lichess.org

If we solve chess, white can force win

Interestingly, I win as Black more often than I win as White, even when playing against myself. Especially in the most aggressive openings for Black, such as the Sicilian or KID. And this even though SF's eval of the latter always shows a slight advantage for White throughout much of the opening. (But I know a lot of people say SF is just plain wrong about the KID.)

So while I'm aware that, historically in high-level competitions, people try to win as White and draw as Black -- which really just means they take more risks as White, and play more defensibly as Black -- my own experience is that I'd rather be Black. Let White over-extend himself in the center; I will punish him for it from the flanks. :)

In fact, I have a pet theory that moving first is actually a disadvantage. As Onyx says, it telegraphs your intentions to your opponent; and this allows him decide how to respond -- such that Black usually ends up choosing the actual opening, and thereby steers the course of the game more than White. And even for White's best first moves, Black's best responses can frustrate, if not outright refute them. So if Black plays well, he _should_ at least draw, if not win. And White may soon come to regret his first move.

Another thing about the KID in particular is that, quite often, Black castles first. In fact, I find I often complete my development before White, who is busy with his pawns for some reason. So if I'm castled and my pieces are deployed, is that central pawn mass really an advantage? Does White really even have his extra tempo, anymore? . . . After all, time in chess isn't just the seconds on your clock _or_ moving first; it can be gained by strategically deploying, or lost by deploying in a sub-optimal way. So for example, when the game goes 1. d4 Nf6, White has spent his first move on a pawn, whereas Black has already moved a piece. Therefore -- arguably -- Black is ahead on time now, not White. Instead of a one-pawn advantage for White, it's a two-pawn advantage for Black (since a knight is worth 3). Or at least they're equal. (I guess it's also worth considering how d4 gives the dark-square bishop some scope.)

So anyway, it's certainly debatable. And therein lies all of chess: an eternal argument between White & Black, each one making his case as eloquently & forcefully as he can.

I don't think chess will ever be solved by humans , it's seems impossible right now...

Try to solve chess by removing all the blunders on every move and then add up all the options left per move.
The one with the most options should be holding a winning advantage. It's not knowing what is good, it's recognizing what is bad. What is left is good blunder free games. Cross train these blunder free games and chess will be solved quicker, without ever knowing all the possible billions of moves.

If engines tried to draw games, before trying so hard to win them, ... things might look much different and main lines might have already been solved. Like seeing only the trunk of a tree and knowing it's a tree without having to count every branch of a tree to say it is there and what type of tree it is. I think the main opening line approach can be solved before solving every branch. Count all the openings that favor white and count every opening that favors black. What is the result? Does it not add up in whites favor?

Since both sides know the advantage of holding a main line .... White would be trying to maintain the main line, while black would be trying to avoid it by transposing. Find a main line that cannot transpose. There are none. Find one that transposes the least in blacks favor and that is the main line that white should want to follow. If you find one, then chess is already getting solved. Once the complete main line is solved chess in this way, as far as I'm concerned chess is then solved.

Prune the variants down. Remove the guaranteed bad branches first, the rest is going to be solved later.

Yeah, forget using them to solve monetary policy and global warming. Let's use the most powerful computers to ruin a game humans play for fun.

#25: Monetary policy doesn't have a solution. If it did, I'd need to find a new line of work.
#21: Actually, White can play in order to make black think he will play in a different manner. To be fair though, I doubt many people look for Nash equilibria when preparing their opening repertoire.

If we solve chess, it will be a draw.

Chess has been solved weakly for a long time, humans don’t stand a chance.

A hard solution, no way. At the moment, you cannot prove that the basic position a Queen up is a win.

en.wikipedia.org/wiki/Shannon_number

How long would it take for a computer to play through every game in the 10^40 figure in the article to find out if chess is a draw, or white has a forced mate or black has a forced mate?

say if chess is a draw (or pick any of the two other options) how many games and how many moves would have to be memorised by a person to play perfectly?

I can't comprehend these large numbers like number of atoms in the universe.

Reconnecting