What is perfect chess?

Suppos, for the sake of argument scientists make a breakthrough in quantum computers and we had computers powerful enough to store a 32 piece tablebase and play "perfect chess". We would probably discover that the all knowing computer regards 1 a3 as just as good as 1 e4 or 1 d4 if, in fact, all those moves lead to a draw with correct play. Perhaps even g4 and f3 are no worse. A game between two perfect players might begin 1 a3, h5 2. Nh6, c6 and it is just as good as Rut Lopez assuming that the draw is maintained.

I doubt it. If chess is eventually "solved", even then there will be starting moves that are worse than others. I'm fairly sure a3 would be among those XD

Edit: Oh, now I see you're point. Unless a move like a3 randomly affected a "perfect" game 30 moves later or something, it really wouldn't matter what move white played. That's really cool.

Actually, patzermike, that's an interesting idea I'd never thought of.  If chess were solved, you could throw out terms like "winning chances" or "advantage" because every position would be either a known theoretical win/loss or a known theoretical draw.  I'd reckon that if chess is a theoretical draw (which I would suspect), then any first move by white is probably just as good as any other first move, in that strict sense of "still a drawn position".  The question of who has "practical" winning chances would be meaningless in this context.

I have a point. If chess with perfect play is a draw, then there's not a first move advantage. Therefore 1. a3 (which is almost a null move) would not make white's best outcome*  worse than the best outcome to black* (which is, by hypothesis, a draw).

*from the starting position

Ok, it took me a while, but I see where you're coming from. Interesting. 

But the question is not about "practical" chess. Let's point some things:

"My own experience on ICCF is trending toward a "formula" in terms of what it takes to win:

For white to win, black must make two key mistakes.

For black to win, white must make three key mistakes." 

As chess is a finite game, there's an algorithm that evaluates every position to the set {W, D, B}, where each value represents the outcome of the game given perfect play from both sides. It's actually easy to give this evaluation, and that's how tablebases are constructed (and since we're in a theoretical discussion, complexity is irrelevant). Based on that, there's no such thing as "For white to win, black must make two key mistakes.", because the evaluation is given to any position. If X is a draw position and black makes a mistake, then the new outcome is the white's victory. If not, then the new outcome is still a draw. 

"As such, none of the possible first moves are enough to lose the game outright"
This is a pretensious claim. What if there's a mate in 80 after 1. f3? 

" I've also found that black MUST want to play double-edged chess in order for white to have a chance at winning"
Again, the question in the thread is theoretical. There's no such thing as "chance" when perfect players are acting.

"It's already to the point where you don't need to a quantum computer to draw using the Berlin defense or the Petrov. If you have a modern multi-core PC, an opening database comprised of the ICCF archives, and played the Berlin or Petrov at correspondence time controls, not even a computer that could theoretically play "perfect" chess could beat you."
O.K., you're saying that Petrov defence leads to a draw with perfect play from both sides.
 

"I went into ICCF 7 years ago not knowing this, and at present I now know this to be fact."

Nice! Anedoctal evidence is better than science!

Seems to me that for an all knowing tablebase, if a game is drawn with best play, then all moves that hold the draw are equally "good". Only in a position that is decisive with best play we can define best move for the winning side as move, or one of moves, that mates fastest. Best move for losing side is move that delays mate longest.

Chess is matter of tactical intelligence for human and gives pleasure of intelligence, but man will make it dead if they solve it completely for "perfect Play". Then we do not need carlsen or anand we need just good memorizer rather than talent, chess will be like tic tac toe and managing sponsor for this game will be ridiculously tough.

So **** up perfect play by 'Super computer', it is enemy of chess.

Even if a computer could solve chess it wouldn't spoil chess for humans as memorizing the "solutions" would be impossible for any human being. Thus the super computers are not enemies of chess as we know it, maybe enemies of C vs C games.

"if chess were solved, you could throw out terms like "winning chances" or "advantage" because every position would be either a known theoretical win/loss or a known theoretical draw."

The winning chances as they are today have a real value to observers as they are a reasonable approximation most of the time of a GM's chance of winning. If chess was solved the absolute answer would be irrelevant in most cases since it had little predictive value to the outcome to H vs H games. If a better analysis provides a more accurate but less useful prediction it is in a practical sense not better and we'd be better of using less sofisticated engines. It is quite possible that we are close to that level already - ever better engines only improving by seeing solutions that no human could be expected to find.

Look you could compare it to a chess puzzle - we all know that with "perfect play" they are all solvable. That doesn't stop a lot of people devoting a lot of energy to solving them. Imagine if chess.com started publishing chess puzzles that might or might not have a solution - not sure if that would be popular or not...

Real chess games today are essentially puzzles where we don't know if there are real solutions. Even if we did know if a theoretical solution existed it would not impact the joy of playing for most people.

Of course, it is an enemy of chess, i do not understand my points.

Memorising will not be matter for full time professionals,
Moreover that chess played top players will be just fake human chess.
After solution their will be nothing for adventure in top playing like WC, besides there will be no interest among sponsors.

After solution i will not obey the world champs of new genartion like kasparov or others. They will be just mere dumb 'tablish' chess player.