least it’s not a debate; and in some cases this may allow players to self mitigate - IE your game above you were given a draw. I’ve got maximum respect for that; that’s something I Will do in the future for less experienced players I face. I would hope others might do that for me :).
In general - my rule is that I always want to play someone at their best; And Have a good game ,if they make mistakes or miss something they should be allowed to correct it within reason. If the dice are horrendous, that is something else to consider .
Also - say I play another master player, barely squeak out a win , but had a handicap of 40 or more , maybe I’m not as good as I think and I need to practice :)
How many dice were thrown in your game Gargantua?
To know the standard deviation of the number of hits, take the square root of number of dice through x 1/2 x 1/2.
E.g., 5,000 throws = standard deviation of 35.
In that case each swing are still within one standard deviation for each player, but both go against you. Maybe that joint outcome would be a 10% worst case for you and 90% best case for him.
These scores even out on a relative basis (per dice thrown), but not on an absolute basis, the deviations can continue to compound one against you or you can recuperate (it’s equally likely).
Would this be posted every time you post a move, or will it be a button in TripleA itself to see all these statistics?
Predicted vs actual TUV swing won’t be as simple as hits vs expected hits, since you need to find the predicted TUV swing. How will that be done?
Yes always play against someone at their best, that’s my philosophy as well.
Good work!
It will be a game menu option and/or it will be posted and the bottom of each post in the report.
As for TUV swing, we’ll have to see about that… each battle would have to be calculated. Prior to rolling. Perhaps when we click DONE on combat move, it would take 3o seconds and calculate all battles? Expected TUV swing vs actual TUV swing would then be what’s measured.
I’ll ask for this; and put some money down for it, but let’s get the easy fluff like casualty records and purchase records; or even a better TUV break down summary (showing unit type by unit type) done first.
If we get some key reports available - the statistics nerds can go nuts and we can have a great new way to look at our favourite game.
I’d really like to be able to remove facilities from TUV. And also be able to see, how much of the TUV is land, air, ships or even show a complete count of what I have.
…I’d really like to be able to remove facilities from TUV…
Remember Veqryn telling Gamerman there was a way to do that. Was several years ago. I don’t remember what it was though
Where’s the thread for the 2016 championship game?
They are starting it now - there is a thread if you look again
FWIW I agree with Adam’s point of view here. When I analyze a game as to who was luckier, I have to do it manually, battle by battle, looking at the game situation at all times. Adam’s example was a great one - the example of a sub trying to clear a destroyer that would lead to a coup, and whether it succeeds or fails. You can’t quantify a lot of the actual good luck or bad luck by an automated process - it’s just not possible. But yes, if you want an automated process the best is probably to calculate actual TUV swings vs. expected. But when I say “best”, I mean least misleading :lol:
Another thing is that your luck in round 1 matters a lot more than your luck in round 10. Early luck or bad luck is normally weightier. Really the best way to evaluate luck in a match is to pay (lol) a master to review it battle by battle and give you a report.
Adam’s example was a great one - the example of a sub trying to clear a destroyer that would lead to a coup, and whether it succeeds or fails.
Probably an even better example is an Italian attempt to clear a territory for Germany that would allow Germany to sack Moscow with a very high IPC swing. Any automated attempt to quantify the luck of this battle would fail miserably. It could be a very small battle with a very small TUV swing, but actual massive game consequences.
Not to downplay your efforts - there would still be a lot of value to a TUV swing calc - would just need caveats to go with it is all we’re saying :-)
is there some way to change the topic on a game?
auto-field mislabeled one of mine.
Let me know.
Yes, not directly, but I can get the results that you want. And that reminds me…. the league threads say 2017 so I’ll do something about that.
This thread is closed (because it says 2017). Please use the new “League General Discussion thread” for league discussion going forward.
