A perfect game of chess is always a draw. Discuss.

OK then.

Bubber off!

I read the first couple of paragraphs of MAR's reply and reading further was unnecessary in this context because I completely agree with MAR that a complete tree leading to a tablebase is unnecessary for a proof. Should be obvious but never mind. It's also at the moment impossible but let's not let that put us off.

@203

This thread is about ultra-weakly solving chess:
determine if chess is a draw, a white win, or a black win.
To ultra-weakly solve chess it is unnecessary to weakly solve chess,
i.e. demonstrate a way to draw for black against all reasonable white moves.
To weakly solving chess it is unnecessary to strongly solve chess, i.e. a 32-men tablebase.

@186

"But the tygxc method starts by ignoring the majority of possible opening moves and responses and pruning the search using the opinions of a small group of masters. Even if this achieves the result you seek it is no proof that other lines would not give different results."
++ I said the present ICCF WC Finals with 100% draws is almost a weak solution to chess.
It is incomplete: it did not yet cover all reasonable white attempts to win.
It is redundant: it provides more lines to draw for black than the one that is needed.
Even if a winning move for white were found in some line, then still the other lines draw.
Even if the French were refuted, then the Najdorf, Petrov, open Ruy Lopez, Berlin lines are available to draw for black against 1 e4.
Even if the Grünfeld Indian Defense, Nimzovich Indian Defense, Queen's Indian Defense were refuted, then the Slav Defense, Queen's Gambit Declined, Queen's Gambit Accepted are available to draw for black against 1 d4.

As for other lines there is also logic.
It is logically clear that 1 a4, 1 a3, 1 b4, 1 b3, 1 c3, 1 d3 are logically inferior to 1 d4.
It is logically clear that 1 e3, 1 f3, 1 f4, 1 g3, 1 g4, 1 h3, 1 h4 are logically inferior to 1 e4.
It is logically clear that 1 Na3, 1 Nc3, 1 Nh3 are logically inferior to 1 Nf3.
Logically inferior does not mean game-theoretically inferior: 19 out of 20 initial white moves draw as well for white, but they make it easier for black to draw than 1 e4, 1 d4, 1 Nf3, 1 c4 in the sense that black has a wider range of moves to draw.
If black can draw against the best white moves, then black a fortiori can draw against the weaker white moves.

I also derived from these 100% drawn results in the ICCF WC Finals the ultra-weak solution to chess: Chess is a draw.

"It's still a matter of how little evidence an individual is willing to accept as "proof positive"."
++ It is also a matter of intelligence to understand a proof.

This is true and can be easily shown. Just go to the analysis and always do the top move on each side and you will get different positions but it will always be a draw at the end.

@194

"prove chess is a draw without any need for searching moves and therefore no need of computers."
++ The 97 ICCF WC draws each represent 2 * 1.5 = 3 years of computer time searching moves.

The late GM Sveshnikov was right when he said:

'Give me five years, good assistants and the latest computers
- I will bring all openings to technical endgames and kill (weakly solve) chess.'

Now the 17 ICCF (grand)masters are the 'good assistants',
and their computers are the 'latest computers'.
The ongoing ICCF World Championship Finals is almost a weak solution to Chess,
and it is redundant.
It does not yet exhaust all reasonable attempts for white to win,
but provides more succesful strategies to draw for black than the one that is necessary.

#205

"Easier"? What is easier to a snake is not to a lion or elephant. If both moves draw there is no "weaker" move between them. They are both equally good.

@208

"If both moves draw there is no "weaker" move between them. They are both equally good."
++ Yes, game-theoretically both 1 e4 and 1 Nh3 draw, but after 1 e4 black has a narrower choice of moves to draw than after 1 Nh3. Thus if black can draw against 1 e4, then he can a fortiori draw against 1 Nh3.

I showed beyond any doubt just two pages back that your derivation doesn't work. I tried it out and it didn't work. Full stop. End of.

You make no comment and just continue posting the same garbage.

It's been pointed out to you before on other threads. Are you trolling or genuinely brain dead?

@210

++ You yourself continue posting the same garbage.
I draw conclusions from 97 perfect games, successes, 5 days/move average.
You erroneously put forward 12 failures not even from the initial position.

Are you trolling or genuinely brain dead?

"A perfect game of chess is always a draw."
++ Yes, here are 97 perfect games, all draws.
https://www.iccf.com/event?id=100104

No - mate can be forced in 36 moves. See here (you can try playing Black).

In that case SF makes a particularly pathetic attempt even for SF (it can't play the endgame in general). A random legal move generator would probably have drawn against it.

there is not a very "perfect game". but if you mean two players very strong yes, its draw. just think of the GM's games, their matches are almost always draw. but if you think at computer's games, for example stockfish or alphazero their games ar rarely draw, although teorically they are perfect players. this is because doesn't exist the definitive perfect player or the definitive perfect game. so, a game of two player very strong yes, it's draw, but the perfect game doesn't exist.

I tryed to play a game stockfish of lichess VS stockfish of lichess. i gave they 5 minutes of calculing and after i maked the move who they think best. the game finished draw with a king and a bishop VS a king and a bishop. so, if you think stockfish of lichess is a perfect player yes, a perfect game is draw. but stockfish isn't the bestplayer (just think aplhazero beated it).

It depends on your point of view.

#214

Stockish and Alphazero are most definitely not perfect players.

And perfect games most definitely do exist, so long as you're talking theoretically.

But otherwise I agree.

@214

"the perfect game doesn't exist"
++ Here are 97 perfect games, all draws
https://www.iccf.com/event?id=100104

"stockfish or alphazero their games ar rarely draw"
++ They are all draws. In the TCEC competition they have to impose dubious openings.

"teorically they are perfect players"
++ Only with enough time per move. ICCF is 5 days/move average.

@213

"What exactly was erroneous about them?"

++ The title of this thread is 'A perfect game of chess is always a draw.'
The answer is: yes, and here are 97 perfect games 5 days/move, ICCF (grand)master + engine:
https://www.iccf.com/event?id=100104

A game of chess starts with the initial position, yours does not.
Your 12 games are certainly not perfect, as you know yourself.
The reason is that you did not give your computer enough time/move.
You cannot draw conclusions from letting your computer blitz against itself.
If you are serious and want to investigate, then go to the mistake it makes and compute from there with more time/move: 5 days/move.

This topic IS a draw, lul.

Yes. Mic drop.