I’m sure someone else has posted the question I’m about to ask, but I don’t know where it is. The question is, if both white and black play a chess game perfectly, who would win, or would it be a draw?

Spock proved its a draw.

Most people, for at least 100 years, assume best play is a draw. Here are a few reasons:

1) Conceptually, the starting position is symmetrical, unorganized (the pieces aren't working together) and there is no contact between the opposing forces. Together this suggests the position is not only equal, but there's no mechanism to gain an advantage.

2) The drawing margin of many endgames is quite large e.g. a pawn up isn't enough to win with perfect play. In other words the difference between the equality of the starting position and the advantage needed to win a technical endgame is quite a lot.

3) Experimentally, as players get better they draw more often. This is true of humans and computer and across various time controls. As a fun example you can check Nakamura's blitz stats on this site, and he has maybe 10% draws. Now check a beginner and they'll probably have less than 1%.

In classical games the best players in the world draw over 50% of their games.

4) Computationally, engines continue to find ways for black to equalize in any opening. When top players want to win with white, long gone are the days of preparing a line that keeps an advantage. These days they prepare lines that try to keep a certain amount of "play" or winning chances in the position (a long dynamic game gives both sides chances to make a mistake).

A perfect game

Chess is extremely drawish! (more and more prominent when players become stronger or get more time control).

Previously people thought perfect chess require extremely precise play by both parties. However, we found out that perfect chess has extremely large rooms of play/ varieties of play as long as one side doesnt do serious blunders or a series of mistakes.

Let us set up 7 men tablebase, and as we can see a lot of moves end as perfect play except dropping pieces . ( This is a condition where one side has one pawn advantage already).

The same condition will happen in 32 men tablebase, (which we can call a solved chess)

If both white and black play a game perfectly, what would the result be

On chess.com they'd be both banned for fairplay

Black would accuse white of cheating/hacking/lag/delaying.

White would accuse black of cheating/hacking/lag/delaying.

Then it would turn racist because white gets to go first.

White would block black.

Black would block white.

The forum would get over 1000 posts.

I was going to add a witty GoP comment, but was to lazy to go back and edit.

It'd likely be a draw but then again, we (by we I mean the dozen or so masters who bothered to analyse them) wrongly evaluated some "simple" endgame positions as draws even thought tablebases eventually proved they were winning, as an example look up the KBBKN endgame. Another example is the following endgame produced by game 50 CCC11 finals played between Lc0 and SF (certainly not wimpy players) which was won by white, the blunder was apparently 43. ... f5 after which you get the following position:

No material advantages (or imbalances) and comparable king activity.

TBH, it doesn't matter that much if X position is winning or not if the technique is beyond human skill, perhaps it's more relevant in advanced chess but even then, the same argument applies but the bar is set much higher.

Hah, that's funny that f5 was the blunder. It's a move most human masters (by which I mean "lowly" 2200 players) would be reluctant to play precisely because such bishop endgames can be lost.

I didn't know top engines still made moves like this. That's fun to know.

Yes I know, facts can be funny.

draw!

Your position was already known by Stockfish programmers. The problem happened when perfect chess ( tablebases results) were interrupted by stupid FIDE rules 50 moves rules/ 75 moves rules.

Why stupid rules are affecting engine's performances? 

The reasons is " the state of game result changed with those rules". For example , SF see mate in 80, but due to 50 move rule, Stockfish store that result in her hash as draw. But when the game actually approaches with another 40 moves, the draw is not the draw anymore. ( actual mate in 80 become mate in 40)

In short, that lead to programmer error resulting Stockfish err. It is nothing to do with state of " being drawish in chess".

P.S. Normally engines stores search data in their hash memory. When the same position happens, engines doesnt need to search again and use stored memory data. But in above case, reuse of store data information become incorrect--》》 causing enging bug.

It would be a theoretical draw. As you see most not all, but most lower rated people don't draw as much as the higher-rated people that's because they make blunders and mistakes more easily. Higher rated people tend to draw more often because they tend not to make big blunders and mistakes, they usually play a-ok perfect game, granted with a couple of inaccuracies and maybe one or two small mistakes. So I think it is close to impossible to play a completely perfect game, but if one day some people do I think it should be a draw or something really close to a draw.

Oh, ok. That makes more sense, because IMO f5 is a pretty basic mistake.

If it would always be a draw, then that would logically mean that there is no advantage to having the white pieces. Think about it. If you go with the theory that it would always be a draw, you have to through out the centuries of thought that white has a slight advantage. You can't have it both ways.

That's very poor logic.

Try again.

No, that's perfect logic. You try again, or tell me how it's wrong.

What do morons consider logical?

I don't know, and I sincerely don't give a **** .