Discussion of computer chess matches and engine tournaments.
Moderators: hgm , Rebel , chrisw
flok
Posts: 481 Joined: Tue Jul 03, 2018 10:19 am
Full name: Folkert van Heusden
Post
by flok » Thu May 23, 2019 3:55 pm
tpoppins wrote: ↑ Thu May 23, 2019 3:42 am
Embla reports some bizarre evals, by the way; e.g. a 106 cp advantage in a KRKR endgame:
Ah thank you for that!
That's a most interesting bug, and not a trivial one:
Code: Select all
folkert@msi:~/Projects/QueenBee/tags/v0.9.1_2205/trunk$ ./Embla
position fen 8/5R2/8/3K2k1/8/8/3r4/8 w - -
eval
# color: white
Total Evaluation: 0.300000 (white side)
Hmmm:
Code: Select all
./Embla
position fen 8/5R2/8/3K2k1/8/8/2r5/8 b - - 33 180
eval
# color: black
Total Evaluation: 0.060000 (white side)
Guenther
Posts: 4611 Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon
Post
by Guenther » Thu May 23, 2019 4:12 pm
flok wrote: ↑ Thu May 23, 2019 3:55 pm
tpoppins wrote: ↑ Thu May 23, 2019 3:42 am
Embla reports some bizarre evals, by the way; e.g. a 106 cp advantage in a KRKR endgame:
Ah thank you for that!
That's a most interesting bug, and not a trivial one:
Hmmm:
Code: Select all
./Embla
position fen 8/5R2/8/3K2k1/8/8/2r5/8 b - - 33 180
eval
# color: black
Total Evaluation: 0.060000 (white side)
Well, there is some difference between +0.06 and +1.06, where does the missing 1.00 eval come from?
https://rwbc-chess.de
trollwatch:
Talkchess nowadays is a joke - it is full of trolls/idiots/people stuck in the pleistocene > 80% of the posts fall into this category...
flok
Posts: 481 Joined: Tue Jul 03, 2018 10:19 am
Full name: Folkert van Heusden
Post
by flok » Thu May 23, 2019 4:16 pm
Guenther wrote: ↑ Thu May 23, 2019 4:12 pm
flok wrote: ↑ Thu May 23, 2019 3:55 pm
tpoppins wrote: ↑ Thu May 23, 2019 3:42 am
Embla reports some bizarre evals, by the way; e.g. a 106 cp advantage in a KRKR endgame:
Ah thank you for that!
That's a most interesting bug, and not a trivial one:
Hmmm:
Code: Select all
./Embla
position fen 8/5R2/8/3K2k1/8/8/2r5/8 b - - 33 180
eval
# color: black
Total Evaluation: 0.060000 (white side)
Well, there is some difference between +0.06 and +1.06, where does the missing 1.00 eval come from?
Exactly my point!
This probably comes from a tt-hit so now way I can reproduce this. That sucks.