ID 216 beats Fire 6.1

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

epanek
Posts: 28
Joined: Fri Mar 23, 2018 9:01 pm

ID 216 beats Fire 6.1

Post by epanek »

https://lichess.org/yPgm53s1

120' 10 TC

[Fire 6.1 x64 popcnt]
Hash=24500
Threads=4
MultiPV=4
Contempt=0
Ponder=false
UCI_Chess960=false
Syzygy50MoveRule=true
SyzygyPath=%ArenaDrive%\Users\Ed Panek\Desktop\TB2\dtz
SyzygyProbeDepth=8
SyzygyProbeLimit=6
jdart
Posts: 4435
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: ID 216 beats Fire 6.1

Post by jdart »

10. Nc3 is considered bad by the Hiarcs book, although it appears that line is not as bad as it first appears. But Black was better out of the opening. At this point:

[d] 4k2r/1p1b1pbp/p1r1p2p/3pP3/3P3P/q4N2/3QNPP1/2R2RK1 b k - 0 18

IMO Black should not have castled but maybe should have played .. Rxc1. However, that is not immediately obvious.

I have to say, this is starting to have some of the AlphaZero characteristics of finding moves that trip up the deep searchers. Still, this is one game and I suspect it would still lose to Fire 6 in a match.

--Jon
epanek
Posts: 28
Joined: Fri Mar 23, 2018 9:01 pm

Re: ID 216 beats Fire 6.1

Post by epanek »

I agree it would lose in a long match. On the other hand if Leela can bring about a game that it has played hundreds of thousands of times before pure brute search wont help much.

In general, unless there is some tactical playout Leela can compete with most engines at LTC>