Re: ChessBase's prophecy of 2004 fulfilled in 2019
Posted: Tue Jun 18, 2019 6:50 am
Well, just checked, 42580 at 1 node is equal to SF_dev (1 thread) depth = 8 .Milos wrote: ↑Mon Jun 17, 2019 5:06 pmI haven't tried Texel, but what I tried is SF with QS only i.e. depth=1 and I know I have struggled most of the time to even hold a draw (SF with QS only would punish each of my tactical blunders ruthlessly). And Lc0 with 11xxx net and no search was slightly better than SF with only QS.Laskos wrote: ↑Mon Jun 17, 2019 4:48 pmI myself was surprised, but Texel 1.07 at 4000 nodes/ move I speculated pretty firmly to be in 2000-2100 FIDE Elo range at 45min + 15s. Can you beat or draw consistently Texel at 4000 nodes/move at this TC? If not, then the result might be an artifact due to Texel's inability to capitalize on tactics and endgame. Also, 42580 with a bit less than 100 nodes/move beat Fruit 2.1 at blitz. Fruit at blitz is about 2800 FIDE blitz level, a very tough nut. Again, Fruit level might not translate into human level against Lc0.
SF depth 3 easily beats me (I think I am ~1700 FIDE or so), so the result is outstanding, pointing to above 2000 Fide level of 42580 1 node. Maybe you played the first run of Leela with those much weaker 15x192 nets? I remember "KingCrusher" (2200 FIDE blitz?) on YouTube playing those, and progressively he started to have difficulties against them, but at some 100 nodes, not 1. There seem some intransitivity, human players might have more difficulties against tactically astute AB engines of similar level than against Kimo's positional play, but tactical misses.