Luck Plays No Part in Diplomacy

[RoganJosh]

Swordsman, it not clear what you mean by in reality. Is 1v1 enough?

jmo, now it seems like you don't know what a "coin flip" refers to. Let me call it a guess instead.

You are missing an important point. When you end up in a situation in which there is just a single guess left, and nothing else. And, where both players are competent enough to realize that it is a guess, and where both player are competent enough to realize that their opponent will try to predict their choice, then any reads become practically useless.

[RoganJosh]

the public chat in this game is quite revealing, but I'm not sure you guys accept 1v1 games
http://webdiplomacy.net/board.php?gameID=228808

[RoganJosh]

Here is one I won after randomizing the Ber/Mun choice in the last move
This is also an example of a position with a pure 50/50 guess

http://webdiplomacy.net/board.php?gameID=224287

[jmo1121109]

Oh, and even if you reject the possibility of two players ever being equal in Diplomacy Skill, then you could still attribute to luck every instance of the player of lower Diplomacy Skill guessing correctly against the player of higher Diplomacy Skill. Which I am CERTAIN you would all agree happens plenty often.

Except I've already addressed why this isn't accurate. Because even people of similar skill do not play with that level of skill consistently across games or even turns in the game. And that the various other factors in the game impact which choices are acceptable to all involved players who have to coordinate. You keep incorrectly trying to dismiss the Diplomatic factors in the game and bring a decision down to 1 or 2 factors when the game is literally designed to prevent this. To the point of needing suspension of disbelief to even be able to consider your scenario as being a true 50/50.

[jmo1121109]

Swordsman, it not clear what you mean by in reality. Is 1v1 enough?

jmo, now it seems like you don't know what a "coin flip" refers to. Let me call it a guess instead.

You are missing an important point. When you end up in a situation in which there is just a single guess left, and nothing else. And, where both players are competent enough to realize that it is a guess, and where both player are competent enough to realize that their opponent will try to predict their choice, then any reads become practically useless.

That scenario does not exist is my counter, where everything comes down to a guess where both parties know their predictions will cancel out. Press games are games of skill and 2 people engaged in a skill based fight are not going to throw that out of the window for an uneducated guess. They'll make the choice with the best information available. In an anon game that includes the game history, their press tendencies, etc. The big problem here is you keep arguing that the press doesn't matter, and you're flat out wrong. It always matters, in every single situation. The proof of burden absolutely falls on you here to find even 1 press scenario where Diplomacy played no part in a "guess" as you're claiming.

I'm perfectly willing to accept that 1 v 1's have game theory guess matches. 1 v 1 is the site's version of a chess match. There's far less factors, and it's only between 2 people.

[RoganJosh]

A full press game can end with only two powers of the board. In some cases that is an agreed two way draw. In some cases it's not. In the latter case, the game works just like a 1v1 game.

I'm sorry, I am not gonna spend hours looking through lists of old games. Especially since I don't think that your objection is particularly serious.

[Your Humble Narrator]

Jmo, do you think that every correct guess is entirely attributable to the Diplomacy Skill of the correctly guessing player?

[Your Humble Narrator]

To expand on what I’m getting at now, I’m not sure that it’s been necessary for me to confine my perceived instances of luck to 50/50 guesses. If, thanks to my diplomatic abilities, I increase a 50/50 chance of success to a 75/25 chance of success, and this game turns out to be that one in four in which I guess incorrectly, how else are we to describe that one in four possibility coming to pass other than by calling it luck? We are assuming either that my opponent has not taken adequate predictive or influential steps to alter the odds of my success, or that the steps that my opponent take were in fact what brought my odds of success to 75/25 (perhaps from 85/15 if both opponent and I had guessed randomly).

[RoganJosh]

Well, it turns out there was a full press example that ended a week or so ago in the ODC tournament. Squigs44 was even playing in it!

http://webdiplomacy.net/board.php?gameID=236425

Position at A14. Possible outcomes is a draw and a victory for Germany. This is solvable by hand. That is, there is nothing left but guesses here, and you can compute the payoff matrices.

PS. Aurelin accepted a draw in S15 from a winning position because he was tired. Maybe we should post these as chess problems? "Germany to play and win in 1."

[David E. Cohen]

This has been an interesting discussion. Not the first time it has taken place, nor is it likely to be the last, but interesting nevertheless. No matter which side you agree with, it can be helpful to look at the fame on this level.

[RoganJosh]

Since it's summer and, whatever... Here's your example with a 50/50 guess for Berlin/Munich in the last phase. Full press. Game was balancing between T solo and a 3 way draw.

http://webdiplomacy.net/board.php?gameI ... #gamePanel

Edit: Note that Turkey's last comment in the public chat is posted 11 days after the game ended.

[Squigs44]

Well, it turns out there was a full press example that ended a week or so ago in the ODC tournament. Squigs44 was even playing in it!

http://webdiplomacy.net/board.php?gameID=236425

Position at A14. Possible outcomes is a draw and a victory for Germany. This is solvable by hand. That is, there is nothing left but guesses here, and you can compute the payoff matrices.

PS. Aurelin accepted a draw in S15 from a winning position because he was tired. Maybe we should post these as chess problems? "Germany to play and win in 1."

I'm still a bit salty about this game.

Italy could have beat Germany had he actually moved his fleets North while I was slowing Germany down (still pretty proud of my moves there to slow him down, less proud of the fact that Germany stabbed me). And Germany could have soloed late in the game.

That game wasn't decided by chance, it was decided by press. Germany convinced Italy not to solo, and Italy convinced Germany to draw. Yes, it's possible that it could have ended in some sort of last minute decision, but it didn't.

[RoganJosh]

Yes, it wasn't a very good example. It was stated in the public chat that Italy accepted the draw because he was "tired," but you probably know better than me.

The second example I put is more to the point!

[Restitution]

If there's no luck in diplomacy, why does nobody have a 100% guess rate?

[Squigs44]

If there's no luck in diplomacy, why does nobody have a 100% guess rate?

That's really not a great example. Basketball players dont have a 100% free throw percentage. Not because they get unlucky, but because the skill level involved to do that is not really achievable by humans. It's possible for someone to outguess their opponents 100% of the time, but it's never going to happen.

[Restitution]

That's really not a great example. Basketball players dont have a 100% free throw percentage. Not because they get unlucky, but because the skill level involved to do that is not really achievable by humans. It's possible for someone to outguess their opponents 100% of the time, but it's never going to happen.

The theoretically best basketball free thrower is a robotic arm that will always get it in, but the theoretically best diplomacy guesser is still going to lose to one of the worst players every now and then.

I mean, if the best player in the world going to lose to a five year old some percentage of the time, it seems to me that that constitutes some degree of luck.

The best RPS algorithms (which outperform any human being in the world) still lose to humans some percentage of the time. So clearly RPS has luck in it. Given that Diplomacy very often has RPS-style conflicts in it, that seems to be an indication that luck exists in Diplomacy. A *perfect* diplomacy strategy will still occasionally lose to some idiot.

If the theoretically best possible player will still lose to a five year old on some occasions, to me, that's just what luck means.

[Restitution]

I mean, shit, only Laplace's Demon wouldn't be able to get 100% guess rate in Diplomacy. There's just no level of skill even theoretically possible where you can consistently win every guess. What is luck if not that? No matter how good you get, you're still going to lose sometimes. When the best player imaginable still loses, what could one call those losses other than "unlucky"?

[jmo1121109]

A full press game can end with only two powers of the board. In some cases that is an agreed two way draw. In some cases it's not. In the latter case, the game works just like a 1v1 game.

I'm sorry, I am not gonna spend hours looking through lists of old games. Especially since I don't think that your objection is particularly serious.

No, my objection is completely serious, I find your entire premise of a 50/50 in classic press game to be complete and utter rubbish. And no game gets to a 17-17 with the people still fighting in classic press.

[Squigs44]

Basically what it comes down to is that you are ascribing an incorrect guess to luck, and I am ascribing an incorrect guess to a players decision. You are saying that luck comes into play when there is incomplete information, and I am saying luck comes into play when the rules call for some sort of rng.

We both agree that a player cannot guarantee a win in certain situations because of incomplete information. We both agree that in a game theory situation, a player can have no greater success rate than a coin flip. We both agree (I think?) that game theory perfection does not exist in real life full press diplomacy.

Does that about sums things up?

[e.m.c^42]

In my small and inexperienced opinion, I feel like most of this discussion could be summed up to (proceed to terrible quotation paraphrasing)

What are the odds? Unbidden, a memory comes to him...The Teacher was describing a battle he fought in on the banks of the Komi River, how the tide had shifted in his favor at the last minute and how they had beaten the enemy into submission as the sun was setting.

Lucky break, he had said around a mouthful, and the Teacher had laughed. You make your own luck, kid.

Not luck, then...

and the debate here is between people who can make that luck, who can tilt those odds in their own favor - and those who can't.