A nice trick in the Zukertort or Hungarian openings (bonus in the Benko Gambit) • page 1/1 • Community Blog Discussions • lichess.org

Comments on https://lichess.org/@/totalnoob69/blog/a-nice-trick-in-the-zukertort-or-hungarian-openings-bonus-in-the-benko-gambit/eQfNI4dq

Messmaker

#2

Nice position, enjoyed analyzing it, but Stockfish 17.1 considers 4. ...b6 as the 22nd best move and 5. ... bc5 as the 7th best move, so I think the chances a decent opponent will allow this are very slim.

TotalNoob69

edited

#3

@Messmaker , it is probably impossible to find a fun position with Stockfish. I am using the Lichess database, so actual people playing actual games. Now, when the games become few, I am left with what a statistically insignificant part of humanity played, but trying to determine what people would play using SF is a losing strategy. Even Maiachess, which has the explicit goal to emulate human play, can't do it. It's impossible to predict people, apparently.

This is what Lichess DB for all games shows:
https://i.imgur.com/vO1bLjh.png

See that 1.54 eval computed for the sum of all games from that position? That's mathematically computed from winning stats, so, based on what all people played, whether is b6 or not, White wins in 59% of the games and thus the human evaluation of the position is 1.5.

@Messmaker , it is probably impossible to find a fun position with Stockfish. I am using the Lichess database, so actual people playing actual games. Now, when the games become few, I am left with what a statistically insignificant part of humanity played, but trying to determine what people would play using SF is a losing strategy. Even Maiachess, which has the explicit goal to emulate human play, can't do it. It's impossible to predict people, apparently. This is what Lichess DB for all games shows: https://i.imgur.com/vO1bLjh.png See that 1.54 eval computed for the sum of all games from that position? That's mathematically computed from winning stats, so, based on what all people played, whether is b6 or not, White wins in 59% of the games and thus the human evaluation of the position is 1.5.

TotalNoob69

#4

And if you think this is a statistical anomaly determined by bullet play by 500 rated players, here is the same thing, for players rated 1800+, Rapid+ and in the last year:

https://i.imgur.com/bJPkbZk.png

And if you think this is a statistical anomaly determined by bullet play by 500 rated players, here is the same thing, for players rated 1800+, Rapid+ and in the last year: https://i.imgur.com/bJPkbZk.png

Adam_Prikler

#5

I'm inspired to try this opening out, thanks for the post. And I found the shortcut links to be very convenient!

Dobby_the_Brave