Rook odds match: human vs engines.

[Fritz 0]

I don't use any GUI and i testing engines with 'chess' python library, I guess some GUI setting slow down the game with useless animation move.
I can confirm, in my slow 2Ghz laptop, a single game Dragon2.6 vs Dragon2.6 about half second
Thread: 1
UCI elo: 2000
Games: 100
Total time: 49s
Average: 0.49s

I have also try with 2300 UCI elo setting:
100 games in 76.8 seconds

And level 17 komodo 12.1.1 against itself:
100 games in 65.7 seconds

I haven't said it before, I always set 'varierty' to 10

You mean, you set "variety" to 10 AND a skill level? Do parameters even work with the levels, or are they set automatically with a certain level?

Yes, but skill levels already have inherent variety

It seems that the variety option has no effect when an Elo level is set; the Elo level has a defined variety value which will override whatever value you set for variety.

And what about other parameters? If I set a skill level and, for example, decrease "selectivity", will it play stronger due to less pruning?

[lkaufman]

I don't use any GUI and i testing engines with 'chess' python library, I guess some GUI setting slow down the game with useless animation move.
I can confirm, in my slow 2Ghz laptop, a single game Dragon2.6 vs Dragon2.6 about half second
Thread: 1
UCI elo: 2000
Games: 100
Total time: 49s
Average: 0.49s

I have also try with 2300 UCI elo setting:
100 games in 76.8 seconds

And level 17 komodo 12.1.1 against itself:
100 games in 65.7 seconds

I haven't said it before, I always set 'varierty' to 10

You mean, you set "variety" to 10 AND a skill level? Do parameters even work with the levels, or are they set automatically with a certain level?

Yes, but skill levels already have inherent variety

It seems that the variety option has no effect when an Elo level is set; the Elo level has a defined variety value which will override whatever value you set for variety.

And what about other parameters? If I set a skill level and, for example, decrease "selectivity", will it play stronger due to less pruning?

Changing parameters like Reduction and Selectivity with an Elo level (above 1700) set may change the play, but it's not so obvious in which direction, because above 1700 the strength is largely limited by nodes, which is quite different than the former method of limiting by depth.

[Fritz 0]

I don't use any GUI and i testing engines with 'chess' python library, I guess some GUI setting slow down the game with useless animation move.
I can confirm, in my slow 2Ghz laptop, a single game Dragon2.6 vs Dragon2.6 about half second
Thread: 1
UCI elo: 2000
Games: 100
Total time: 49s
Average: 0.49s

I have also try with 2300 UCI elo setting:
100 games in 76.8 seconds

And level 17 komodo 12.1.1 against itself:
100 games in 65.7 seconds

I haven't said it before, I always set 'varierty' to 10

You mean, you set "variety" to 10 AND a skill level? Do parameters even work with the levels, or are they set automatically with a certain level?

Yes, but skill levels already have inherent variety

It seems that the variety option has no effect when an Elo level is set; the Elo level has a defined variety value which will override whatever value you set for variety.

And what about other parameters? If I set a skill level and, for example, decrease "selectivity", will it play stronger due to less pruning?

Changing parameters like Reduction and Selectivity with an Elo level (above 1700) set may change the play, but it's not so obvious in which direction, because above 1700 the strength is largely limited by nodes, which is quite different than the former method of limiting by depth.

I run a 3600 game match overnight between Komodo 14.1 level 19 with Selectivity set to minimum (10) and regular Komodo 14.1 level 19. The former won by 38 Elo. Then started a match between Komodo 14.1 level 19 with Selectivity set to minimum (10) and Reduction set to minimum (-500) and regular Komodo 14.1 level 19. After 400+ games the former leads by 210 Elo. It seems that the quality of search is dramatically improved this way at this low depth (6 ply).

Now I will start the second test with Komodo 14.1 level 24 (12 ply).

[Fritz 0]

Wow. After 1600+ games, Komodo 14.1 level 24 with minimum Selectivity and Reduction has +236 Elo. But it doesn't move quite instantly, it takes maybe 15-20 seconds per game.

[lkaufman]

Wow. After 1600+ games, Komodo 14.1 level 24 with minimum Selectivity and Reduction has +236 Elo. But it doesn't move quite instantly, it takes maybe 15-20 seconds per game.

My comments were regarding Dragon 2.6 (or later), which uses nodes in the Elo settings. For older versions like K14.1, depth was the main parameter at the higher levels, so obviously using little selectivity or reduction will make it play much better and much slower.

[Fritz 0]

Wow. After 1600+ games, Komodo 14.1 level 24 with minimum Selectivity and Reduction has +236 Elo. But it doesn't move quite instantly, it takes maybe 15-20 seconds per game.

My comments were regarding Dragon 2.6 (or later), which uses nodes in the Elo settings. For older versions like K14.1, depth was the main parameter at the higher levels, so obviously using little selectivity or reduction will make it play much better and much slower.

Yes, I know. But I find it interesting that a full 12 ply search is perhaps World Champion strength in the classical time control.

[lkaufman]

Wow. After 1600+ games, Komodo 14.1 level 24 with minimum Selectivity and Reduction has +236 Elo. But it doesn't move quite instantly, it takes maybe 15-20 seconds per game.

My comments were regarding Dragon 2.6 (or later), which uses nodes in the Elo settings. For older versions like K14.1, depth was the main parameter at the higher levels, so obviously using little selectivity or reduction will make it play much better and much slower.

Yes, I know. But I find it interesting that a full 12 ply search is perhaps World Champion strength in the classical time control.

I don't recall precisely, but I think Deep Blue typically did something around 12 ply or so in the match where it beat Kasparov; if so this isn't surprising.

[Fritz 0]

Wow. After 1600+ games, Komodo 14.1 level 24 with minimum Selectivity and Reduction has +236 Elo. But it doesn't move quite instantly, it takes maybe 15-20 seconds per game.

My comments were regarding Dragon 2.6 (or later), which uses nodes in the Elo settings. For older versions like K14.1, depth was the main parameter at the higher levels, so obviously using little selectivity or reduction will make it play much better and much slower.

Yes, I know. But I find it interesting that a full 12 ply search is perhaps World Champion strength in the classical time control.

I don't recall precisely, but I think Deep Blue typically did something around 12 ply or so in the match where it beat Kasparov; if so this isn't surprising.

I think it was 14 ply, but obviously the knowlwdge of Komodo is not comparable to a brute force engine from 25 years ago.

[pepechuy]

Hi.

The fourth game against Komodo, removing the h1 rook.
Another loss for me.
Komodo deviated from the previous games with 5. g4 (instead of 5. Nc3, 5. e3 and 5. ba6).
I do not know if Komodo has some sort of inherent randomness, or if simply the amount of time it ponders on my turn can explain this variety of moves in the same position. ¿Or is it the room temperature?
Anyway, 5. g4 looks weird, but actually I think it is a good move, perhaps it has been played against the Volga gambit in non-handicap games.
Of course, I missed the mate in 3 at the end of the game. I was playing quite slowly, and was down to six minutes at that point... still, more than enough time to spot such a simple forced sequence. Guess, today was not my day.

[pgn]
[Event "banksia game"]
[Site "?"]
[Date "2022.05.01"]
[Round "?"]
[White "Komodo 12.1.1 64-bit"]
[Black "hm::Garcia Ruvalcaba, Jose de Jesus"]
[Result "1-0"]
[WhiteElo "3500"]
[Time "16:39:39"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBN1 w Qkq - 0 1"]
[Termination "mate"]

1. d4 {-3.64/24 59722 47970336} 1... Nf6 2. c4 {-3.64/23 44049 34765880} 2... c5
3. d5 {-3.36/25 40754 34633122} 3... b5 4. cxb5 {-3.45/23 28625 25038657} 4...
a6 5. g4 {-3.54/24 97364 79141583} 5... Nxg4 6. e4 {-3.52/23 24113 22130689}
6... Nxh2 7. Be2 {-3.79/22 35877 36115560} 7... axb5 8. Bf4 {-2.71/24 57858
49860004} 8... Qa5+ 9. Nc3 {-2.64/25 55559 52057502} 9... d6 10. Bxh2 {-2.61/24
61657 54781631} 10... b4 11. Nb5 {-2.69/30 10 168671204} 11... Ba6 12. a4
{-1.54/26 32860 32082510} 12... bxa3+ 13. Kf1 {-1.49/31 10 96346986} 13... a2
14. b4 {-1.59/26 26793 27593324} 14... cxb4 15. Qc2 {-1.47/24 9207 25323341}
15... Bxb5 16. Qc8+ {M+3/99 85 74793} 16... Qd8 17. Bxb5+ {M+2/99 10 3896} 17...
Nd7 18. Bxd7# {M+1/99 10 3762} 1-0
[/pgn]

After four rounds, the results (all from my perspective)

Crafty, a1 rook: 1/2, 1
Komodo, h1 rook: 0, 1/2, 0, 0
Lc0, a1 rook: 0, 1
Stockfish, h1 rook: 0, 1
Crafty, h1 rook: 1
Komodo, a1 rook: 0, 1
Lc0, h1 rook: 1
Stockfish, a1 rook: 1

I have dropped 7 points out of 15 games.
Taking the a1 rook: 2.5 dropped points in 7 games.
Taking the h1 rook: 4.5 dropped points in 8 games.
Against Crafty: 0.5 dropped points in 3 games.
Against Komodo: 4.5 dropped points in 6 games.
Against Lc0: 1 dropped point in 3 games.
Against Stockfish: 1 dropped point in 3 games.

The next game, again against Komodo removing the h1 rook, when I get the chance.

Greetings.

[pepechuy]

Hi to all.

Finally, the last game of the challenge.
The key to winning this game was just being patient.
When Komodo offered the queen exchange, I knew I was going to win (Komodo was low on time at that point, playing almost only on the increment).
I deviated from a previous game with 6... d6 (instead of 6... g6). I think both moves are good.

[pgn]
[Event "banksia game"]
[Site "?"]
[Date "2022.05.07"]
[Round "?"]
[White "Komodo 12.1.1 64-bit"]
[Black "hm::Garcia Ruvalcaba, Jose de Jesus"]
[Result "0-1"]
[WhiteElo "3500"]
[Time "18:32:25"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBN1 w Qkq - 0 1"]
[Termination "mate"]

1. d4 {-3.64/24 54246 47970336} 1... Nf6 2. c4 {-3.58/23 35205 30095482} 2... c5
3. d5 {-3.34/26 36878 33782745} 3... b5 4. cxb5 {-3.54/24 41781 37627930} 4...
a6 5. bxa6 {-3.63/24 104772 89217783} 5... Bxa6 6. Nc3 {-3.17/23 28097 21297037}
6... d6 7. Bg5 {-3.39/24 54711 41754023} 7... Nbd7 8. e4 {-3.37/25 39784
38399200} 8... Bxf1 9. Kxf1 {-3.14/24 17229 14608981} 9... g6 10. Nge2 {-3.26/24
30272 28227612} 10... Bg7 11. Kg1 {-3.37/24 78849 82921861} 11... O-O 12. b3
{-3.56/24 69066 65057073} 12... Qa5 13. Rc1 {-3.71/23 47951 44040992} 13... Rfb8
14. Be3 {-3.71/22 46130 39271045} 14... Ne8 15. h3 {-3.56/24 53753 47244284}
15... Nc7 16. Rc2 {-3.68/26 12230 17270077} 16... Nb5 17. Nxb5 {-3.82/28 23530
35609190} 17... Rxb5 18. Kh2 {-3.71/25 53422 51041591} 18... Nf6 19. f3
{-3.53/22 15868 14607531} 19... Ne8 20. Nc1 {-3.50/24 50353 44128340} 20... Nc7
21. a4 {-3.61/22 34989 32622045} 21... Rb7 22. Bd2 {-3.83/24 43017 40256178}
22... Qa7 23. g3 {-4.00/23 39307 37540529} 23... Na6 24. Be3 {-4.31/26 71048
72734481} 24... Nb4 25. Rc4 {-4.57/23 12572 12497729} 25... Qa6 26. Qe2
{-4.92/24 22815 24679641} 26... Rab8 27. Kg2 {-5.28/28 10 26017880} 27... Qa5
28. h4 {-3.70/22 35891 34271589} 28... Na6 29. Bd2 {-3.37/22 12858 11261524}
29... Qd8 30. h5 {-3.51/21 6520 9345515} 30... Nc7 31. Qd3 {-3.66/21 25727
24850743} 31... Na8 32. hxg6 {-3.39/21 10630 10186180} 32... hxg6 33. Rc2
{-3.38/21 10 4532672} 33... Nb6 34. a5 {-3.56/22 19610 19122159} 34... Nd7 35.
Qc4 {-3.70/22 10 10130083} 35... Ne5 36. Qe2 {-3.66/22 9462 9457145} 36... Qd7
37. Qe3 {-3.74/23 39 11170027} 37... Qb5 38. Be1 {-4.13/23 30053 30628557} 38...
Qa6 39. Qe2 {-4.42/22 10803 11857885} 39... Qxe2+ 40. Rxe2 {-4.51/24 10
14251100} 40... Ra8 41. Rc2 {-4.19/24 15614 16906204} 41... Nd7 42. Bd2
{-4.45/24 10 13152187} 42... Nb8 43. b4 {-4.71/24 11412 11512127} 43... Na6 44.
bxc5 {-5.00/28 10 44775486} 44... Nxc5 45. Kf2 {-4.93/26 16319 20659749} 45...
Nb3 46. Nxb3 {-3.86/22 9098 10645546} 46... Rxb3 47. Ke2 {-4.44/23 14856
19680121} 47... Rb2 48. Rxb2 {-3.84/26 11478 14776506} 48... Bxb2 49. Kd3
{-4.08/27 7976 13586812} 49... Kf8 50. Kc2 {-4.17/25 13649 17616423} 50... Be5
51. f4 {-4.00/24 13785 14760591} 51... Bd4 52. Kb3 {-4.14/24 15577 20820649}
52... Ke8 53. Kc4 {-4.20/25 10 7612815} 53... Bf2 54. g4 {-4.30/24 11269
14822672} 54... Kd7 55. f5 {-4.37/26 3262 12228669} 55... gxf5 56. gxf5
{-4.55/26 9726 14099765} 56... e6 57. dxe6+ {-3.79/23 10184 15168908} 57... fxe6
58. Bc3 {-4.19/26 13023 24407992} 58... exf5 59. Kd3 {-6.90/25 11901 19364708}
59... fxe4+ 60. Kxe4 {-6.85/27 3754 6509379} 60... Kc6 61. Kd3 {-7.24/24 7345
12349055} 61... Kb5 62. Ke4 {-7.35/30 13006 35278788} 62... Bh4 63. Kd5
{-7.09/28 8677 15023943} 63... Ra6 64. Ke4 {-10.96/31 11245 21729994} 64... Bd8
65. Kd5 {-11.02/32 7138 10491759} 65... Bxa5 66. Bd4 {-11.35/36 10 21059306}
66... Bb6 67. Bf6 {-250.00/34 18795 34831368} 67... Bc5 68. Bg7 {-250.00/22 2120
3943513} 68... Ra3 69. Ke6 {-250.00/23 3113 5390592} 69... Kc4 70. Bf6
{-250.00/20 2149 3950290} 70... d5 71. Bg5 {-250.00/24 39801 74440062} 71... d4
72. Kf5 {M-45/19 1388 2231651} 72... d3 73. Ke4 {M-42/18 1569 2090990} 73... Bb4
74. Kf3 {M-49/18 2362 4059957} 74... d2+ 75. Ke2 {M-6/21 1965 4033619} 75... Ra1
76. Bxd2 {M-69/22 1210 2475370} 76... Bxd2 77. Kxd2 {M-10/29 1366 3184146} 77...
Ra3 78. Ke2 {M-9/34 10 4523330} 78... Kd4 79. Kf2 {M-8/38 10 7456586} 79... Ke4
80. Kg2 {M-7/44 10 4164762} 80... Kf4 81. Kf2 {M-6/99 10 12005103} 81... Ra2+
82. Ke1 {M-6/69 1423 4152936} 82... Ke3 83. Kd1 {M-5/99 10 1165023} 83... Rh2
84. Kc1 {M-4/99 85 135243} 84... Kd3 85. Kb1 {M-3/99 10 6213} 85... Kc3 86. Ka1
{M-2/99 10 2767} 86... Kb3 87. Kb1 {M-1/99 10 197} 87... Rh1# 0-1
[/pgn]

I will follow up with a recap, and a preview of the next challenges, in a few hours.

Greetings.