The dice is fair: By Cow - rank #1 season 1 and 2 allies

[HOCHI MINH]

If you have ever sat down and flipped a coin for 100 or so times or played poker and other card games for long enough you will notice something.

If you flip a coin 1000 times you will likely get around 500 heads and 500 tails. During this time there will be long streaks of 10+ straight heads and tails. What does this mean? In a dice game sometimes your aa guns will hit every single roll and sometimes you won't shoot a single bomber down all game. Those 2 starting bombers can do a 100ipc worth of damage and never get hit or the 5 bombers you bought this game will all get shot down before breaking 20 ipc damage.

Bad luck is most painful round 1. Beyond this it is far easier to recover so long as your bad luck is not on say Berlin from a last ditch effort from the allies to take your capital at a 2% chance. This can happen. It has happened. I have been both the recipient and a giver of brutal dicing.

The server has fairly close to natural dice. Of course being a server using programming comprising of 0s and 1s.... it can never fully replicate your hand throwing dice across a room (barring quantum computing machines). That being said I do feel the dice is close to natural.

In a live game I have shot down 9 out 9 fighters defending Moscow on the 50th anniversary board. I a live game I have lost a carrier, battleship, and 2 fighters to a single fighter that lived. I have seen 2 german subs sink bship, a loaded carrier, 2 destroyers and a cruiser on uk (this was a fast loss or the allies).

I have played over 1000 games over the past 15+ years. I have seen some things. Sometimes the dicing is your opponent hitting 50% of the time throughout the game, your bad luck spread out and you can't hit when you need it the most. It happens.

I had a live party game someone felt the need to track all of our dice to see who is the luckiest. I was rolling 22% 1s 23% 2s and 20% 3s, while playing Russia and I smashed Germany. Hitting close to 50% of the time for 100+ rolls. I been on the receiving end of this and the giving side of this on this game in ranked.

The dice can be extreme in your favor or against your favor. It happens 1 in 10 or so games. I have had 4 brutal games in a row online. I actually clawed my way back into a few of those.

What separates the top 10 players from everyone else is not just having good strategy and adaptation to what your opponent is doing, but also adapting to the dice outcomes and understanding where you are at. Sometimes all is lost and all you can do is a last ditch effort attack at Russia or Berlin or a VC win by trying to hold them all. Because if you still have a 2% chance to win and that is the best it will ever get, then you take it or fade away with a 0% chance to win.

You will get diced.
You will dice someone.

It is inevitable. Embrace it.

[1baddude]

From what you've said so far I take it you are lucky with the dice. Because you only presented experiences where you won. While I can only tell experiences where I lost to brutal dices, some of them also posted with pictures in the forum.

Anyway, all we unlucky ppl ask is a low luck option in order to enjoy our games with more strategy and less 'unluck'.

There will still be room for mad streaks to express themselves but at least we can say bye bye to fights were 20+ units score 0 hits in the first round while the enemy smashes with half of his units.

[krystalsmoke]

Thanks for the nice post Cow. For what its worth, I agree with you. And let me add that I love this game also because its in part a dice game. I think it sits halfway between e.g. chess (not much luck there) and risk (mostly luck there :)).

[Julius Borisov]

Thanks, Cow, and congratulations to you and Snuggles for getting #1 spots.

Originally posted by GOD EMPEROR TRUMP:

What separates the top 10 players from everyone else is not just having good strategy and adaptation to what your opponent is doing, but also adapting to the dice outcomes and understanding where you are at.

You nailed this, in my opinion.

It would be great to hear from Snuggles and other Top 10 players as well.

[Zukov]

Good post

I totally agree!, have 20+ years experience with A&A, with live, pbem, online games. Here only 30+ games

Originally posted by author:

Bad luck is most painful round 1. Beyond this it is far easier to recover so long as your bad luck is not on say Berlin from a last ditch effort from the allies to take your capital at a 2% chance. This can happen. It has happened. I have been both the recipient and a giver of brutal dicing.

And this is so true! It is hard to bounce back when you Russia kills WR and Ukraine without losses and then you lose big time against UK fleet, but it is possible.

[doggy_dady]

Well said!!

[Kilroy]

If there were a like or thumbs up flag on for this forum, I bet this would be a top ten post instantly. Thanks for taking the time to articulate something I feel every time I read someone complain about the dice.

[sirhahnsirhahn]

Thanks, Cow. I can't agree more with everything you said. Hope we get a chance to play sometime.

[WebbsNYC]

Originally posted by GOD EMPEROR TRUMP:

If you flip a coin 1000 times you will likely get around 500 heads and 500 tails. During this time there will be long streaks of 10+ straight heads and tails.

The odds of flipping heads or tails 10 times in a row is slightly over 1 in 1,000. Meaning you would need over 1,000 trials of 10 flips to get that result. If you're telling me this happens regularly and there are frequently strings of 10+ straight heads and tails, I'd say you are experiencing something well outside the norm...which is exactly the point that I, and many others, complain about in regard to the dice. Low probability outcomes happening once in a while makes sense. Low probability outcomes happening all of the time do not make sense and indicate the dice are not "natural."

Enjoy the game...in fact, I hope and am glad that everyone on this thread enjoys the game. My criticism of the dice stands. And if you want to prove your point about the coin flips, I'll go $100 with you that you can't flip a string of 10+ heads in a row in 1,000 coin flips.

[sirhahnsirhahn]

No where in his post did he say that flipping heads or tails 10 times in a row happens regularly, or frequently. He said there will be streaks.

Please stop complaining about the dice.

[aardvarkpepper]

Thread title:

Originally posted by GOD EMPEROR TRUMP:

The dice is fair: By Cow - rank #1 season 1.

Dice being fair or unfair has nothing to do with rank. It's a total non-argument.

The rest of the post has no masses of data drawn from in-game play, only conceptual examples. As such, it's simply anecdotal; in terms of supporting an logical argument has less validity (as it has no data) than the data provided by those that argued the PRNG resulted in 24 million plus and 1 in 37 million plus cases.

I'll also point out the inherent bias in a top-ranked player reporting they find the system fair. If you've spent dozens or hundreds of hours working your way to a top rank, do you really want to call the validity of your ranking into question? Of course not. Of course you want to say the system *is* valid.

If top ranked players did a load of games at the start of a season to load up on easy points, if they took 24 hours to move between turns to try to time out opponents that missed a login, if they started a load of games right before major holidays anticipating more players would drop - those are all legitimate non-hacking ways to gain in rank as well.

Which is not to say that Cow or any player in particular did these things. I'm simply pointing out there is reason for ranked players to sell the legitimacy of the system.

Really, rank and PRNG fairness are two completely separate things.

If the #1 ranked player said "PRNG was garbage" with the same level of evidence, I'd argue against that too.

Time played, rank, anecdotal evidence - these are not things to rely on when evaluating PRNG fairness. What's needed is mathematics and data.

Mind, I don't think any poster is claiming otherwise. Even the post I'm replying to could be interpreted as reading the dice are "fair" in that even if the PRNG does have issues both players have to deal with it. Though I'll point out as I have previously - I disagree with any claim that "PRNG applies to both sides" is a legitimate line of argument to discount any PRNG issues. As I've pointed out, the setup is asymmetrical, the lines of play are asymmetrical. Also Axis and Allies is a dice game; distorted probabilities cause players to make uninformed and improper decisions based on their anticipation of proper RNG modeling meeting improper PRNG implementation (if the PRNG *is* an issue).

==

For other points made in the OP - comment -

Originally posted by GOD EMPEROR TRUMP:

What separates the top 10 players from everyone else is not just having good strategy and adaptation to what your opponent is doing, but also adapting to the dice outcomes and understanding where you are at.

I agree with the general idea. Readers that want high rank should pay attention to that statement.

Only thing I'd say different is good strategy and tactics *requires* being prepared for a range of dice results. If a strategy depends on questionable tactics, it can't be a good strategy.

To elaborate:

A) Just because you have 85% odds on an attack doesn't mean it's a "good" attack; if your opponent can't stop you from making it a 99% attack next turn and there's no real reason to press prematurely, then best wait.

B) Even with favorable odds, there's no reason to take unnecessary risks. Suppose you're playing Germany, you control most of Africa and Europe and are about to capture Russia this turn, and Japan is doing very well. You could buy fighters on Berlin and leave the Allies with only a 10% to capture; say if the Allies attempt to capture they'll likely lose a load of valuable air, leaving Germany an easier time of breaking the Allied Atlantic fleets on its next turn. But at that point the Allies are pretty much dead anyways. So really Germany shouldn't even leave that 10% chance, just buy a chunk of infantry and leave a 0.1% chance to capture, then grind the game out to the win.

C) As with B, but if the Allies are pressing the Axis hard, maybe Germany leaves the Allies with an attack on Berlin, gambling that if the Allies try and lose it'll cost valuable Allied air, and if the Allies don't try then Germany can consolidate its position elsewhere. That is, Germany could win or lose but it doesn't have a "safe' path that leads to success, so it needs to figure out what's best for that situation.

When players read general advice, they often think they're *already* doing things like "adapting to opponent moves" or "adapting to dice outcomes". So I wrote a few specific examples. Though eh, maybe I should write a bunch more guides for A&A, pop in a bunch of mathematics and build some tools and things. Eh.

[aardvarkpepper]

Originally posted by WebbsNYC:

Originally posted by GOD EMPEROR TRUMP:

If you flip a coin 1000 times you will likely get around 500 heads and 500 tails. During this time there will be long streaks of 10+ straight heads and tails.

The odds of flipping heads or tails 10 times in a row is slightly over 1 in 1,000. Meaning you would need over 1,000 trials of 10 flips to get that result. If you're telling me this happens regularly and there are frequently strings of 10+ straight heads and tails, I'd say you are experiencing something well outside the norm...which is exactly the point that I, and many others, complain about in regard to the dice. Low probability outcomes happening once in a while makes sense. Low probability outcomes happening all of the time do not make sense and indicate the dice are not "natural."

Enjoy the game...in fact, I hope and am glad that everyone on this thread enjoys the game. My criticism of the dice stands. And if you want to prove your point about the coin flips, I'll go $100 with you that you can't flip a string of 10+ heads in a row in 1,000 coin flips.

1) @GOD EMPEROR TRUMP: The reports of in-game data I linked to in the sticky thread are on the order of 1 in 24 million and 1 in 37 million. If you combine the odds of 10 heads and 10 tails it comes to the order of 1 in 512. It's just not the same thing at all. Please don't trivialize things by making such incredibly inaccurate comparisons.

2) GOD EMPEROR TRUMP is saying if you have a bunch of long streaks of 10+ straight heads and tails in 1000 coin flips, that's normal. WebbsNYC is saying that's not normal. WebbsNYC is correct - MATHEMATICALLY correct.

My impression is GOD EMPEROR TRUMP is saying you get abnormal results and that's to be expected. Then for (his?) example, he is saying you should get not just one streak of 10 heads or tails, but MULTIPLE ("streaks") of 10+ in a series of 1000 coin flips. That just isn't *probable*. If it happens, it's abnormal, and if it keeps happening it's a sign that there's something wrong.

A lot of readers that don't care to understand the mathematics want to dismiss the whole argument. Vaguely, some may think that there's some sort of argument that improbable events occur, and so long as it doesn't involve them understanding the mathematics, they're happy to dismiss the whole thing out of hand and proclaim that whoever said there wasn't a problem.

If there's no problem, readers are personally absolved of any responsibility of informing themselves of whether there is actually an issue or not, and further absolved of any responsibility for seeing anything gets done.

If there's no problem, there's no need to understand anything. So simple.

But really? What some are saying is the issue is there's *too many* improbable results. Casinos, insurance, oddsmakers, all run on the *numbers*. Businesses and fortunes are made on running and understanding the numbers. Numbers are not a trivial thing.

The nature of the internet being what it is, points have to be repeated again and again. I understand that. But really, the repeated denial and misrepresentations of the PRNG issues is not good. It's not a healthy or realistic environment.

If you're somewhere where the water is a weird orange color and there's a load of frogs and cows with two heads walking around, and some well-meaning people are walking around saying "frogs have two heads all the time! cows too!" and you go over to a barn and open the door and 45 out of 47 cows have two heads and you say that's not normal and they say "nothing that can't be explained by chance!" look okay. Yeah maybe it CAN be explained by chance. But that's not normal. It's reason to think there's something weird.

If you're in a casino and the players are winning way more than expected, you think the bosses are going to settle for "it's just chance!"? All those cameras with the zoom lenses and the security aren't just for show. There's going to be investigations. It's a serious matter.

If posters don't understand the issue, that's their prerogative. If they want to express an anecdotal opinion, that fair enough. But once posters actually trying to make arguments without data or mathematics or reasoning supporting, and *saying that things should or should not happen on basis of those supposed arguments*, well.

[WebbsNYC]

Originally posted by sirhahnsirhahn:

No where in his post did he say that flipping heads or tails 10 times in a row happens regularly, or frequently. He said there will be streaks.

Please stop complaining about the dice.

He said that during a stretch of 1,000 flips, there would be streaks (plural, mind you) of 10+ heads or 10+ tails. I reject that premise. There will almost certainly NOT be multiple streaks of 10+ of anything. One? Maybe. Two or three? No. Highly unlikely. Not without a weighted or trick coin.

In my opinion, this is evidence that his view on what is "probable" or "likely" is somewhat flawed and this misapplication of probabilities to coin flips very possibly carries over to his misapplication of probabilities to dice rolls and things he assumes are just "normal" are, in fact, not at all normal.

[Comrade Bernie]

Thank you cow, and no thank you Aardvarkpepper, you always seem to be a huge downer on the conversations. Sure, it sucks to be on the recieving end ALL THE TIME, and trust me, I know what that feels like. I have played over 415 hours of this game in the past two months, and the highest rank I ever got was 119th in silver as the Axis. I have a 56% win rate, and most of my wins are from other people forfieting because they got bored or something. I don't even remember the last time I went on a bombing raid, and didn't lose at least one bomber. But the thing is, I don't care. It's just a game, and a nice way to pass the time, so when I lose, sure it's a bummer, but who cares? It's not like being number one in ranked play is going to 'look good on your resume." I can understand if you just hate being a loser, and want a refund for the game, but if you are going to argue so much, you might as well sue them and get a lawyer. But by george, this is not a courtroom!

[Nikako]

Maybe in the aggregate? What I've tended to notice is that I have games where I'm -incredibly- lucky in every fight and other games where every fight is absolutely stacked against me. It doesn't seem like the distribution is even.

[1baddude]

Originally posted by individualistds:

I didn't read OP's post as saying the dice are fair. Quite the opposite.

I read the post as saying "the dice are unfair, deal with it. It's how you deal with it is what separates the great players from the good.

You actually have a good point here! :)