Luck Plays No Part in Diplomacy

[Restitution]

Yes, you can say you got lucky in chess. In fact, top players regularly do. The flip side of accepting what you could have done better in defeats is accepting what your opponent could have done better in victories, in both Diplomacy and chess. Attributing opponents' mistakes to your skill rather than forces outside your control is sometimes reasonable, but clearly not always. Now, that doesn't say much about the game itself, but I do believe that is the context in which most decent Diplomacy players use the phrase, and it's surprising to me that this isn't already the consensus in this conversation. I don't think defining luck in a way that means it's present everywhere means that you can't discuss to what degree a game is determined by it. So the statement "chess has luck" may be technically correct but entirely meaningless, because everything "has luck" in that sense. Diplomacy differs from chess because of (1) concurrent moves meaning your best move is determined by the opponent's decision, leading to the debate over whether "50-50" situations should be thought of probabilistically; and (2) limited influence over 3rd parties' actions outside the battles you are fighting. Since you can do both these things better or worse, but even the best players can get defeated by circumstances outside their control in both instances, IMO that's enough to draw a meaningful distinction between those two games. Luck plays a bigger role in Diplomacy than in chess, but obviously less than in Risk. That doesn't mean that better players don't overcome that variance and perform better consistently over time, and therefore improving your skills in these areas increases your chances of winning.

You put this better than I was able to, thank you.

[Carl Tuckerson]

I confess that I couldn't hope to be a bigger expert on self delusion than your good self, Wusti :)

Nevertheless my opinion remains the same. There are a great number of factors over which you have no or minimal control. If these happen to fall your way you will have a far easier time of it than if they fall against you. That, in my view, is clearly luck.

If you claim these factors don't exist, you are mistaken

If you claim that the positive or negative impact of these factors somehow don't count as luck under your definition then you are free to do so, but I'd argue that whatever definition you are using is so limited as to be useless.

So if I'm playing online chess and my opponent is so distracted by his real life that he plays very poorly, I got lucky, and therefore chess has luck? I actually feel that your definition is too broad to be useful in any sense.

Yes, you can say you got lucky in chess. In fact, top players regularly do. The flip side of accepting what you could have done better in defeats is accepting what your opponent could have done better in victories, in both Diplomacy and chess. Attributing opponents' mistakes to your skill rather than forces outside your control is sometimes reasonable, but clearly not always.

Perfectly put.
This distinction is what boiled my blood about the comment that people only believe this game has luck because they're excusing their losses. I'm excusing my successes FAR more than my failures when I evaluate my performance in postgame discussions, because I'm well aware that I caught a break somewhere that wasn't necessarily earned. I can point to actions I took that made the break more likely to happen, but the break happening was outside of my control. By calling it "lucky" I am being humble enough to recognize that--a lesson some people in this conversation could stand to learn.

[RoganJosh]

Sorry for the delay.

For http://webdiplomacy.net/board.php?gameID=163988 which part do you consider a 50/50 guess? And why do you believe that press played no impact in that especially when you see that France and England coordinated in their choice? Once I hear your reply to make sure we're not discussing different parts of that game I'll reply in detail.

Last season, in Autumn 11. Turkey needs either Berlin or Munich to win the game.

Turkey has two options:
A. Defend Munich with 3 + Attack Ber with 1
B. Defend Munich with 2 + Attack Ber with 2

England/France has two options
C. Attack Munich with 4 + Defend Ber with 1
D. Attack Munich with 3 + Defend Ber with 2

[Note: For both sides, there are a bunch of different sets of orders that realizes each of the options, but it doesn't matter exactly which set they pick.]

{A vs D} or {B vs C} and Turkey reaches 18 and wins
{A vs C} or {B vs D} and England/France completes the stalemate line and it is a draw.

Either way, the game will be over after this season.

The only coordination of orders is the France should order Bur S to Mun from Ber. That order works both with C and D.

Press? Well, Turkey could try to convince France to throw the game. But I think it is safe to say that's not gonna happen.

[Squigs44]

(Are you proud of me Squigs? I'm trying to use your mafia argument template, with kool formatting :D)

Lol

[Squigs44]

I confess that I couldn't hope to be a bigger expert on self delusion than your good self, Wusti :)

Nevertheless my opinion remains the same. There are a great number of factors over which you have no or minimal control. If these happen to fall your way you will have a far easier time of it than if they fall against you. That, in my view, is clearly luck.

If you claim these factors don't exist, you are mistaken

If you claim that the positive or negative impact of these factors somehow don't count as luck under your definition then you are free to do so, but I'd argue that whatever definition you are using is so limited as to be useless.

So if I'm playing online chess and my opponent is so distracted by his real life that he plays very poorly, I got lucky, and therefore chess has luck? I actually feel that your definition is too broad to be useful in any sense.

Yes, you can say you got lucky in chess. In fact, top players regularly do. The flip side of accepting what you could have done better in defeats is accepting what your opponent could have done better in victories, in both Diplomacy and chess. Attributing opponents' mistakes to your skill rather than forces outside your control is sometimes reasonable, but clearly not always. Now, that doesn't say much about the game itself, but I do believe that is the context in which most decent Diplomacy players use the phrase, and it's surprising to me that this isn't already the consensus in this conversation. I don't think defining luck in a way that means it's present everywhere means that you can't discuss to what degree a game is determined by it. So the statement "chess has luck" may be technically correct but entirely meaningless, because everything "has luck" in that sense. Diplomacy differs from chess because of (1) concurrent moves meaning your best move is determined by the opponent's decision, leading to the debate over whether "50-50" situations should be thought of probabilistically; and (2) limited influence over 3rd parties' actions outside the battles you are fighting. Since you can do both these things better or worse, but even the best players can get defeated by circumstances outside their control in both instances, IMO that's enough to draw a meaningful distinction between those two games. Luck plays a bigger role in Diplomacy than in chess, but obviously less than in Risk. That doesn't mean that better players don't overcome that variance and perform better consistently over time, and therefore improving your skills in these areas increases your chances of winning.

I like the perspective you have here, and think that it is a very useful way of looking at this question. Yes, there is a distinction between chess and diplomacy, in that there is imperfect information. We have different definitions of luck, but that doesn't matter much in the scheme of things. My main argument that I feel might not be fully accepted yet is that there is also a distinction between outguessing a human and guessing a coin flip, even in a 50/50 situation. If a player thinks that they can do no better than a coin flip in a 50/50, then they fail to recognize this distinction. There is no situation in diplomacy where the optimal play is to resign yourself to complete chance. Basically, imperfect information means that there is still information out there for you to sift through and inform your decision making process.

[Squigs44]

Axiom 1: No two players are of exactly the same skill.
Axiom 2: For each opponent, there is a nonzero chance that they will try to outguess me rather than flip a coin.

I completely missed this entry and it makes an interesting point.

It's the second axiom that is questionable. If the worse player knows that they are the worse player, then deciding to outguess rather than to flip a coin is to deliberately lower their own winning chances. So, you are assuming that for any player there is a nonzero chance that they will not do their best.

(Sure, some player will not use RNG's by principle...)

But you are touching upon an interesting aspect of the whole dynamics. In practice, if you know that you are the better player, then you might have nothing to gain from using a random number generator. We saw this before in the exchange about whether jmo or jmo's opponents should have used an RNG.

Another thing to add to this, is that part of getting in someone's head is to get them to overestimate their own skill level, and make them believe that they can outguess you. Maybe jmo can comment more, but it sounds like he draws people in emotionally, makes them want to prove themselves by outguessing him.

[ssorenn]

Any time chance is involved , there’s luck. Plain and simple.

[tr1285]

If a player thinks that they can do no better than a coin flip in a 50/50, then they fail to recognize this distinction. There is no situation in diplomacy where the optimal play is to resign yourself to complete chance. Basically, imperfect information means that there is still information out there for you to sift through and inform your decision making process.

But your opponent also has information about you. If you think your opponent is better at reading you than you are of them, wouldn't a coin flip at least even the odds again?

[Squigs44]

If a player thinks that they can do no better than a coin flip in a 50/50, then they fail to recognize this distinction. There is no situation in diplomacy where the optimal play is to resign yourself to complete chance. Basically, imperfect information means that there is still information out there for you to sift through and inform your decision making process.

But your opponent also has information about you. If you think your opponent is better at reading you than you are of them, wouldn't a coin flip at least even the odds again?

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

[Restitution]

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

The point of the article is that "Luck Plays No Part In Diplomacy", not "Being good at the game tilts the odds in your favor and minimizes the degree to which you need to rely on luck to succeed".

[Squigs44]

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

The point of the article is that "Luck Plays No Part In Diplomacy", not "Being good at the game tilts the odds in your favor and minimizes the degree to which you need to rely on luck to succeed".

Okay, you're playing word games again. First off, I made a disclaimer saying it was my opinion. Second off, the main point of an article is not always the title, it is the sum of its pieces and the conclusion drawn from the article. I will admit that my statement did fail to include several positions put forth in the article, and so I failed to summarize the whole point and rather gave a major point.

Take a look at this paragraph drawn from the second article:

"If my blog has a theme, the theme is self improvement in general and getting better at Diplomacy in particular. So let’s set aside the question of “What is Luck?” and instead ask what perspective on the role of luck is more helpful to you as a competitor?"

Given that the latter half of this article follows from this question and how to improve your skills, and that one of the three numbered points in the first article is titled: "Outguessing Your Opponents is a Skill; Maybe You Just Suck" (and the content after that point talks about improving your skills when in a guessing scenario), combined with the fact that the author of the articles +1'd my post, I would say that I was at least on the right track with that post.

[Restitution]

Maybe the author should try to write articles that aren't so badly written that they cause pages of argument and confusion about what the fuck he's talking about.

And not flipped out when people agreed with his thesis all along, apparently?

[Restitution]

Like, how is it even possible for somebody to explain the extremely basic concept of Yomi layers in a way that causes 9 pages of (apparent) misinterpretation and argument?

Either there's a conspiracy against the author, we're a bunch of braindead dullards, or the article is badly written at best.

[Carl Tuckerson]

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

The point of the article is that "Luck Plays No Part In Diplomacy", not "Being good at the game tilts the odds in your favor and minimizes the degree to which you need to rely on luck to succeed".

Okay, you're playing word games again. First off, I made a disclaimer saying it was my opinion. Second off, the main point of an article is not always the title, it is the sum of its pieces and the conclusion drawn from the article. I will admit that my statement did fail to include several positions put forth in the article, and so I failed to summarize the whole point and rather gave a major point.

Take a look at this paragraph drawn from the second article:

"If my blog has a theme, the theme is self improvement in general and getting better at Diplomacy in particular. So let’s set aside the question of “What is Luck?” and instead ask what perspective on the role of luck is more helpful to you as a competitor?"

Given that the latter half of this article follows from this question and how to improve your skills, and that one of the three numbered points in the first article is titled: "Outguessing Your Opponents is a Skill; Maybe You Just Suck" (and the content after that point talks about improving your skills when in a guessing scenario), combined with the fact that the author of the articles +1'd my post, I would say that I was at least on the right track with that post.

If that were it then this would have ended pages ago when it turned out that everyone quoted in the article as being in opposition to this message said they agreed with that message.

[tr1285]

If a player thinks that they can do no better than a coin flip in a 50/50, then they fail to recognize this distinction. There is no situation in diplomacy where the optimal play is to resign yourself to complete chance. Basically, imperfect information means that there is still information out there for you to sift through and inform your decision making process.

But your opponent also has information about you. If you think your opponent is better at reading you than you are of them, wouldn't a coin flip at least even the odds again?

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

We always want to improve our skills, however if outguessing your opponent is entirely based on a skill then your assertion that "There is no situation in diplomacy where the optimal play is to resign yourself to complete chance," is only true for one person, the highest-skilled player in the world. For anyone else playing against him or her, their optimal play should be determined by a coin flip. Because our skill is lower, we don't have a better than 50-50 chance to outguess.

[Squigs44]

But your opponent also has information about you. If you think your opponent is better at reading you than you are of them, wouldn't a coin flip at least even the odds again?

To this I would say what *I think* is the whole point of the article. Improve your skills so that you don't ever find yourself in a situation where you rely on pure chance.

We always want to improve our skills, however if outguessing your opponent is entirely based on a skill then your assertion that "There is no situation in diplomacy where the optimal play is to resign yourself to complete chance," is only true for one person, the highest-skilled player in the world. For anyone else playing against him or her, their optimal play should be determined by a coin flip. Because our skill is lower, we don't have a better than 50-50 chance to outguess.

My point is that anyone* can increase their skill and become higher skilled than their opponents. If you are the lower skilled player, your optimal play is not to say "Oh bother, I guess I should just flip a coin whenever I am in this situation", it is to practice, to read some articles, to get a mentor, to sleep with a picture of jmo under your pillow and hope that his skill seeps into you by osmosis, etc.

I guess if you find yourself in a situation where you have 30 seconds to make a decision and you feel that your opponent will outguess you, then picking randomly will yield a higher success rate, but if you find yourself in that situation, you did not play the game optimally.

So yes, I stand by my assertion that: "There is no situation in diplomacy where the optimal play is to resign yourself to complete chance", and I will continue to defend this assertion until people stop disagreeing, or someone proves me wrong.

*okay, obviously not anyone - there are people who are mentally handicapped

[jmo1121109]

Sorry for the delay.

For http://webdiplomacy.net/board.php?gameID=163988 which part do you consider a 50/50 guess? And why do you believe that press played no impact in that especially when you see that France and England coordinated in their choice? Once I hear your reply to make sure we're not discussing different parts of that game I'll reply in detail.

Last season, in Autumn 11. Turkey needs either Berlin or Munich to win the game.

Turkey has two options:
A. Defend Munich with 3 + Attack Ber with 1
B. Defend Munich with 2 + Attack Ber with 2

England/France has two options
C. Attack Munich with 4 + Defend Ber with 1
D. Attack Munich with 3 + Defend Ber with 2

[Note: For both sides, there are a bunch of different sets of orders that realizes each of the options, but it doesn't matter exactly which set they pick.]

{A vs D} or {B vs C} and Turkey reaches 18 and wins
{A vs C} or {B vs D} and England/France completes the stalemate line and it is a draw.

Either way, the game will be over after this season.

The only coordination of orders is the France should order Bur S to Mun from Ber. That order works both with C and D.

Press? Well, Turkey could try to convince France to throw the game. But I think it is safe to say that's not gonna happen.

So this game is interesting because it is anything but a coin flip. What I mean by that is there were 2 options available to England and France. The first was to engage in a single turn gamble where it was ensured that Munich could be taken, but not ensured that Turkey would not then take Berlin.

Or there was the ensured route of stopping the solo which was unfortunately missed.

Pie->Venice
Tyr support Apu to Rome or Tyr move to Naples
Tus support Apu to Rome
Apu->Rome

results in a forced disband of the Turkish fleet, and with the supported move into North Africa, the following turn Tyr Sea taps Ion Sea and Western Med and North Africa take Tunis. Game over, almost no risk, it's a very atypical move and easily missed because everyone tends to hyper focus on Germany for stalemate lines. Turkey would have had to take multiple risks that simply did not make sense for him to even have a hope of preventing that. You can swap around which unit attacks quite a lot so this scenario is pretty complex, not just a 50/50.

Now that said, let's pretend that Rome was locked down hard.

There are 3 enemy units bordering Berlin at the end of Spring 11. And 4 enemy units bordering Munich.

Everyone on the board should realize that England and France can take Munich without fail but only if they move from Berlin into Munich. Now Turkey actually did not realize this which we know from his global press statement of:

"5 Autumn, 1911: You know what sucks? If I'd supported myself to Berlin like i did in the Spring, I WOULD HAVE WON"

So his odds of solo'ing were actually 0, because he'd never considered the correct counter.

So right there I have to think that anyone playing with him all game may have been able to pick up diplomatically that he was weak tactically. This was also reflect in earlier failures like Spring 10 where he should have bounced England in Gulf of Lyons. But we can also see England failed to take Tyr in Autumn 10, so there was some lacking tactical skill all around.

So there's definitely a decent chance that any of the players involved could have been getting press reads around the skill failures to understand who was likely to do what.

Now getting back to what everyone should have known, it was that Turkey's only chance of surviving was using 2 units to support himself into Berlin.

Again everyone should see this, but might not. So this is where diplomacy has to play a part. Who's seen the possibilities, has the opponent said anything like "this is gonna be close" or is the press more like "you know you can't win can you just draw the game already?". And figuring out who's seen the options without revealing them to someone who hasn't...which takes some skill and probing. At that point you're introducing different factors that change anything from being luck by using your press to attempt to influence the outcome. Italy is still alive and possibly mad that he knows he's going to be eliminated because his allies are dumb, so Turkey can try working with his anger to get info on what England and France are going to do. There's a number of skill based things that should have been going on here.

So overall result is this game was not anywhere close to relying on a coin flip or luck. The most luck I'll give this ending, was that Turkey wasn't faced with anyone with any tactical skill who would have shut down even the slim solo chance he had sooner...though if he'd had any tactical skill he could have solo'ed with better moves in the 2 prior turns. But you get my point.

This game is so complex with so much going on that an entire game coming down to a single coin flip just is not feasible.

[jmo1121109]

Axiom 1: No two players are of exactly the same skill.
Axiom 2: For each opponent, there is a nonzero chance that they will try to outguess me rather than flip a coin.

I completely missed this entry and it makes an interesting point.

It's the second axiom that is questionable. If the worse player knows that they are the worse player, then deciding to outguess rather than to flip a coin is to deliberately lower their own winning chances. So, you are assuming that for any player there is a nonzero chance that they will not do their best.

(Sure, some player will not use RNG's by principle...)

But you are touching upon an interesting aspect of the whole dynamics. In practice, if you know that you are the better player, then you might have nothing to gain from using a random number generator. We saw this before in the exchange about whether jmo or jmo's opponents should have used an RNG.

Another thing to add to this, is that part of getting in someone's head is to get them to overestimate their own skill level, and make them believe that they can outguess you. Maybe jmo can comment more, but it sounds like he draws people in emotionally, makes them want to prove themselves by outguessing him.

So I do a number of things and it depends on the game. If the game is anonymous then I'm fine. What I'll do there is send stupid sounding press in a difficult scenario making it seem like I'm too dumb to have understood all the possible options and seem proud I figured out one. Done convincingly you can get stupidly high success rates with this.

I'll drop press hints to their ally along the lines of "look, you know that your ally is going to lose this center, it's not holdable, so you really need to consider stabbing them because they're an idiot". Which tends to have some success if you can pull off the bumbling idiot...though I've had this work in getting an ally to stab their ally...which was unexpected.

Sometimes I've even let someone take centers faster when I could have held them off because I want people who were counting on me to slow a threat down to have to break off and help me to stop a solo before they counted on it to prevent my eventual elimination.

In non anon games I play so infrequently that people tend to forget I'm halfway decent...which might be harder after this thread. And I'll just feel them out with various press/tactics/and how they previously moved in this game and prior ones.

Some of it is just gut feel too. Sometimes I'll plan out a game changing move years in advance. For example, http://webdiplomacy.net/board.php?gameID=226807. I spent years of game press stating over and over and over that I could not take Tunis to establish a pattern in my behavior of being defensive and not mentally considering aggressive possibilities down in the med. I continued with that loudly and aggressively in chainsaw fashion to draw attention away from that and then set myself up to have a shot at Venice so Italy could provide a trap which would temporarily need the use of Ionion in Adr, which should have, at that point, seemed safe because I was a defensive player. And then I took Tunis.

Press can be methodical with layers in layers and be planned years ahead to go hand in hand with tactical plays. How I handle it isn't the only way I've seen successful either. Players like Balki, MadMarx, Babak, etc were so smooth with press that you never ended up fighting them or just ended up so convinced you were going to lose you'd just enter all holds and let them win. It's magical to see lol. Other player use less press then me but more tactical analysis. It all depends, but the success and solo rates speak for themselves.

[RoganJosh]

Turkey has two options:
A. Defend Munich with 3 + Attack Ber with 1
B. Defend Munich with 2 + Attack Ber with 2

England/France has two options
C. Attack Munich with 4 + Defend Ber with 1
D. Attack Munich with 3 + Defend Ber with 2

Or there was the ensured route of stopping the solo which was unfortunately missed.

Pie->Venice
Tyr support Apu to Rome or Tyr move to Naples
Tus support Apu to Rome
Apu->Rome

This does not ensure to stop the solo. But it was silly of me to miss it. If TYR does not attack Tun, then Turkey can use ION to cut support.

Tun > TYS
Ven > Tus
ION > Apu
Rom > Tus
Tyr > Pie

Doesn't matter what attacking order England/France plays, the only center Turkey can loose is Tunis. Which means that - yes! - this is also a 50/50! Turkey has to risk either Rome or Tunis.

So there was actually two 50/50 guesses on the board, out of which Turkey had to get both right to win. So, in total, it is a 25/75.

Everyone on the board should realize that England and France can take Munich without fail but only if they move from Berlin into Munich. Now Turkey actually did not realize this which we know from his global press statement of:

"5 Autumn, 1911: You know what sucks? If I'd supported myself to Berlin like i did in the Spring, I WOULD HAVE WON"

So his odds of solo'ing were actually 0, because he'd never considered the correct counter.

You are jumping conclusions. There is nothing in his statement that implies that he missed the move.

Yes, everyone should realize that England/France can pick C. And that Turkey's reply is B. And that England/France's reply to B is D. And that Turkey's reply to D is A. And that England/France's reply to A is C. Actually, anyone who does not realize this is tactically inept. And that seems to be the only argument you put forward: "the players were probably tactically inept, and didn't realize that these were 50/50 guesses".

I think the players did realize these were 50/50 guesses. But it doesn't really matter. This is a position from a full press game which has two 50/50's left and nothing else. My claimed was that also full press game can end up in these positions, and, voilà!

[jmo1121109]

Turkey has two options:
A. Defend Munich with 3 + Attack Ber with 1
B. Defend Munich with 2 + Attack Ber with 2

England/France has two options
C. Attack Munich with 4 + Defend Ber with 1
D. Attack Munich with 3 + Defend Ber with 2

Or there was the ensured route of stopping the solo which was unfortunately missed.

Pie->Venice
Tyr support Apu to Rome or Tyr move to Naples
Tus support Apu to Rome
Apu->Rome

This does not ensure to stop the solo. But it was silly of me to miss it. If TYR does not attack Tun, then Turkey can use ION to cut support.

Tun > TYS
Ven > Tus
ION > Apu
Rom > Tus
Tyr > Pie

Doesn't matter what attacking order England/France plays, the only center Turkey can loose is Tunis. Which means that - yes! - this is also a 50/50! Turkey has to risk either Rome or Tunis.

So there was actually two 50/50 guesses on the board, out of which Turkey had to get both right to win. So, in total, it is a 25/75.

Everyone on the board should realize that England and France can take Munich without fail but only if they move from Berlin into Munich. Now Turkey actually did not realize this which we know from his global press statement of:

"5 Autumn, 1911: You know what sucks? If I'd supported myself to Berlin like i did in the Spring, I WOULD HAVE WON"

So his odds of solo'ing were actually 0, because he'd never considered the correct counter.

You are jumping conclusions. There is nothing in his statement that implies that he missed the move.

Yes, everyone should realize that England/France can pick C. And that Turkey's reply is B. And that England/France's reply to B is D. And that Turkey's reply to D is A. And that England/France's reply to A is C. Actually, anyone who does not realize this is tactically inept. And that seems to be the only argument you put forward: "the players were probably tactically inept, and didn't realize that these were 50/50 guesses".

I think the players did realize these were 50/50 guesses. But it doesn't really matter. This is a position from a full press game which has two 50/50's left and nothing else. My claimed was that also full press game can end up in these positions, and, voilà!

Literally nothing in your post was correct from a game analysis standpoint. I'll have to over the weekend to explain why.