lichess.org
Donate

an opening variation that computers can't understand

veritely I won 2 times against Fritz 14,I have already posted the games,and everybody can win against a computer using that Traxler variation.
Please could you post that benko gambit starting position ? thank you
all of you are knowledgeable and presumptuos,I,m very stupid writing here
I looked at the Traxler in detail in the past and there is a certain amount of truth on both sides of the argument.

Lello33leone is right that it is very sharp and computer evaluations will swing wildly.
But I believe others are right and that the counter Traxler Gambit is losing by force, i.e. white wins.

Sidonia's solution to the sharpness of the opening is quite simple: it has winning lines in its opening book. I believe both KxBf2 and Kf1 are winning, but as I remember Kf1 gives a bigger quick material advantage and wins more easily.

If you check through Sidonia's games on here, it may appear like I am talking rubbish. I recall Sidonia losing at least one Traxler game as white. However I am sure that the reason for that is just that a 1500 Sidonia engine against a 2500 Leela meant she was simply outclassed once Sidonia was out of book.

So.
Yes its sharp and engine evaluations will swing wildly.
Yes engines will make mistakes if playing rapidly without a decent opening book, so a well prepared human might beat an unprepared program.

But you can put the winning lines in an opening book for a decent computer program, and the program will win. (Decent I mean 3000+, not a home brew program like Sidonia.)
Thank you,I think the opening book is the most important thing,for example Fritz 14 won 4-2
against Kramnik,but its book,chessbase opening book, is very weak in Traxler

This topic has been archived and can no longer be replied to.