Lol, AI Department... Deep Learning? Can someone direct me to the correct development person?

Sort:
westnwestn

Good Morning, 

Was doing the normal chess routine this morning, and finally wanted to check in on this analysis question. I'll post this morning's game for a reference.

I was performing the ol' "bait and slaughter" technique and checked the analysis on it. The system treats it as a blunder and then rewards...bring you back to barely above zero.

As a fellow developer, Is there someone working on this? I have a couple of ideas that might lead to some big advances or shot dead in the street....who knows maybe I should just stick to my legos. But any info would be great! Thank You.

[Event "Live Chess"]
[Site "Chess.com"]
[Date "2022.02.28"]
[Round "-"]
[White "westnwestn"]
[Black "mrityunjaya99"]
[Result "1-0"]
[CurrentPosition "k5r1/p7/1pQ2p2/1N1r2np/7N/8/PPP2PPP/R4K2 b - -"]
[Timezone "UTC"]
[ECO "B01"]
[ECOUrl "https://www.chess.com/openings/Scandinavian-Defense"]
[UTCDate "2022.02.28"]
[UTCTime "12:22:07"]
[WhiteElo "808"]
[BlackElo "831"]
[TimeControl "600"]
[Termination "westnwestn won by resignation"]
[StartTime "12:22:07"]
[EndDate "2022.02.28"]
[EndTime "12:31:44"]
[Link "https://www.chess.com/game/live/39790117313"]

1. e4 d5 2. e5 f6 3. Nf3 Nc6 4. exf6 gxf6 5. Nh4 Be6 6. Bd3 Nh6 7. O-O Qd6 8.
Bg6+ $4 $17 {BLUNDER (-7.88)} ({(-1.17) The best move was} 8. Be2 O-O-O 9. d4 Nf5
10. Nxf5 Bxf5 11. c3 Rg8 12. Re1 e5 13. Be3 Bh3 14. g3 Qd7 15. Nd2 Kb8 16. Bf1
Bxf1) 8... Bf7 $4 $12 {BLUNDER (-0.43) Critical mistake.} ({(-7.88) The best move
was} 8... hxg6 9. Nf3 Ng4 10. h3 Nd4 11. Re1 Rxh3 12. gxh3 Nxf3+ 13. Qxf3 Qh2+
14. Kf1 Ne5 15. Qg3 Bxh3+ 16. Ke2 Bg4+ 17. Ke3 Bh6+ 18. Kd4 Qxg3 19. fxg3 Nf3+
20. Kc3 Nxe1 21. Kb3 Bh3 22. Nc3) 9. Bxf7+ Nxf7 {Critical move.} 10. Qg4 e6 11.
Re1 O-O-O 12. Rxe6 $2 $17 {MISTAKE (-4.44) Critical mistake.} ({(-0.80) The best
move was} 12. d4 Qd7 13. c3 Re8 14. Bf4 Bd6 15. Bxd6 Nxd6 16. Nd2 Rhg8 17. Qf3
f5 18. g3 Ne4 19. Qd1 e5 20. dxe5 Nxe5) 12... Qc5 $4 $16 {BLUNDER (+10.40) Critical
mistake. From winning to losing.} ({(-4.44) The best move was} 12... Qd7 13. Nf3
Nfe5 14. Nxe5 fxe5 15. c3 Bc5 16. Qh3 Rdf8 17. d4 exd4 18. b4 Bb6 19. Bh6 Rf7
20. Bg5 d3 21. Bh4 Nd8) 13. Rxc6+ Kb8 14. Rxc5 Bxc5 15. d4 Bd6 16. Bf4 h5 17.
Qg3 Rdg8 18. Qf3 Ng5 19. Qxd5 Bxf4 20. Nc3 Rd8 21. Qc4 Rhg8 22. d5 Nh3+ 23. Kf1
Ng5 $2 {MISTAKE (+11.40)} ({(+6.87) The best move was} 23... Rg4 24. Nf3 Ng5 25.
Nd4 Be5 26. Nce2 Ne4 27. Rd1 Bxh2 28. f3 Ng3+ 29. Kf2 Nxe2 30. fxg4 Nxd4 31.
Qxd4 hxg4 32. Qxf6 Rg8 33. c4 Bd6 34. Qf5 Ka8 35. b3 Bc5+ 36. Kg3) 24. Qxf4 b6
25. Nb5 Rxd5 $4 {BLUNDER (♔ Mate in 2)} ({(+11.77) The best move was} 25... Rd7
26. Nxa7 Rxd5 27. Nc6+ Kb7 28. Ne7 Rgd8 29. Nxd5 Rxd5 30. Ng6 h4 31. Qxf6 Ne4
32. Qxh4 Nd6 33. b4 Rd2 34. c4 Kb8 35. a3 Nf5 36. Qe4) 26. Qxc7+ Ka8 □ 27. Qc6+
{FASTER MATE (♔ Mate in 2)} ({(♔ Mate in 1) Checkmate after} 27. Qxa7#) (27.
Qc6+ Kb8 28. Qc7+ Ka8 29. Qxa7#) 1-0

Martin_Stahl

This club really isn't the right place for this, but what move are you concerned with? Evaluation accuracy is going to come down to evaluation depth and there may be some positions where the engine has to go really deep to see the promise of a move.