Search found 5799 matches

by zullil
Fri Feb 14, 2020 11:10 am
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

1.Rb1 cxb1B 2.Bxb1 e3 3.Bxf5 e2 4.Bg4 e1Q 5.h3 Qe7 6.Kg1 Qb7 7.Kh2 = (-0.05) depth: 7 00:00:01 1226N, tb=22 Well, thanks for posting it because we won't know if you don't post it. But what is tb=22? Is that tablebase hits? What tablebases are you using? You should turn all TBs off. Maybe if it's ju...
by zullil
Thu Feb 13, 2020 8:22 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: mmt Vs. Ovyron (G4 D5 BG2)
Replies: 374
Views: 13062

Re: mmt Vs. Ovyron (G4 D5 BG2)

jp wrote:
Thu Feb 13, 2020 8:17 pm
Jouni wrote:
Thu Feb 13, 2020 6:05 pm
1 min/move selfplay from current position. Komodo 0-1 Houdini 0-1 Stockfish 0-1.
Can you post the games here?
Please, not while the game is ongoing!
by zullil
Thu Feb 13, 2020 6:13 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: mmt Vs. Ovyron (G4 D5 BG2)
Replies: 374
Views: 13062

Re: mmt Vs. Ovyron (G4 D5 BG2)

Jouni wrote:
Thu Feb 13, 2020 6:05 pm
1 min/move selfplay from current position. Komodo 0-1 Houdini 0-1 Stockfish 0-1.
What about with Komodo or Houdini playing Black against Stockfish?
by zullil
Thu Feb 13, 2020 12:24 pm
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

Interesting. Maybe because I have a higher starting depth, due to 2x GPUs? I have done a second run and it started with Rxf7. I was waiting 1:41 minutes and nothing changed. I have done a third run. From 0 to 25 seconds Rxf7 was the move. From 26 to 1:13 Nd6. From 1:14 to the end Rxf7 was the move....
by zullil
Thu Feb 13, 2020 11:45 am
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

Maybe you are using the wrong net or your GPUs are to weak. Running on my pc it doesn't switched and I have let it run up to depth 25. I'm sure it doesn't switched up to depth 21 and later to switch to Nd6 and back to Rxf7 doesn't make sense, because Nd6 needs to have better evaluation than +2.09, ...
by zullil
Thu Feb 13, 2020 11:19 am
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

LC0 Sergio 40x512 (320) The position is very easy for the 40x512 net and 40x512 runs up to depth 18 extreme fast. 1.Rb1 cxb1Q 2.Bxb1 e3 3.Bxf5 e2 4.f3 e1Q 5.Bg4 = (-0.05) depth: 6 00:00:0 1.Rh8+ Kg4 2.Rf8 c1Q 3.h3+ Kh5 4.Rxf5 Qd2 5.Bf7+ Kh4 6.Re5 Qf4 7.Re8 Qf3+ 8.Kg1 g4 9.Rh8+ Kg5 -/+ (-0.89) depth...
by zullil
Thu Feb 13, 2020 11:11 am
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

Maybe you are using the wrong net or your GPUs are to weak. Running on my pc it doesn't switched and I have let it run up to depth 25. I'm sure it doesn't switched up to depth 21 and later to switch to Nd6 and back to Rxf7 doesn't make sense, because Nd6 needs to have better evaluation than +2.09, ...
by zullil
Thu Feb 13, 2020 9:50 am
Forum: Computer Chess Club: General Topics
Topic: 83 nice Testpositions (Long and big post).
Replies: 102
Views: 24852

Re: 83 nice Testpositions (Long and big post).

... Look at first game (first diagram) here : http://www.chessbase.com/newsdetail.asp?newsid=4007 ... Blast from the past. This position still not easy (sometimes more than 1 minutes with 4 cores) for today : Rxf7!! [d]r1b1r1k1/1pBnnpp1/p1q1p2p/8/4N3/6Q1/2PRB1PP/5R1K w - - 2 2 This position is very...
by zullil
Wed Feb 12, 2020 6:12 pm
Forum: Computer Chess Club: General Topics
Topic: Funny fortress position
Replies: 30
Views: 2697

Re: Funny fortress position

Just asking for better static evals of near-tablebase positions :D. Bruh, that is 12m.... we don't even have 8m. Meanwhile, the method I implemented handles arbitrary piece count.... Bonus: If you happen to read the code, you will see it accomplishes the results it does by altering the static eval ...
by zullil
Wed Feb 12, 2020 5:18 pm
Forum: Computer Chess Club: General Topics
Topic: Funny fortress position
Replies: 30
Views: 2697

Re: Funny fortress position

Well, when someone creates a perfect eval, I'll be happy to implement that and do away with search entirely. I mean hey, it is just 32m TBs... easy right? Just asking for better static evals of near-tablebase positions :D. Yes, this is very hard, and not particularly interesting. This is currently ...