Stockfish Family Gauntlet Series — Bullet

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

goni-K26
Posts: 217
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Stockfish Family Gauntlet Series — Bullet

Post by goni-K26 »

Tournament Setup (Stockfish Family Gauntlet)

Code: Select all

------------------|----------------------------------------------|
Feature           | Tournament Details                           |
------------------|----------------------------------------------|
Tournament Name:  | Stockfish Family Gauntlet                    |
Time Control:     | 9s + 0.1s (5% of Standard Blitz 3m + 2s)     |
Hash:             | 16 MB                                        |
Threads:          | Single Thread                                |
Tournament Type:  | Gauntlet                                     |
Rounds:           | Dependent on participants (50-100 Rounds)    |
Openings:         | Various openings                             |
Tablebases:       | 5-men Syzygy                                 |
GUI:              | Arena Chess GUI                              |
Hardware:         | Intel(R) Core(TM) i5-4570 3.20GHz, 8GB RAM   |
Operating System: | Windows 10/11 Pro x64                        |
------------------|----------------------------------------------|
Testing Stockfish 13 (HCE-5Pct)

Stockfish 13 x64 (HCE build) played 40 games against each of 50 opponent engines, 2000 games total. Final result: 1192.0/2000 (59.6%).

Analysis

A 59.6% overall score looks solid on paper, but it hides a huge spread: SF13's result against individual opponents ranges from 34% all the way up to 78% — a 44-point swing depending purely on who it faced.

Toughest opponents (SF13 came out behind):
Caissa 1.17 (BMI2) — 34% (Delta -117) — the single hardest test in the whole field
Starzix 6.0 and Motor 0.8.0 — both 40% (-70)
Caissa 1.19 (BMI2) — 41% (-61)
Caissa 1.15 (BMI2) and Lizard 11.1 — both 42% (-53)

Easiest opponents (SF13 dominated):
Quanticade 0.9 — 78% (+215) — SF13's best result of the whole gauntlet
Berserk 11.1 (BMI2) — 76% (+203)
Akimbo 0.8.0, Caissa 1.8.0 (BMI2), Viridithas 11.0.0, Lizard 10.2 (BMI2) — all tied at 74% (+179)

Version-to-Version Comparisons

Caissa shows the clearest development arc in the dataset. SF13 cruises against the old builds — 74% vs 1.8.0, 71% vs 1.9.0 — but from 1.13.1 onward the gap collapses fast: 50% vs 1.13.1, then 42% and 34% against 1.15/1.17, before a partial recovery to 41% vs 1.19. That's roughly a 40-point swing across the version range — somewhere between 1.9 and 1.17, Caissa closed almost the entire gap to SF13 and briefly overtook it.

Lizard jumps sharply between the 10.x and 11.x lines: 74% vs 10.2, 70% vs 10.4, then a steep drop to 46% vs 11.0 and 42% vs 11.1 — a major strength jump concentrated right at the 11.0 release.

Stormphrax follows the same shape: 69% vs 4.0.0, 61% vs 4.1.0, falling to 49% by 6.0.0.

Motor tightens with each release in a clean, steady line: 62% (0.5.0) > 46% (0.6.0) > 40% (0.8.0).

Velvet is noisier but trends the same way overall: 68% (6.0.0) > 62% (7.0.0) > 56% (7.1.0) > 55% (7.2.0) — then an unexpected rebound to 66% at 7.3.0, which breaks the otherwise clean downward curve and is worth a second look.

Build-variant note: a few engines appear twice under different CPU builds (Caissa 1.17 BMI2 at 34% vs AVX2 at 46%; Berserk 11.1 BMI2 at 76% vs AVX2 at 64%). Since BMI2/AVX2/SSE3 builds of the same version should be equal in playing strength, these gaps are most likely just normal statistical noise from a 40-game sample rather than a real strength difference.

Summary

Stockfish 13 (HCE) posted a comfortable 59.6% across the full 50-engine field, but that average hides real range — clearly outmatched by the strongest modern Caissa and Lizard builds, while crushing older or less mature engines. The real story isn't the headline score, it's how fast Caissa, Lizard, and Stormphrax closed the gap within just a few point releases.

Tournament Standings after 2000 Games vs 50 Teams

Code: Select all

+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
| Rank | Name                             | Points | Games | Live Elo | Start Elo |   +/- | Score% |
+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
|    1 | Stockfish 13 x64(HCE-5Pct)       | 1192,0 |  2000 |   2727,0 |    2731,0 |  -4,0 |   59,6 |
|    2 | Caissa 1.17 x64(BMI2-5Pct)       |   26,5 |    40 |   2811,2 |    2783,0 | +28,2 |   66,2 |
|    3 | Starzix 6.0 x64(AVX2-5Pct)       |   24,0 |    40 |   2790,7 |    2782,0 |  +8,7 |   60,0 |
|    4 | Motor 0.8.0 x64(AVX2-5Pct)       |   24,0 |    40 |   2776,0 |    2755,0 | +21,0 |   60,0 |
|    5 | Caissa 1.19 x64(BMI2-5Pct)       |   23,5 |    40 |   2792,2 |    2792,0 |  +0,2 |   58,8 |
|    6 | Caissa 1.15 x64(BMI2-5Pct)       |   23,0 |    40 |   2767,4 |    2754,0 | +13,4 |   57,5 |
|    7 | Lizard 11.1 x64(AVX2-5Pct)       |   23,0 |    40 |   2739,3 |    2702,0 | +37,3 |   57,5 |
|    8 | Caissa 1.17 x64(AVX2-5Pct)       |   21,5 |    40 |   2766,3 |    2774,0 |  -7,7 |   53,8 |
|    9 | Integral 5.0 x64(BMI2-5Pct)      |   21,5 |    40 |   2750,6 |    2745,0 |  +5,6 |   53,8 |
|   10 | Lizard 11.0 x64(AVX2-5Pct)       |   21,5 |    40 |   2727,3 |    2702,0 | +25,3 |   53,8 |
|   11 | Motor 0.6.0 x64(AVX2-5Pct)       |   21,5 |    40 |   2724,6 |    2697,0 | +27,6 |   53,8 |
|   12 | Igel 3.5.0 x64(AVX2-5Pct)        |   21,0 |    40 |   2701,4 |    2662,0 | +39,4 |   52,5 |
|   13 | Stormphrax 6.0.0 x64(BMI2-5Pct)  |   20,5 |    40 |   2742,0 |    2744,0 |  -2,0 |   51,2 |
|   14 | Clarity 7.0.0 x64(BMI2-5Pct)     |   20,5 |    40 |   2710,1 |    2685,0 | +25,1 |   51,2 |
|   15 | Caissa 1.13.1 x64(5Pct)          |   20,0 |    40 |   2693,9 |    2663,0 | +30,9 |   50,0 |
|   16 | Integral 5.0 x64(AVX2-5Pct)      |   19,5 |    40 |   2734,6 |    2745,0 | -10,4 |   48,8 |
|   17 | Caissa 1.15 x64(5Pct)            |   19,0 |    40 |   2722,5 |    2730,0 |  -7,5 |   47,5 |
|   18 | Obsidian 12.0 x64(AVX2-5Pct)     |   18,0 |    40 |   2744,9 |    2786,0 | -41,1 |   45,0 |
|   19 | Caissa 1.16 x64(5Pct)            |   18,0 |    40 |   2716,1 |    2733,0 | -16,9 |   45,0 |
|   20 | Velvet 7.2.0 x64(AVX2-5Pct)      |   18,0 |    40 |   2663,8 |    2638,0 | +25,8 |   45,0 |
|   21 | Velvet 7.1.0 x64(AVX2-5Pct)      |   17,5 |    40 |   2653,5 |    2627,0 | +26,5 |   43,8 |
|   22 | Caissa 1.15 x64(AVX2-5Pct)       |   17,0 |    40 |   2714,0 |    2744,0 | -30,0 |   42,5 |
|   23 | Viridithas 13.0.0 x64(AVX2-5Pct) |   16,5 |    40 |   2674,8 |    2679,0 |  -4,2 |   41,2 |
|   24 | PlentyChess 0.3.0 x64(AVX2-5Pct) |   16,5 |    40 |   2664,2 |    2660,0 |  +4,2 |   41,2 |
|   25 | Stormphrax 4.1.0 x64(BMI2-5Pct)  |   15,5 |    40 |   2630,5 |    2615,0 | +15,5 |   38,8 |
|   26 | Booot 7.1 x64(AVX2-5Pct)         |   15,5 |    40 |   2607,0 |    2576,0 | +31,0 |   38,8 |
|   27 | Motor 0.5.0 x64(AVX2-5Pct)       |   15,0 |    40 |   2647,8 |    2652,0 |  -4,2 |   37,5 |
|   28 | Clarity 6.0.0 x64(BMI2-5Pct)     |   15,0 |    40 |   2641,5 |    2641,0 |  +0,5 |   37,5 |
|   29 | Velvet 7.0.0 x64(AVX2-5Pct)      |   15,0 |    40 |   2630,0 |    2621,0 |  +9,0 |   37,5 |
|   30 | Berserk 11.1 x64(AVX2-5Pct)      |   14,5 |    40 |   2607,6 |    2590,0 | +17,6 |   36,2 |
|   31 | Komodo Dragon 1.0 x64(5Pct)      |   14,0 |    40 |   2628,4 |    2632,0 |  -3,6 |   35,0 |
|   32 | Peacekeeper 3.00 x64(AVX2-5Pct)  |   14,0 |    40 |   2606,0 |    2594,0 | +12,0 |   35,0 |
|   33 | Velvet 7.3.0 x64(AVX2-5Pct)      |   13,5 |    40 |   2635,8 |    2652,0 | -16,2 |   33,8 |
|   34 | Velvet 6.0.0 x64(AVX2-5Pct)      |   13,0 |    40 |   2563,4 |    2539,0 | +24,4 |   32,5 |
|   35 | Stormphrax 4.0.0 x64(BMI2-5Pct)  |   12,5 |    40 |   2601,8 |    2607,0 |  -5,2 |   31,2 |
|   36 | Rice 8.0 x64(AVX2-5Pct)          |   12,5 |    40 |   2592,2 |    2591,0 |  +1,2 |   31,2 |
|   37 | ChessCpp 4.10 x64(AVX2-5Pct)     |   12,5 |    40 |   2590,4 |    2588,0 |  +2,4 |   31,2 |
|   38 | Smallbrain 7.0 x64(AVX2-5Pct)    |   12,5 |    40 |   2550,8 |    2526,0 | +24,8 |   31,2 |
|   39 | Lizard 10.4 x64(AVX2-5Pct)       |   12,0 |    40 |   2600,2 |    2611,0 | -10,8 |   30,0 |
|   40 | RofChade 3.0 x64(SSE3-5Pct)      |   12,0 |    40 |   2576,5 |    2572,0 |  +4,5 |   30,0 |
|   41 | RubiChess 20220813 x64(5Pct)     |   11,5 |    40 |   2606,1 |    2628,0 | -21,9 |   28,7 |
|   42 | Wasp 7.00 x64(AVX-5Pct)          |   11,5 |    40 |   2593,8 |    2607,0 | -13,2 |   28,7 |
|   43 | Caissa 1.9.0 x64(BMI2-5Pct)      |   11,5 |    40 |   2558,6 |    2550,0 |  +8,6 |   28,7 |
|   44 | SlowChess Classic 2.9 x64(5Pct)  |   11,0 |    40 |   2569,1 |    2573,0 |  -3,9 |   27,5 |
|   45 | Arasan 24.2.1 x64(AVX2-5Pct)     |   11,0 |    40 |   2566,0 |    2568,0 |  -2,0 |   27,5 |
|   46 | Akimbo 0.8.0 x64(AVX2-5Pct)      |   10,5 |    40 |   2591,7 |    2617,0 | -25,3 |   26,2 |
|   47 | Caissa 1.8.0 x64(BMI2-5Pct)      |   10,5 |    40 |   2565,1 |    2573,0 |  -7,9 |   26,2 |
|   48 | Lizard 10.2 x64(BMI2-5Pct)       |   10,5 |    40 |   2562,0 |    2568,0 |  -6,0 |   26,2 |
|   49 | Viridithas 11.0.0 x64(5Pct)      |   10,5 |    40 |   2562,0 |    2568,0 |  -6,0 |   26,2 |
|   50 | Berserk 11.1 x64(BMI2-5Pct)      |    9,5 |    40 |   2575,4 |    2603,0 | -27,6 |   23,8 |
|   51 | Quanticade 0.9 x64(BMI2-5Pct)    |    9,0 |    40 |   2505,8 |    2501,0 |  +4,8 |   22,5 |
+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
HEAD-TO-HEAD MATRIX

Code: Select all

+----+----------------------------------+------+------+-------+-------+-------+
| No | Opponent                         | Wins | Loss | Draws | Games | Score |
+----+----------------------------------+------+------+-------+-------+-------+
|  1 | Caissa 1.17 x64(BMI2-5Pct)       |    4 |   17 |    19 |    40 | 33,8% |
|  2 | Starzix 6.0 x64(AVX2-5Pct)       |    2 |   10 |    28 |    40 | 40,0% |
|  3 | Motor 0.8.0 x64(AVX2-5Pct)       |    3 |   11 |    26 |    40 | 40,0% |
|  4 | Caissa 1.19 x64(BMI2-5Pct)       |    3 |   10 |    27 |    40 | 41,2% |
|  5 | Caissa 1.15 x64(BMI2-5Pct)       |    5 |   11 |    24 |    40 | 42,5% |
|  6 | Lizard 11.1 x64(AVX2-5Pct)       |    7 |   13 |    20 |    40 | 42,5% |
|  7 | Caissa 1.17 x64(AVX2-5Pct)       |    5 |    8 |    27 |    40 | 46,2% |
|  8 | Integral 5.0 x64(BMI2-5Pct)      |    7 |   10 |    23 |    40 | 46,2% |
|  9 | Lizard 11.0 x64(AVX2-5Pct)       |    6 |    9 |    25 |    40 | 46,2% |
| 10 | Motor 0.6.0 x64(AVX2-5Pct)       |    9 |   12 |    19 |    40 | 46,2% |
| 11 | Igel 3.5.0 x64(AVX2-5Pct)        |    7 |    9 |    24 |    40 | 47,5% |
| 12 | Stormphrax 6.0.0 x64(BMI2-5Pct)  |    8 |    9 |    23 |    40 | 48,8% |
| 13 | Clarity 7.0.0 x64(BMI2-5Pct)     |    6 |    7 |    27 |    40 | 48,8% |
| 14 | Caissa 1.13.1 x64(5Pct)          |    9 |    9 |    22 |    40 | 50,0% |
| 15 | Integral 5.0 x64(AVX2-5Pct)      |    9 |    8 |    23 |    40 | 51,2% |
| 16 | Caissa 1.15 x64(5Pct)            |    7 |    5 |    28 |    40 | 52,5% |
| 17 | Obsidian 12.0 x64(AVX2-5Pct)     |   10 |    6 |    24 |    40 | 55,0% |
| 18 | Caissa 1.16 x64(5Pct)            |   10 |    6 |    24 |    40 | 55,0% |
| 19 | Velvet 7.2.0 x64(AVX2-5Pct)      |   14 |   10 |    16 |    40 | 55,0% |
| 20 | Velvet 7.1.0 x64(AVX2-5Pct)      |   18 |   13 |     9 |    40 | 56,2% |
| 21 | Caissa 1.15 x64(AVX2-5Pct)       |   12 |    6 |    22 |    40 | 57,5% |
| 22 | Viridithas 13.0.0 x64(AVX2-5Pct) |   15 |    8 |    17 |    40 | 58,8% |
| 23 | PlentyChess 0.3.0 x64(AVX2-5Pct) |   13 |    6 |    21 |    40 | 58,8% |
| 24 | Stormphrax 4.1.0 x64(BMI2-5Pct)  |   13 |    4 |    23 |    40 | 61,3% |
| 25 | Booot 7.1 x64(AVX2-5Pct)         |   17 |    8 |    15 |    40 | 61,3% |
| 26 | Motor 0.5.0 x64(AVX2-5Pct)       |   16 |    6 |    18 |    40 | 62,5% |
| 27 | Clarity 6.0.0 x64(BMI2-5Pct)     |   19 |    9 |    12 |    40 | 62,5% |
| 28 | Velvet 7.0.0 x64(AVX2-5Pct)      |   18 |    8 |    14 |    40 | 62,5% |
| 29 | Berserk 11.1 x64(AVX2-5Pct)      |   14 |    3 |    23 |    40 | 63,7% |
| 30 | Komodo Dragon 1.0 x64(5Pct)      |   13 |    1 |    26 |    40 | 65,0% |
| 31 | Peacekeeper 3.00 x64(AVX2-5Pct)  |   14 |    2 |    24 |    40 | 65,0% |
| 32 | Velvet 7.3.0 x64(AVX2-5Pct)      |   18 |    5 |    17 |    40 | 66,2% |
| 33 | Velvet 6.0.0 x64(AVX2-5Pct)      |   19 |    5 |    16 |    40 | 67,5% |
| 34 | Stormphrax 4.0.0 x64(BMI2-5Pct)  |   20 |    5 |    15 |    40 | 68,8% |
| 35 | Rice 8.0 x64(AVX2-5Pct)          |   17 |    2 |    21 |    40 | 68,8% |
| 36 | ChessCpp 4.10 x64(AVX2-5Pct)     |   21 |    6 |    13 |    40 | 68,8% |
| 37 | Smallbrain 7.0 x64(AVX2-5Pct)    |   23 |    8 |     9 |    40 | 68,8% |
| 38 | Lizard 10.4 x64(AVX2-5Pct)       |   17 |    1 |    22 |    40 | 70,0% |
| 39 | RofChade 3.0 x64(SSE3-5Pct)      |   20 |    4 |    16 |    40 | 70,0% |
| 40 | RubiChess 20220813 x64(5Pct)     |   19 |    2 |    19 |    40 | 71,2% |
| 41 | Wasp 7.00 x64(AVX-5Pct)          |   19 |    2 |    19 |    40 | 71,2% |
| 42 | Caissa 1.9.0 x64(BMI2-5Pct)      |   20 |    3 |    17 |    40 | 71,2% |
| 43 | SlowChess Classic 2.9 x64(5Pct)  |   21 |    3 |    16 |    40 | 72,5% |
| 44 | Arasan 24.2.1 x64(AVX2-5Pct)     |   22 |    4 |    14 |    40 | 72,5% |
| 45 | Akimbo 0.8.0 x64(AVX2-5Pct)      |   24 |    5 |    11 |    40 | 73,8% |
| 46 | Caissa 1.8.0 x64(BMI2-5Pct)      |   24 |    5 |    11 |    40 | 73,8% |
| 47 | Lizard 10.2 x64(BMI2-5Pct)       |   23 |    4 |    13 |    40 | 73,8% |
| 48 | Viridithas 11.0.0 x64(5Pct)      |   21 |    2 |    17 |    40 | 73,8% |
| 49 | Berserk 11.1 x64(BMI2-5Pct)      |   24 |    3 |    13 |    40 | 76,2% |
| 50 | Quanticade 0.9 x64(BMI2-5Pct)    |   25 |    3 |    12 |    40 | 77,5% |
+----+----------------------------------+------+------+-------+-------+-------+
Note: Scores are reported from Stockfish's perspective against each opponent.
________________________________________________________________________________
Download Pgn Games(Zip)

Tournament Info

Code: Select all

Event:               Testing Stockfish 13 x64(HCE-5Pct)
Site/Hardware:       Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM(HP)
TimeControl:         9s + 0.1s Inc.
Participants:        51

Start:               2026.07.21 09:20:12
End:                 2026.07.22 22:27:57
Duration:            1d 13:07:45
Avg Game Duration:   00:01:06
Total Games:         2000

White Wins:               580(29,0%)
Black Wins:               456(22,8%)
Draws:                    964(48,2%)
Decisive Games (W/B):     1036(51,8%)
White/Black Win Ratio:    127,2%
White Scoring Probability: 53,1%

Longest Decisive Game:    Stockfish 13 x64(HCE-5Pct) vs Stormphrax 4.0.0 x64(BMI2-5Pct)  1-0  (260 plies)
Shortest Decisive Game:   Lizard 10.2 x64(BMI2-5Pct) vs Stockfish 13 x64(HCE-5Pct)  0-1  (45 plies)
Longest Draw:             Motor 0.6.0 x64(AVX2-5Pct) vs Stockfish 13 x64(HCE-5Pct)  1/2-1/2  (300 plies)
Shortest Draw:            RubiChess 20220813 x64(5Pct) vs Stockfish 13 x64(HCE-5Pct)  1/2-1/2  (64 plies)

Games with 30-60 moves:   918(45,9%)
Games with 61-90 moves:   831(41,5%)
Games with +90 moves:     234(11,7%)
Participants

Code: Select all

+----+----------------------------------+------+
| No | Name                             |  Elo |
+----+----------------------------------+------+
|  - | Stockfish 13 x64(HCE-5Pct)       | 2731 |
+----+----------------------------------+------+
|  1 | Caissa 1.19 x64(BMI2-5Pct)       | 2792 |
|  2 | Obsidian 12.0 x64(AVX2-5Pct)     | 2786 |
|  3 | Caissa 1.17 x64(BMI2-5Pct)       | 2783 |
|  4 | Starzix 6.0 x64(AVX2-5Pct)       | 2782 |
|  5 | Caissa 1.17 x64(AVX2-5Pct)       | 2774 |
|  6 | Motor 0.8.0 x64(AVX2-5Pct)       | 2755 |
|  7 | Caissa 1.15 x64(BMI2-5Pct)       | 2754 |
|  8 | Integral 5.0 x64(BMI2-5Pct)      | 2745 |
|  9 | Integral 5.0 x64(AVX2-5Pct)      | 2745 |
| 10 | Caissa 1.15 x64(AVX2-5Pct)       | 2744 |
| 11 | Stormphrax 6.0.0 x64(BMI2-5Pct)  | 2744 |
| 12 | Caissa 1.16 x64(5Pct)            | 2733 |
| 13 | Caissa 1.15 x64(5Pct)            | 2730 |
| 14 | Lizard 11.1 x64(AVX2-5Pct)       | 2702 |
| 15 | Lizard 11.0 x64(AVX2-5Pct)       | 2702 |
| 16 | Motor 0.6.0 x64(AVX2-5Pct)       | 2697 |
| 17 | Clarity 7.0.0 x64(BMI2-5Pct)     | 2685 |
| 18 | Viridithas 13.0.0 x64(AVX2-5Pct) | 2679 |
| 19 | Caissa 1.13.1 x64(5Pct)          | 2663 |
| 20 | Igel 3.5.0 x64(AVX2-5Pct)        | 2662 |
| 21 | PlentyChess 0.3.0 x64(AVX2-5Pct) | 2660 |
| 22 | Velvet 7.3.0 x64(AVX2-5Pct)      | 2652 |
| 23 | Motor 0.5.0 x64(AVX2-5Pct)       | 2652 |
| 24 | Clarity 6.0.0 x64(BMI2-5Pct)     | 2641 |
| 25 | Velvet 7.2.0 x64(AVX2-5Pct)      | 2638 |
| 26 | Komodo Dragon 1.0 x64(5Pct)      | 2632 |
| 27 | RubiChess 20220813 x64(5Pct)     | 2628 |
| 28 | Velvet 7.1.0 x64(AVX2-5Pct)      | 2627 |
| 29 | Velvet 7.0.0 x64(AVX2-5Pct)      | 2621 |
| 30 | Akimbo 0.8.0 x64(AVX2-5Pct)      | 2617 |
| 31 | Stormphrax 4.1.0 x64(BMI2-5Pct)  | 2615 |
| 32 | Lizard 10.4 x64(AVX2-5Pct)       | 2611 |
| 33 | Stormphrax 4.0.0 x64(BMI2-5Pct)  | 2607 |
| 34 | Wasp 7.00 x64(AVX-5Pct)          | 2607 |
| 35 | Berserk 11.1 x64(BMI2-5Pct)      | 2603 |
| 36 | Peacekeeper 3.00 x64(AVX2-5Pct)  | 2594 |
| 37 | Rice 8.0 x64(AVX2-5Pct)          | 2591 |
| 38 | Berserk 11.1 x64(AVX2-5Pct)      | 2590 |
| 39 | ChessCpp 4.10 x64(AVX2-5Pct)     | 2588 |
| 40 | Booot 7.1 x64(AVX2-5Pct)         | 2576 |
| 41 | Caissa 1.8.0 x64(BMI2-5Pct)      | 2573 |
| 42 | SlowChess Classic 2.9 x64(5Pct)  | 2573 |
| 43 | RofChade 3.0 x64(SSE3-5Pct)      | 2572 |
| 44 | Arasan 24.2.1 x64(AVX2-5Pct)     | 2568 |
| 45 | Viridithas 11.0.0 x64(5Pct)      | 2568 |
| 46 | Lizard 10.2 x64(BMI2-5Pct)       | 2568 |
| 47 | Caissa 1.9.0 x64(BMI2-5Pct)      | 2550 |
| 48 | Velvet 6.0.0 x64(AVX2-5Pct)      | 2539 |
| 49 | Smallbrain 7.0 x64(AVX2-5Pct)    | 2526 |
| 50 | Quanticade 0.9 x64(BMI2-5Pct)    | 2501 |
+----+----------------------------------+------+
Openings

Code: Select all

+----+-----+-------------------------+----------------------------------------------+------------------------------------------+-------+---------+---------+-------+
|  # | ECO | Opening                 | Variation                                    | Moves                                    | Count | White W | Black W | Draws |
+----+-----+-------------------------+----------------------------------------------+------------------------------------------+-------+---------+---------+-------+
|  1 | E60 | King's Indian Defense   | Normal Variation, King's Knight Variation    | 1. d4 Nf6 2. c4 g6 3. Nf3                |   100 |      34 |      16 |    50 |
|  2 | D10 | Slav Defense            | General                                      | 1. d4 d5 2. c4 c6                        |   100 |      27 |      17 |    56 |
|  3 | E10 | Indian Game             | Anti-Nimzo-indian                            | 1. d4 Nf6 2. c4 e6 3. Nf3                |   100 |      29 |      15 |    56 |
|  4 | D30 | Queen's Gambit Declined | General                                      | 1. d4 d5 2. c4 e6                        |   100 |      29 |      19 |    52 |
|  5 | C41 | Philidor Defense        | Lion Variation                               | 1. e4 e5 2. Nf3 d6 3. d4 Nd7 4. Nc3 Ngf6 |    95 |      27 |      29 |    39 |
|  6 | D11 | Slav Defense            | Modern Line                                  | 1. d4 d5 2. c4 c6 3. Nf3                 |   100 |      34 |      19 |    47 |
|  7 | C55 | Italian Game            | Two Knights Defense, Modern Bishop's Opening | 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3     |   100 |      26 |      24 |    50 |
|  8 | B50 | Sicilian Defense        | Delayed Alapin                               | 1. e4 c5 2. Nf3 d6 3. c3                 |   100 |      37 |      19 |    44 |
|  9 | B50 | Sicilian Defense        | Modern Variations                            | 1. e4 c5 2. Nf3 d6                       |   100 |      27 |      20 |    53 |
| 10 | B25 | Sicilian Defense        | Closed Variation, Traditional                | 1. e4 c5 2. Nc3 Nc6                      |   100 |      32 |      32 |    36 |
| 11 | E00 | Catalan Opening         | General                                      | 1. d4 Nf6 2. c4 e6 3. g3                 |   100 |      27 |      28 |    45 |
| 12 | D02 | Queen Pawn Game         | Symmetrical Variation                        | 1. d4 d5 2. Nf3 Nf6                      |   100 |      18 |      26 |    56 |
| 13 | B13 | Caro-Kann Defense       | Exchange Variation                           | 1. e4 c6 2. d4 d5 3. exd5                |   100 |      34 |      17 |    49 |
| 14 | E32 | Nimzo-Indian Defense    | Classical Variation                          | 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2     |   100 |      21 |      27 |    52 |
| 15 | B40 | Sicilian Defense        | Delayed Alapin Variation                     | 1. e4 c5 2. Nf3 e6 3. c3                 |   100 |      30 |      16 |    54 |
| 16 | A48 | Indian Game             | London System                                | 1. d4 Nf6 2. Nf3 g6 3. Bf4               |   100 |      31 |      24 |    45 |
| 17 | B23 | Sicilian Defense        | Grand Prix Attack                            | 1. e4 c5 2. Nc3 Nc6 3. f4                |   100 |      26 |      23 |    51 |
| 18 | B53 | Sicilian Defense        | Chekhover Variation                          | 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Qxd4    |   100 |      34 |      22 |    44 |
| 19 | B12 | Caro-Kann Defense       | Advance Variation, Short Variation           | 1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3       |   100 |      31 |      37 |    32 |
| 20 | B24 | Sicilian Defense        | Closed Variation, Fianchetto Variation       | 1. e4 c5 2. Nc3 Nc6 3. g3                |   100 |      23 |      25 |    52 |
+----+-----+-------------------------+----------------------------------------------+------------------------------------------+-------+---------+---------+-------+
Top
goni-K26
Posts: 217
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Re: Stockfish Family Gauntlet Series — Bullet

Post by goni-K26 »

Testing Stockfish 13

Stockfish 13 x64 (HCE build) played 80 games against each of 50 opponent engines, 4.000 games total. Final result: 2395,5/4000 (59.88%).

Analysis

Doubling the sample to 80 games/pairing didn't change the shape of the picture from the first round — it just made it sharper. SF13's score still swings from 38% to 77% depending on the opponent.

Hardest opponents:
Caissa 1.15 (BMI2) — 38% (Delta -84)
Caissa 1.17 (BMI2), Starzix 6.0, Motor 0.8.0 — all 41%

Easiest opponents:
Berserk 11.1 (BMI2) — 77% (+209)
Smallbrain 7.0, Viridithas 11.0.0 — both 76% (+203)

Notable version trend: The Caissa family remains the standout — SF13 struggles against every recent Caissa build (1.13.1 through 1.19, all in the 38-52% range) while crushing the older 1.8.0/1.9.0 builds (69-74%). This confirms the pattern from the first round and is now backed by double the games.

Report and analysis generated with the help of Claude (Anthropic)

Tournament Standings after 4.000 Games vs 50 Teams

Code: Select all

+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
| Rank | Name                             | Points | Games | Live Elo | Start Elo |   +/- | Score% |
+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
|    1 | Stockfish 13 x64(HCE-5Pct)       | 2395,5 |  4000 |   2725,8 |    2731,0 |  -5,2 |   59,9 |
|    2 | Caissa 1.15 x64(BMI2-5Pct)       |   49,5 |    80 |   2799,7 |    2754,0 | +45,7 |   61,9 |
|    3 | Caissa 1.17 x64(BMI2-5Pct)       |   47,0 |    80 |   2790,1 |    2783,0 |  +7,1 |   58,8 |
|    4 | Starzix 6.0 x64(AVX2-5Pct)       |   47,0 |    80 |   2789,8 |    2782,0 |  +7,8 |   58,8 |
|    5 | Motor 0.8.0 x64(AVX2-5Pct)       |   47,0 |    80 |   2783,3 |    2755,0 | +28,3 |   58,8 |
|    6 | Caissa 1.19 x64(BMI2-5Pct)       |   45,5 |    80 |   2782,3 |    2792,0 |  -9,7 |   56,9 |
|    7 | Lizard 11.1 x64(AVX2-5Pct)       |   44,0 |    80 |   2750,9 |    2702,0 | +48,9 |   55,0 |
|    8 | Caissa 1.17 x64(AVX2-5Pct)       |   42,5 |    80 |   2757,8 |    2774,0 | -16,2 |   53,1 |
|    9 | Stormphrax 6.0.0 x64(BMI2-5Pct)  |   42,0 |    80 |   2747,4 |    2744,0 |  +3,4 |   52,5 |
|   10 | Lizard 11.0 x64(AVX2-5Pct)       |   42,0 |    80 |   2737,5 |    2702,0 | +35,5 |   52,5 |
|   11 | Obsidian 12.0 x64(AVX2-5Pct)     |   41,5 |    80 |   2754,1 |    2786,0 | -31,9 |   51,9 |
|   12 | Integral 5.0 x64(BMI2-5Pct)      |   41,5 |    80 |   2744,3 |    2745,0 |  -0,7 |   51,9 |
|   13 | Motor 0.6.0 x64(AVX2-5Pct)       |   40,5 |    80 |   2726,3 |    2697,0 | +29,3 |   50,6 |
|   14 | Caissa 1.16 x64(5Pct)            |   40,0 |    80 |   2731,5 |    2733,0 |  -1,5 |   50,0 |
|   15 | Caissa 1.15 x64(AVX2-5Pct)       |   39,5 |    80 |   2730,7 |    2744,0 | -13,3 |   49,4 |
|   16 | Caissa 1.15 x64(5Pct)            |   38,5 |    80 |   2720,8 |    2730,0 |  -9,2 |   48,1 |
|   17 | Caissa 1.13.1 x64(5Pct)          |   38,5 |    80 |   2704,5 |    2663,0 | +41,5 |   48,1 |
|   18 | Clarity 7.0.0 x64(BMI2-5Pct)     |   38,0 |    80 |   2706,8 |    2685,0 | +21,8 |   47,5 |
|   19 | Integral 5.0 x64(AVX2-5Pct)      |   37,0 |    80 |   2714,3 |    2745,0 | -30,7 |   46,2 |
|   20 | Velvet 7.2.0 x64(AVX2-5Pct)      |   36,0 |    80 |   2681,1 |    2638,0 | +43,1 |   45,0 |
|   21 | Igel 3.5.0 x64(AVX2-5Pct)        |   35,5 |    80 |   2684,3 |    2662,0 | +22,3 |   44,4 |
|   22 | Viridithas 13.0.0 x64(AVX2-5Pct) |   34,0 |    80 |   2678,6 |    2679,0 |  -0,4 |   42,5 |
|   23 | PlentyChess 0.3.0 x64(AVX2-5Pct) |   30,5 |    80 |   2650,4 |    2660,0 |  -9,6 |   38,1 |
|   24 | Velvet 7.1.0 x64(AVX2-5Pct)      |   30,5 |    80 |   2641,2 |    2627,0 | +14,2 |   38,1 |
|   25 | Stormphrax 4.1.0 x64(BMI2-5Pct)  |   30,0 |    80 |   2634,2 |    2615,0 | +19,2 |   37,5 |
|   26 | Clarity 6.0.0 x64(BMI2-5Pct)     |   29,5 |    80 |   2638,6 |    2641,0 |  -2,4 |   36,9 |
|   27 | Velvet 7.3.0 x64(AVX2-5Pct)      |   29,0 |    80 |   2638,3 |    2652,0 | -13,7 |   36,2 |
|   28 | Velvet 7.0.0 x64(AVX2-5Pct)      |   29,0 |    80 |   2629,4 |    2621,0 |  +8,4 |   36,2 |
|   29 | Motor 0.5.0 x64(AVX2-5Pct)       |   27,5 |    80 |   2628,3 |    2652,0 | -23,7 |   34,4 |
|   30 | Wasp 7.00 x64(AVX-5Pct)          |   27,5 |    80 |   2615,0 |    2607,0 |  +8,0 |   34,4 |
|   31 | Rice 8.0 x64(AVX2-5Pct)          |   27,5 |    80 |   2609,7 |    2591,0 | +18,7 |   34,4 |
|   32 | Berserk 11.1 x64(AVX2-5Pct)      |   27,5 |    80 |   2609,3 |    2590,0 | +19,3 |   34,4 |
|   33 | RofChade 3.0 x64(SSE3-5Pct)      |   27,5 |    80 |   2602,8 |    2572,0 | +30,8 |   34,4 |
|   34 | Komodo Dragon 1.0 x64(5Pct)      |   27,0 |    80 |   2619,3 |    2632,0 | -12,7 |   33,8 |
|   35 | Peacekeeper 3.00 x64(AVX2-5Pct)  |   26,5 |    80 |   2604,0 |    2594,0 | +10,0 |   33,1 |
|   36 | Stormphrax 4.0.0 x64(BMI2-5Pct)  |   26,0 |    80 |   2605,0 |    2607,0 |  -2,0 |   32,5 |
|   37 | Arasan 24.2.1 x64(AVX2-5Pct)     |   26,0 |    80 |   2591,3 |    2568,0 | +23,3 |   32,5 |
|   38 | Lizard 10.4 x64(AVX2-5Pct)       |   25,5 |    80 |   2602,9 |    2611,0 |  -8,1 |   31,9 |
|   39 | Caissa 1.8.0 x64(BMI2-5Pct)      |   25,0 |    80 |   2586,5 |    2573,0 | +13,5 |   31,2 |
|   40 | RubiChess 20220813 x64(5Pct)     |   24,5 |    80 |   2601,5 |    2628,0 | -26,5 |   30,6 |
|   41 | Velvet 6.0.0 x64(AVX2-5Pct)      |   24,5 |    80 |   2569,7 |    2539,0 | +30,7 |   30,6 |
|   42 | ChessCpp 4.10 x64(AVX2-5Pct)     |   24,0 |    80 |   2585,3 |    2588,0 |  -2,7 |   30,0 |
|   43 | Booot 7.1 x64(AVX2-5Pct)         |   24,0 |    80 |   2581,0 |    2576,0 |  +5,0 |   30,0 |
|   44 | Akimbo 0.8.0 x64(AVX2-5Pct)      |   23,5 |    80 |   2591,5 |    2617,0 | -25,5 |   29,4 |
|   45 | Lizard 10.2 x64(BMI2-5Pct)       |   23,5 |    80 |   2574,7 |    2568,0 |  +6,7 |   29,4 |
|   46 | Quanticade 0.9 x64(BMI2-5Pct)    |   23,5 |    80 |   2545,6 |    2501,0 | +44,6 |   29,4 |
|   47 | SlowChess Classic 2.9 x64(5Pct)  |   21,0 |    80 |   2559,9 |    2573,0 | -13,1 |   26,2 |
|   48 | Caissa 1.9.0 x64(BMI2-5Pct)      |   20,5 |    80 |   2547,6 |    2550,0 |  -2,4 |   25,6 |
|   49 | Viridithas 11.0.0 x64(5Pct)      |   19,0 |    80 |   2544,7 |    2568,0 | -23,3 |   23,8 |
|   50 | Smallbrain 7.0 x64(AVX2-5Pct)    |   19,0 |    80 |   2527,3 |    2526,0 |  +1,3 |   23,8 |
|   51 | Berserk 11.1 x64(BMI2-5Pct)      |   18,5 |    80 |   2553,7 |    2603,0 | -49,3 |   23,1 |
+------+----------------------------------+--------+-------+----------+-----------+-------+--------+
HEAD-TO-HEAD MATRIX

Code: Select all

+----+----------------------------------+------+------+-------+-------+-------+
| No | Opponent                         | Wins | Loss | Draws | Games | Score |
+----+----------------------------------+------+------+-------+-------+-------+
|  1 | Caissa 1.15 x64(BMI2-5Pct)       |    7 |   26 |    47 |    80 | 38,1% |
|  2 | Caissa 1.17 x64(BMI2-5Pct)       |   12 |   26 |    42 |    80 | 41,2% |
|  3 | Starzix 6.0 x64(AVX2-5Pct)       |    6 |   20 |    54 |    80 | 41,2% |
|  4 | Motor 0.8.0 x64(AVX2-5Pct)       |    8 |   22 |    50 |    80 | 41,2% |
|  5 | Caissa 1.19 x64(BMI2-5Pct)       |    7 |   18 |    55 |    80 | 43,1% |
|  6 | Lizard 11.1 x64(AVX2-5Pct)       |   12 |   20 |    48 |    80 | 45,0% |
|  7 | Caissa 1.17 x64(AVX2-5Pct)       |   11 |   16 |    53 |    80 | 46,9% |
|  8 | Stormphrax 6.0.0 x64(BMI2-5Pct)  |   15 |   19 |    46 |    80 | 47,5% |
|  9 | Lizard 11.0 x64(AVX2-5Pct)       |   15 |   19 |    46 |    80 | 47,5% |
| 10 | Obsidian 12.0 x64(AVX2-5Pct)     |   15 |   18 |    47 |    80 | 48,1% |
| 11 | Integral 5.0 x64(BMI2-5Pct)      |   16 |   19 |    45 |    80 | 48,1% |
| 12 | Motor 0.6.0 x64(AVX2-5Pct)       |   18 |   19 |    43 |    80 | 49,4% |
| 13 | Caissa 1.16 x64(5Pct)            |   16 |   16 |    48 |    80 | 50,0% |
| 14 | Caissa 1.15 x64(AVX2-5Pct)       |   17 |   16 |    47 |    80 | 50,6% |
| 15 | Caissa 1.15 x64(5Pct)            |   18 |   15 |    47 |    80 | 51,9% |
| 16 | Caissa 1.13.1 x64(5Pct)          |   19 |   16 |    45 |    80 | 51,9% |
| 17 | Clarity 7.0.0 x64(BMI2-5Pct)     |   17 |   13 |    50 |    80 | 52,5% |
| 18 | Integral 5.0 x64(AVX2-5Pct)      |   19 |   13 |    48 |    80 | 53,8% |
| 19 | Velvet 7.2.0 x64(AVX2-5Pct)      |   25 |   17 |    38 |    80 | 55,0% |
| 20 | Igel 3.5.0 x64(AVX2-5Pct)        |   23 |   14 |    43 |    80 | 55,6% |
| 21 | Viridithas 13.0.0 x64(AVX2-5Pct) |   25 |   13 |    42 |    80 | 57,5% |
| 22 | PlentyChess 0.3.0 x64(AVX2-5Pct) |   30 |   11 |    39 |    80 | 61,9% |
| 23 | Velvet 7.1.0 x64(AVX2-5Pct)      |   36 |   17 |    27 |    80 | 61,9% |
| 24 | Stormphrax 4.1.0 x64(BMI2-5Pct)  |   31 |   11 |    38 |    80 | 62,5% |
| 25 | Clarity 6.0.0 x64(BMI2-5Pct)     |   35 |   14 |    31 |    80 | 63,1% |
| 26 | Velvet 7.3.0 x64(AVX2-5Pct)      |   34 |   12 |    34 |    80 | 63,7% |
| 27 | Velvet 7.0.0 x64(AVX2-5Pct)      |   34 |   12 |    34 |    80 | 63,7% |
| 28 | Motor 0.5.0 x64(AVX2-5Pct)       |   33 |    8 |    39 |    80 | 65,6% |
| 29 | Wasp 7.00 x64(AVX-5Pct)          |   36 |   11 |    33 |    80 | 65,6% |
| 30 | Rice 8.0 x64(AVX2-5Pct)          |   30 |    5 |    45 |    80 | 65,6% |
| 31 | Berserk 11.1 x64(AVX2-5Pct)      |   36 |   11 |    33 |    80 | 65,6% |
| 32 | RofChade 3.0 x64(SSE3-5Pct)      |   32 |    7 |    41 |    80 | 65,6% |
| 33 | Komodo Dragon 1.0 x64(5Pct)      |   31 |    5 |    44 |    80 | 66,2% |
| 34 | Peacekeeper 3.00 x64(AVX2-5Pct)  |   33 |    6 |    41 |    80 | 66,9% |
| 35 | Stormphrax 4.0.0 x64(BMI2-5Pct)  |   38 |   10 |    32 |    80 | 67,5% |
| 36 | Arasan 24.2.1 x64(AVX2-5Pct)     |   39 |   11 |    30 |    80 | 67,5% |
| 37 | Lizard 10.4 x64(AVX2-5Pct)       |   34 |    5 |    41 |    80 | 68,1% |
| 38 | Caissa 1.8.0 x64(BMI2-5Pct)      |   40 |   10 |    30 |    80 | 68,8% |
| 39 | RubiChess 20220813 x64(5Pct)     |   36 |    5 |    39 |    80 | 69,4% |
| 40 | Velvet 6.0.0 x64(AVX2-5Pct)      |   39 |    8 |    33 |    80 | 69,4% |
| 41 | ChessCpp 4.10 x64(AVX2-5Pct)     |   42 |   10 |    28 |    80 | 70,0% |
| 42 | Booot 7.1 x64(AVX2-5Pct)         |   43 |   11 |    26 |    80 | 70,0% |
| 43 | Akimbo 0.8.0 x64(AVX2-5Pct)      |   44 |   11 |    25 |    80 | 70,6% |
| 44 | Lizard 10.2 x64(BMI2-5Pct)       |   43 |   10 |    27 |    80 | 70,6% |
| 45 | Quanticade 0.9 x64(BMI2-5Pct)    |   41 |    8 |    31 |    80 | 70,6% |
| 46 | SlowChess Classic 2.9 x64(5Pct)  |   43 |    5 |    32 |    80 | 73,8% |
| 47 | Caissa 1.9.0 x64(BMI2-5Pct)      |   44 |    5 |    31 |    80 | 74,4% |
| 48 | Viridithas 11.0.0 x64(5Pct)      |   47 |    5 |    28 |    80 | 76,2% |
| 49 | Smallbrain 7.0 x64(AVX2-5Pct)    |   52 |   10 |    18 |    80 | 76,2% |
| 50 | Berserk 11.1 x64(BMI2-5Pct)      |   48 |    5 |    27 |    80 | 76,9% |
+----+----------------------------------+------+------+-------+-------+-------+
Note: Scores are reported from Stockfish's perspective against each opponent.
________________________________________________________________________________
Download Pgn Games(Zip)

Tournament Info

Code: Select all

[color=#000088][b]Tournament Info[/b][/color]
[code]
Event:                     Testing Stockfish 13 x64(HCE-5Pct)
Site/Hardware:             Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM
TimeControl:               9s + 0.1s Inc.
Participants:              51
                           
Start:                     2026.07.21 09:20:12
End:                       2026.07.24 12:30:34
Duration:                  3d 03:10:22
Avg Game Duration:         00:01:07
Total Games:               4.000
                           
White Wins:                1196(29,9%)
Black Wins:                863(21,6%)
Draws:                     1941(48,5%)
Decisive Games (W/B):      2059(51,5%)
White/Black Win Ratio:     138,6%
White Scoring Probability:  54,2%

Longest Decisive Game:     Booot 7.1 x64(AVX2-5Pct) vs Stockfish 13 x64(HCE-5Pct)  0-1  (274 plies)
Shortest Decisive Game:    Stockfish 13 x64(HCE-5Pct) vs RofChade 3.0 x64(SSE3-5Pct)  0-1  (39 plies)
Longest Draw:              Motor 0.6.0 x64(AVX2-5Pct) vs Stockfish 13 x64(HCE-5Pct)  1/2-1/2  (300 plies)
Shortest Draw:             RubiChess 20220813 x64(5Pct) vs Stockfish 13 x64(HCE-5Pct)  1/2-1/2  (64 plies)
                           
Games with 30-60 moves:    1828(45,7%)
Games with 61-90 moves:    1639(41,0%)
Games with +90 moves:      503(12,6%)
Openings

Code: Select all

+----+-----+-----------------------------------------+----------------------------------------------+------------------------------------------------+-------+---------+---------+-------+
|  # | ECO | Opening                                 | Variation                                    | Moves                                          | Count | White W | Black W | Draws |
+----+-----+-----------------------------------------+----------------------------------------------+------------------------------------------------+-------+---------+---------+-------+
|  1 | E60 | King's Indian Defense                   | Normal Variation, King's Knight Variation    | 1. d4 Nf6 2. c4 g6 3. Nf3                      |   100 |      34 |      16 |    50 |
|  2 | D10 | Slav Defense                            | General                                      | 1. d4 d5 2. c4 c6                              |   100 |      27 |      17 |    56 |
|  3 | E10 | Indian Game                             | Anti-Nimzo-indian                            | 1. d4 Nf6 2. c4 e6 3. Nf3                      |   100 |      29 |      15 |    56 |
|  4 | D30 | Queen's Gambit Declined                 | General                                      | 1. d4 d5 2. c4 e6                              |   100 |      29 |      19 |    52 |
|  5 | C41 | Philidor Defense                        | Lion Variation                               | 1. e4 e5 2. Nf3 d6 3. d4 Nd7 4. Nc3 Ngf6       |    95 |      27 |      29 |    39 |
|  6 | D11 | Slav Defense                            | Modern Line                                  | 1. d4 d5 2. c4 c6 3. Nf3                       |    99 |      33 |      19 |    47 |
|  7 | C55 | Italian Game                            | Two Knights Defense, Modern Bishop's Opening | 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3           |   100 |      26 |      24 |    50 |
|  8 | B50 | Sicilian Defense                        | Delayed Alapin                               | 1. e4 c5 2. Nf3 d6 3. c3                       |   100 |      37 |      19 |    44 |
|  9 | B50 | Sicilian Defense                        | Modern Variations                            | 1. e4 c5 2. Nf3 d6                             |   100 |      27 |      20 |    53 |
| 10 | B25 | Sicilian Defense                        | Closed Variation, Traditional                | 1. e4 c5 2. Nc3 Nc6                            |   100 |      32 |      32 |    36 |
| 11 | E00 | Catalan Opening                         | General                                      | 1. d4 Nf6 2. c4 e6 3. g3                       |   100 |      27 |      28 |    45 |
| 12 | D02 | Queen Pawn Game                         | Symmetrical Variation                        | 1. d4 d5 2. Nf3 Nf6                            |   100 |      18 |      26 |    56 |
| 13 | B13 | Caro-Kann Defense                       | Exchange Variation                           | 1. e4 c6 2. d4 d5 3. exd5                      |   100 |      34 |      17 |    49 |
| 14 | E32 | Nimzo-Indian Defense                    | Classical Variation                          | 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2           |   100 |      21 |      27 |    52 |
| 15 | B40 | Sicilian Defense                        | Delayed Alapin Variation                     | 1. e4 c5 2. Nf3 e6 3. c3                       |   100 |      30 |      16 |    54 |
| 16 | A48 | Indian Game                             | London System                                | 1. d4 Nf6 2. Nf3 g6 3. Bf4                     |   100 |      31 |      24 |    45 |
| 17 | B23 | Sicilian Defense                        | Grand Prix Attack                            | 1. e4 c5 2. Nc3 Nc6 3. f4                      |   100 |      26 |      23 |    51 |
| 18 | B53 | Sicilian Defense                        | Chekhover Variation                          | 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Qxd4          |   100 |      34 |      22 |    44 |
| 19 | B12 | Caro-Kann Defense                       | Advance Variation, Short Variation           | 1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3             |   100 |      31 |      37 |    32 |
| 20 | B24 | Sicilian Defense                        | Closed Variation, Fianchetto Variation       | 1. e4 c5 2. Nc3 Nc6 3. g3                      |   100 |      23 |      25 |    52 |
| 21 | A49 | Indian Game                             | Pseudo-King's Indian Variation               | 1. d4 Nf6 2. Nf3 g6                            |   100 |      28 |      19 |    53 |
| 22 | D85 | Gruenfeld Defense                       | Exchange Variation                           | 1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. cxd5 Nxd5      |   100 |      34 |      14 |    52 |
| 23 | D10 | Slav Defense                            | Exchange Variation                           | 1. d4 d5 2. c4 c6 3. cxd5                      |   100 |      31 |      18 |    51 |
| 24 | A46 | Indian Game                             | London System                                | 1. d4 Nf6 2. Nf3 e6 3. Bf4                     |   100 |      26 |      20 |    54 |
| 25 | C60 | Spanish Game                            | General                                      | 1. e4 e5 2. Nf3 Nc6 3. Bb5                     |   100 |      34 |      24 |    42 |
| 26 | D12 | Slav Defense                            | Quiet Variation, Schallopp Defense           | 1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. e3 Bf5         |   100 |      28 |      19 |    53 |
| 27 | E11 | Bogo-Indian Defense                     | GrA?nfeld Variation                          | 1. d4 Nf6 2. c4 e6 3. Nf3 Bb4+ 4. Nbd2         |   200 |      69 |      43 |    88 |
| 28 | D37 | Queen's Gambit Declined                 | Three Knights Variation, General             | 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Nf3            |   200 |      52 |      19 |   129 |
| 29 | A49 | Indian Game                             | Przepiorka Variation                         | 1. d4 Nf6 2. Nf3 g6 3. g3                      |   199 |      55 |      46 |    98 |
| 30 | D11 | Slav Defense                            | Quiet Variation                              | 1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. e3             |   200 |      59 |      37 |   104 |
| 31 | A46 | Indian Game                             | Yusupov-Rubinstein System                    | 1. d4 Nf6 2. Nf3 e6 3. e3                      |   200 |      57 |      48 |    95 |
| 32 | C53 | Italian Game, Classical Variation       | Giuoco Pianissimo                            | 1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. c3 Nf6 5. d3 |   100 |      40 |      24 |    36 |
| 33 | E04 | Catalan Opening                         | Open Defense                                 | 1. d4 Nf6 2. c4 e6 3. g3 d5 4. Bg2 dxc4 5. Nf3 |   100 |      33 |      22 |    45 |
| 34 | C77 | Spanish Game, Morphy Defense            | Anderssen Variation                          | 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. d3 |   100 |      35 |      31 |    34 |
| 35 | E90 | King's Indian Defense, Normal Variation | Rare Defenses                                | 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3  |   100 |      35 |      23 |    42 |
+----+-----+-----------------------------------------+----------------------------------------------+------------------------------------------------+-------+---------+---------+-------+
Top
goni-K26
Posts: 217
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Re: Stockfish Family Gauntlet Series — Bullet

Post by goni-K26 »

Testing Stockfish 17

Stockfish 17 — two "collapse" losses in an otherwise dominant 30-engine gauntlet

Stockfish 17 finished a clear #1 in a 1,500-game gauntlet against 30 opponents (69.7% score, +9.4 live Elo,
next-best entry at only 40%) — completely as expected for a 2880-rated engine against a 2700–2810 field.
But two individual losses stood out enough to dig into the PGN annotations, and both show the same signature.

Setup
Hardware: Intel i5-4570 3.20GHz, 8GB RAM
Time control: 9+0.1 (the "5Pct" tag in the engine names — 5% of a standard 3+2 blitz baseline)
Games: 1,500 total, 50 games per opponent, 30 opponents

The signature: search depth collapse, not an eval or strength problem

In both losses, Stockfish's own reported search depth (the number after the "/" in each move comment)
crashes to 1–6 ply for extended stretches — right at the moments the position falls apart.
The eval stays confidently positive the whole time; it just isn't backed by any real search.

Game 1 — vs PlentyChess 2.0.0 (start Elo 2801)
SF searches a healthy 14–22 ply the entire game and is up nearly +4 by move 41. Then:

Code: Select all

44. Bf5   {+2.78/15}
45. Bd2   {+2.78/6}    <- depth crashes
46. Bxe3  {+3.65/2}    <- rock bottom
47. Rfg1  {+2.43/5}
48. Qg4   {+1.51/6}
49. Rg3   {+0.33/1}    <- 1 ply
50. Rf1   {-1.31/4}
51. Kg2   {-6.93/20}   <- depth recovers, too late
Exactly in that 1–6 ply window, Black's connected e/f-pawns run down the board and deliver 50...e2! 51...Rxh2+! 52...exf1=N+! — an underpromotion fork that wins material outright. SF never got the depth to see it coming.

Game 2 — vs RubiChess 20221203 (start Elo 2709, 171 points below SF)
This one's more extreme. From roughly move 47 onward, across a 112-move game, SF spends huge stretches at depth 1:

Code: Select all

47. Qc2 {+2.23/1}    54. Nf3 {+1.22/1}    76. Re3 {+1.35/2}    86. Rf3 {+0.96/1}
48. Nc7 {+2.07/1}    55. Rh2 {+1.95/1}    77. Kd1 {+1.35/1}    87. Kd1 {+0.96/1}
50. b3  {+1.83/1}    56. Rc3 {+1.61/1}    78. Ke2 {+1.35/1}    88. Nf6 {+0.96/1}
                      57. Re3 {+1.95/1}    79. Kd1 {+1.35/1}    89. Kd2 {+1.28/1}
                      58. Nd2 {+1.93/1}    80. Kc2 {+1.35/1}
...continuing intermittently to move 110.
Dozens of moves at essentially zero search across the entire second half of the game.
The eval sits comfortably around +1 to +2 throughout — SF genuinely was better — but with no search behind it,
it can't convert, and finally collapses in the last few moves (eval goes from +17.85 to -199.95 in three moves).

Likely cause
This looks like a time-management edge case rather than a strength or eval problem.
At 9+0.1, the total clock budget for an entire ~112-move game like Game 2 is only around 9 + (112 x 0.1) ? 20 seconds for that side,
for the whole game — the 0.1s increment barely moves the needle. This hardware (i5-4570) is also noticeably older/slower than other rigs
I've tested on. My read: SF overspent early (depth 17–23 in the middlegame of game 2,
which isn't cheap at this budget) and was left running on fumes for the remaining 60+ moves,
eventually losing the thread completely.

Context
Worth stressing this isn't representative — SF still won the event comfortably at 69.7% with a wide margin over 2nd.
These read like rare time-trouble outliers rather than a systemic weakness, but flagging in case the depth-1 stretches
point to something checkable in time allocation at this specific TC/hardware combination.

Participants

Code: Select all

+----+---------------------------------------+------+
| No | Name                                  |  Elo |
+----+---------------------------------------+------+
|  - | Stockfish 17 x64(AVX2-5Pct)           | 2880 |
+----+---------------------------------------+------+
|  1 | Obsidian 13.0 x64(BMI2-5Pct)          | 2808 |
|  2 | PlentyChess 2.0.0 x64(BMI2-5Pct)      | 2801 |
|  3 | Caissa 1.19 x64(AVX2-5Pct)            | 2800 |
|  4 | Caissa 1.18 x64(AVX2-5Pct)            | 2796 |
|  5 | PlentyChess 2.1.0 x64(BMI2-5Pct)      | 2796 |
|  6 | Caissa 1.19 x64(BMI2-5Pct)            | 2793 |
|  7 | Caissa 1.17 x64(BMI2-5Pct)            | 2790 |
|  8 | RubiChess 20240112 x64(AVX2-5Pct)     | 2787 |
|  9 | Obsidian 12.0 x64(BMI2-5Pct)          | 2786 |
| 10 | Obsidian 12.0 x64(AVX2-5Pct)          | 2784 |
| 11 | Caissa 1.17 x64(AVX2-5Pct)            | 2772 |
| 12 | Alexandria 6.1.0 x64(BMI2-5Pct)       | 2759 |
| 13 | Caissa 1.15 x64(BMI2-5Pct)            | 2757 |
| 14 | RubiChess 20230918 x64(5Pct)          | 2752 |
| 15 | Motor 0.7.0 x64(AVX2-5Pct)            | 2745 |
| 16 | Obsidian 10.0 x64(AVX2-5Pct)          | 2745 |
| 17 | PlentyChess 1.0.0 x64(BMI2-5Pct)      | 2744 |
| 18 | Caissa 1.15 x64(AVX2-5Pct)            | 2744 |
| 19 | Caissa 1.16 x64(5Pct)                 | 2737 |
| 20 | Caissa 1.15 x64(5Pct)                 | 2732 |
| 21 | Clover 6.2 x64(AVX2-5Pct)             | 2732 |
| 22 | Caissa 1.14.1 x64(AVX2-5Pct)          | 2729 |
| 23 | Alexandria 6.0.0 x64(BMI2-5Pct)       | 2726 |
| 24 | Obsidian 9.0 x64(AVX2-5Pct)           | 2725 |
| 25 | Spaghet BologNNese 1.0 x64(BMI2-5Pct) | 2721 |
| 26 | Clarity 7.2.0 x64(BMI2-5Pct)          | 2718 |
| 27 | Titan 1.1 x64(AVX2-5Pct)              | 2718 |
| 28 | RubiChess 20221203 x64(BMI2-5Pct)     | 2715 |
| 29 | Titan 1.0 x64(AVX2-5Pct)              | 2712 |
| 30 | RubiChess 20221203 x64(AVX2-5Pct)     | 2709 |
+----+---------------------------------------+------+
Openings

Code: Select all

+----+-----+----------------------------------------+--------------------------------------------+-----------------------------------------------------------+-------+---------+---------+-------+--------+
|  # | ECO | Opening                                | Variation                                  | Moves                                                     | Count | White W | Black W | Draws | QDraws |
+----+-----+----------------------------------------+--------------------------------------------+-----------------------------------------------------------+-------+---------+---------+-------+--------+
|  1 |  ?  | ?                                      | -                                          | 1. e4                                                     |    60 |      25 |      11 |    24 |      0 |
|  2 | D30 | Queen's Gambit Declined                | General                                    | 1. d4 d5 2. c4 e6                                         |    60 |      18 |       7 |    35 |      0 |
|  3 | D02 | Queen Pawn Game                        | Symmetrical Variation                      | 1. d4 d5 2. Nf3 Nf6                                       |    60 |      20 |      10 |    30 |      0 |
|  4 | B07 | Pirc Defense                           | General                                    | 1. e4 d6 2. d4 Nf6                                        |    60 |      31 |       3 |    26 |      0 |
|  5 | B30 | Sicilian Defense                       | Nyezhmetdinov-Rossolimo Attack             | 1. e4 c5 2. Nf3 Nc6 3. Bb5                                |    60 |      23 |      10 |    27 |      0 |
|  6 | D31 | Semi-Slav Defense                      | Accelerated Move Order                     | 1. d4 d5 2. c4 e6 3. Nc3 c6                               |    60 |      23 |       7 |    30 |      0 |
|  7 | A35 | English Opening, Symmetrical Variation | Four Knights Variation                     | 1. c4 c5 2. Nc3 Nc6 3. Nf3 Nf6                            |    60 |      15 |       5 |    40 |      0 |
|  8 | B01 | Scandinavian Defense                   | Modern Variation                           | 1. e4 d5 2. exd5 Nf6                                      |    60 |      21 |      10 |    29 |      0 |
|  9 | B12 | Caro-Kann Defense, Advance Variation   | Tal Variation                              | 1. e4 c6 2. d4 d5 3. e5 Bf5 4. h4                         |    60 |      23 |      13 |    24 |      0 |
| 10 | D50 | Queen's Gambit Declined                | Modern Variation                           | 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Bg5                       |    60 |      18 |       8 |    34 |      0 |
| 11 | A06 | Nimzowitsch-Larsen Attack              | General                                    | 1. Nf3 d5 2. b3                                           |    60 |      19 |      12 |    29 |      0 |
| 12 | A04 | Zukertort Opening                      | Nimzo-Larsen Variation                     | 1. Nf3 Nf6 2. b3                                          |    60 |      20 |      14 |    26 |      0 |
| 13 | A07 | King's Indian Attack                   | Yugoslav Variation                         | 1. Nf3 d5 2. g3 Nf6 3. Bg2 c6 4. O-O Bg4                  |    60 |      23 |      10 |    27 |      0 |
| 14 | A31 | English Opening, Symmetrical Variation | Anti-Benoni Variation                      | 1. c4 c5 2. Nf3 Nf6 3. d4                                 |    60 |      14 |      12 |    34 |      0 |
| 15 | E71 | King's Indian Defense                  | Makagonov Variation                        | 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. h3              |    60 |      29 |       7 |    24 |      0 |
| 16 | B41 | Sicilian Defense, Kan Variation        | Maroczy Bind, Reti Variation               | 1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 a6 5. c4            |    60 |      24 |       5 |    31 |      0 |
| 17 | A37 | English Opening, Symmetrical Variation | Two Knights Line                           | 1. c4 c5 2. Nc3 Nc6 3. g3 g6 4. Bg2 Bg7 5. Nf3            |    60 |      16 |      11 |    33 |      0 |
| 18 | B90 | Sicilian Defense                       | Najdorf Variation, Adams Attack            | 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. h3 |    60 |      23 |      12 |    25 |      0 |
| 19 | B12 | Caro-Kann Defense                      | Advance Variation, Botvinnik-Carls Defense | 1. e4 c6 2. d4 d5 3. e5 c5                                |    60 |      20 |      10 |    30 |      0 |
| 20 | B18 | Caro-Kann Defense                      | Classical Variation                        | 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5                 |    60 |      18 |      10 |    32 |      0 |
| 21 | B10 | Caro-Kann Defense                      | Two Knights Attack                         | 1. e4 c6 2. Nc3 d5 3. Nf3                                 |    60 |      20 |       8 |    32 |      0 |
| 22 | A46 | Torre Attack                           | Classical Defense                          | 1. d4 Nf6 2. Nf3 e6 3. Bg5                                |    60 |      14 |      14 |    32 |      0 |
| 23 | B01 | Scandinavian Defense                   | Gubinsky-Melts Defense                     | 1. e4 d5 2. exd5 Qxd5 3. Nc3 Qd6                          |    60 |      27 |       6 |    27 |      0 |
| 24 | B46 | Sicilian Defense                       | Paulsen Variation                          | 1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nc6 5. Nc3 a6       |    60 |      27 |      17 |    16 |      0 |
+----+-----+----------------------------------------+--------------------------------------------+-----------------------------------------------------------+-------+---------+---------+-------+--------+
Tournament Standings after 1.500 Games vs 30 Opponents

Code: Select all

[code]
+------+---------------------------------------+--------+-------+----------+-----------+-------+--------+
| Rank | Name                                  | Points | Games | Live Elo | Start Elo |   +/- | Score% |
+------+---------------------------------------+--------+-------+----------+-----------+-------+--------+
|    1 | Stockfish 17 x64(AVX2-5Pct)           | 1045,0 |  1500 |   2889,4 |    2880,0 |  +9,4 |   69,7 |
|    2 | PlentyChess 2.0.0 x64(BMI2-5Pct)      |   20,0 |    50 |   2805,5 |    2801,0 |  +4,5 |   40,0 |
|    3 | Caissa 1.18 x64(AVX2-5Pct)            |   20,0 |    50 |   2803,1 |    2796,0 |  +7,1 |   40,0 |
|    4 | RubiChess 20240112 x64(AVX2-5Pct)     |   19,5 |    50 |   2794,9 |    2787,0 |  +7,9 |   39,0 |
|    5 | Obsidian 13.0 x64(BMI2-5Pct)          |   18,5 |    50 |   2797,4 |    2808,0 | -10,6 |   37,0 |
|    6 | RubiChess 20230918 x64(5Pct)          |   18,5 |    50 |   2769,6 |    2752,0 | +17,6 |   37,0 |
|    7 | Caissa 1.19 x64(AVX2-5Pct)            |   17,5 |    50 |   2786,0 |    2800,0 | -14,0 |   35,0 |
|    8 | Obsidian 12.0 x64(AVX2-5Pct)          |   16,5 |    50 |   2770,6 |    2784,0 | -13,4 |   33,0 |
|    9 | RubiChess 20221203 x64(AVX2-5Pct)     |   16,5 |    50 |   2731,1 |    2709,0 | +22,1 |   33,0 |
|   10 | PlentyChess 2.1.0 x64(BMI2-5Pct)      |   16,0 |    50 |   2772,6 |    2796,0 | -23,4 |   32,0 |
|   11 | Obsidian 12.0 x64(BMI2-5Pct)          |   16,0 |    50 |   2767,7 |    2786,0 | -18,3 |   32,0 |
|   12 | Caissa 1.15 x64(BMI2-5Pct)            |   16,0 |    50 |   2753,2 |    2757,0 |  -3,8 |   32,0 |
|   13 | Caissa 1.19 x64(BMI2-5Pct)            |   15,5 |    50 |   2767,3 |    2793,0 | -25,7 |   31,0 |
|   14 | Caissa 1.15 x64(AVX2-5Pct)            |   15,5 |    50 |   2742,6 |    2744,0 |  -1,4 |   31,0 |
|   15 | Caissa 1.17 x64(BMI2-5Pct)            |   15,0 |    50 |   2762,1 |    2790,0 | -27,9 |   30,0 |
|   16 | Caissa 1.17 x64(AVX2-5Pct)            |   15,0 |    50 |   2753,2 |    2772,0 | -18,8 |   30,0 |
|   17 | Alexandria 6.1.0 x64(BMI2-5Pct)       |   15,0 |    50 |   2746,6 |    2759,0 | -12,4 |   30,0 |
|   18 | PlentyChess 1.0.0 x64(BMI2-5Pct)      |   15,0 |    50 |   2738,8 |    2744,0 |  -5,2 |   30,0 |
|   19 | Clover 6.2 x64(AVX2-5Pct)             |   15,0 |    50 |   2732,4 |    2732,0 |  +0,4 |   30,0 |
|   20 | Obsidian 9.0 x64(AVX2-5Pct)           |   15,0 |    50 |   2728,6 |    2725,0 |  +3,6 |   30,0 |
|   21 | Caissa 1.16 x64(5Pct)                 |   14,5 |    50 |   2731,3 |    2737,0 |  -5,7 |   29,0 |
|   22 | Motor 0.7.0 x64(AVX2-5Pct)            |   13,5 |    50 |   2727,9 |    2745,0 | -17,1 |   27,0 |
|   23 | Alexandria 6.0.0 x64(BMI2-5Pct)       |   13,5 |    50 |   2717,7 |    2726,0 |  -8,3 |   27,0 |
|   24 | Caissa 1.14.1 x64(AVX2-5Pct)          |   12,5 |    50 |   2711,7 |    2729,0 | -17,3 |   25,0 |
|   25 | Spaghet BologNNese 1.0 x64(BMI2-5Pct) |   12,5 |    50 |   2707,3 |    2721,0 | -13,7 |   25,0 |
|   26 | Titan 1.1 x64(AVX2-5Pct)              |   12,5 |    50 |   2705,7 |    2718,0 | -12,3 |   25,0 |
|   27 | Obsidian 10.0 x64(AVX2-5Pct)          |   12,0 |    50 |   2716,5 |    2745,0 | -28,5 |   24,0 |
|   28 | Caissa 1.15 x64(5Pct)                 |   12,0 |    50 |   2709,5 |    2732,0 | -22,5 |   24,0 |
|   29 | Clarity 7.2.0 x64(BMI2-5Pct)          |   12,0 |    50 |   2701,8 |    2718,0 | -16,2 |   24,0 |
|   30 | RubiChess 20221203 x64(BMI2-5Pct)     |   12,0 |    50 |   2700,2 |    2715,0 | -14,8 |   24,0 |
|   31 | Titan 1.0 x64(AVX2-5Pct)              |   12,0 |    50 |   2698,5 |    2712,0 | -13,5 |   24,0 |
+------+---------------------------------------+--------+-------+----------+-----------+-------+--------+
HEAD-TO-HEAD MATRIX

Code: Select all

+----+---------------------------------------+------+------+-------+-------+-------+
| No | Opponent                              | Wins | Loss | Draws | Games | Score |
+----+---------------------------------------+------+------+-------+-------+-------+
|  1 | PlentyChess 2.0.0 x64(BMI2-5Pct)      |   13 |    3 |    34 |    50 | 60,0% |
|  2 | Caissa 1.18 x64(AVX2-5Pct)            |   16 |    6 |    28 |    50 | 60,0% |
|  3 | RubiChess 20240112 x64(AVX2-5Pct)     |   18 |    7 |    25 |    50 | 61,0% |
|  4 | Obsidian 13.0 x64(BMI2-5Pct)          |   17 |    4 |    29 |    50 | 63,0% |
|  5 | RubiChess 20230918 x64(5Pct)          |   15 |    2 |    33 |    50 | 63,0% |
|  6 | Caissa 1.19 x64(AVX2-5Pct)            |   20 |    5 |    25 |    50 | 65,0% |
|  7 | Obsidian 12.0 x64(AVX2-5Pct)          |   20 |    3 |    27 |    50 | 67,0% |
|  8 | RubiChess 20221203 x64(AVX2-5Pct)     |   22 |    5 |    23 |    50 | 67,0% |
|  9 | PlentyChess 2.1.0 x64(BMI2-5Pct)      |   19 |    1 |    30 |    50 | 68,0% |
| 10 | Obsidian 12.0 x64(BMI2-5Pct)          |   21 |    3 |    26 |    50 | 68,0% |
| 11 | Caissa 1.15 x64(BMI2-5Pct)            |   23 |    5 |    22 |    50 | 68,0% |
| 12 | Caissa 1.19 x64(BMI2-5Pct)            |   20 |    1 |    29 |    50 | 69,0% |
| 13 | Caissa 1.15 x64(AVX2-5Pct)            |   25 |    6 |    19 |    50 | 69,0% |
| 14 | Caissa 1.17 x64(BMI2-5Pct)            |   23 |    3 |    24 |    50 | 70,0% |
| 15 | Caissa 1.17 x64(AVX2-5Pct)            |   23 |    3 |    24 |    50 | 70,0% |
| 16 | Alexandria 6.1.0 x64(BMI2-5Pct)       |   22 |    2 |    26 |    50 | 70,0% |
| 17 | PlentyChess 1.0.0 x64(BMI2-5Pct)      |   22 |    2 |    26 |    50 | 70,0% |
| 18 | Clover 6.2 x64(AVX2-5Pct)             |   26 |    6 |    18 |    50 | 70,0% |
| 19 | Obsidian 9.0 x64(AVX2-5Pct)           |   24 |    4 |    22 |    50 | 70,0% |
| 20 | Caissa 1.16 x64(5Pct)                 |   23 |    2 |    25 |    50 | 71,0% |
| 21 | Motor 0.7.0 x64(AVX2-5Pct)            |   23 |    0 |    27 |    50 | 73,0% |
| 22 | Alexandria 6.0.0 x64(BMI2-5Pct)       |   26 |    3 |    21 |    50 | 73,0% |
| 23 | Caissa 1.14.1 x64(AVX2-5Pct)          |   28 |    3 |    19 |    50 | 75,0% |
| 24 | Spaghet BologNNese 1.0 x64(BMI2-5Pct) |   27 |    2 |    21 |    50 | 75,0% |
| 25 | Titan 1.1 x64(AVX2-5Pct)              |   27 |    2 |    21 |    50 | 75,0% |
| 26 | Obsidian 10.0 x64(AVX2-5Pct)          |   27 |    1 |    22 |    50 | 76,0% |
| 27 | Caissa 1.15 x64(5Pct)                 |   29 |    3 |    18 |    50 | 76,0% |
| 28 | Clarity 7.2.0 x64(BMI2-5Pct)          |   27 |    1 |    22 |    50 | 76,0% |
| 29 | RubiChess 20221203 x64(BMI2-5Pct)     |   28 |    2 |    20 |    50 | 76,0% |
| 30 | Titan 1.0 x64(AVX2-5Pct)              |   28 |    2 |    20 |    50 | 76,0% |
+----+---------------------------------------+------+------+-------+-------+-------+
Note: Scores are reported from Stockfish's perspective against each opponent.
________________________________________________________________________________
Download Pgn Games(Zip)

Tournament Info

Code: Select all

Event:                      Testing - Stockfish 17 x64(AVX2-5Pct)
Site/Hardware:              Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM(HP)
TimeControl:                9s + 0.1s Inc.
Participants:               31

Tourney Start:              07/09/2024 01:43:44 
Tourney End:                08/09/2024 03:26:59 
Duration:                   1d 01:43:15
Avg Game Duration:          00:01:00
Total Games:                1.500

White Wins:                 534(35,6%)
Black Wins:                 240(16,0%)
Draws:                      726(48,4%)

Decisive Games (W/B):       774(51,6%)
White/Black Win Ratio:      222,5%
White Scoring Probability:   59,8%

Longest Decisive Game:      PlentyChess 1.0.0 x64(BMI2-5Pct) vs Stockfish 17 x64(AVX2-5Pct)  0-1  (291 plies)
Shortest Decisive Game:     Stockfish 17 x64(AVX2-5Pct) vs Motor 0.7.0 x64(AVX2-5Pct)  1-0  (44 plies)
Longest Draw:               RubiChess 20240112 x64(AVX2-5Pct) vs Stockfish 17 x64(AVX2-5Pct)  1/2-1/2  (300 plies)
Shortest Draw:              Caissa 1.14.1 x64(AVX2-5Pct) vs Stockfish 17 x64(AVX2-5Pct)  1/2-1/2  (60 plies)

Games with 30-60 moves:     726(48,4%)
Games with 61-90 moves:     626(41,7%)
Games with +90 moves:       139(9,3%)
Moves per Game(Avg.):        63 (127) Plies.
Game 1: Stockfish 17 vs PlentyChess 2.0

[pgn][Event "Testing - Stockfish 17 x64(AVX2-5Pct)"]
[Site "Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM"]
[Date "2024.09.08"]
[Round "25"]
[White "Stockfish 17 x64(AVX2-5Pct)"]
[Black "PlentyChess 2.0.0 x64(BMI2-5Pct)"]
[Result "0-1"]
[BlackElo "2801"]
[ECO "A04"]
[Opening "Zukertort Opening"]
[Time "15:56:26"]
[Variation "Nimzo-Larsen Variation"]
[WhiteElo "2880"]
[TimeControl "9+0"]
[Termination "adjudication"]
[PlyCount "127"]

1. Nf3 Nf6 2. b3 d5 {0.00/19 0} 3. e3 {-0.07/21 0} Bf5 {0.00/18 0} 4. Be2
{-0.03/20 0} e6 {+0.01/18 0} 5. Nh4 {+0.02/17 0} Bg6 {-0.02/18 0} 6. Bb2
{+0.02/17 0} c5 {+0.06/19 0} 7. d3 {+0.01/20 0} Nc6 {+0.03/18 0} 8. O-O
{-0.07/19 0} Rc8 {+0.03/18 0} 9. g3 {+0.11/17 0} Be7 {+0.02/18 0} 10. Bf3
{+0.11/18 0} O-O {-0.03/17 0} 11. c4 {+0.21/17 0} Re8 {-0.05/19 0} 12. Qe2
{+0.24/19 0} d4 {-0.04/18 0} 13. Nxg6 {+0.18/18 0} hxg6 {-0.09/18 0} 14. e4
{+0.18/17 0} Nd7 {-0.05/17 0} 15. Bg2 {+0.19/20 0} g5 {-0.13/18 0} 16. Nd2
{+0.23/18 0} g6 {+0.02/17 0} 17. f4 {+0.22/19 0} gxf4 {+0.14/17 0} 18. gxf4
{+0.21/16 0} e5 {-0.09/17 0} 19. f5 {+0.24/15 0} Bg5 {-0.13/16 0} 20. Bh3
{+0.30/18 0} Kg7 {-0.22/17 0} 21. Nf3 {+0.30/16 0} Rh8 {-0.19/19 0} 22. Qg2
{+0.49/14 0} Be3+ {-0.21/18 0} 23. Kh1 {+0.58/14 0} Qe7 {-0.23/17 0} 24.
f6+ {+0.43/22 0} Nxf6 {+0.12/16 0} 25. Bxc8 {+0.66/14 0} Nb4 {+0.01/17 0}
26. Rad1 {+0.09/21 0} Nh5 {0.00/17 0} 27. Rfe1 {+1.16/22 0} Nf4 {-1.11/15
0} 28. Qf1 {+1.12/14 0} Nh5 {-0.99/16 0} 29. Bh3 {+2.23/15 0} Bf4 {-1.00/14
0} 30. Re2 {+2.39/16 0} Nc6 {-1.88/14 0} 31. Ng1 {+2.81/18 0} Qh4 {-1.66/16
0} 32. Qf2 {+3.00/20 0} Qg5 {-1.97/14 0} 33. Rf1 {+3.10/18 0} f6 {-2.06/16
0} 34. Qf3 {+3.10/17 0} Be3 {-2.38/16 0} 35. Rg2 {+3.26/18 0} Qf4 {-2.14/16
0} 36. Qd1 {+3.48/19 0} Qh6 {-2.80/14 0} 37. Qg4 {+3.42/18 0} Rd8 {-3.09/14
0} 38. Ne2 {+3.44/18 0} Nf4 {-2.86/14 0} 39. Nxf4 {+3.59/16 0} exf4
{-3.18/13 0} 40. Qe6 {+3.63/19 0} Qh5 {-2.63/15 0} 41. Bc1 {+3.98/20 0} Qe5
{-2.27/14 0} 42. Qg4 {+3.82/18 0} g5 {-1.70/14 0} 43. Qe2 {+2.79/21 0} Qd6
{-2.00/14 0} 44. Bf5 {+2.78/15 0} Rh8 {-1.93/13 0} 45. Bd2 {+2.78/6 0} Kf7
{-2.06/16 0} 46. Bxe3 {+3.65/2 0} dxe3 {-0.03/14 0} 47. Rfg1 {+2.43/5 0}
Nd4 {+0.53/14 0} 48. Qg4 {+1.51/6 0} f3 {+2.79/14 0} 49. Rg3 {+0.33/1 0} f2
{+3.18/13 0} 50. Rf1 {-1.31/4 0} e2 {+6.64/15 0} 51. Kg2 {-6.93/20 0} Rxh2+
{+7.36/17 0} 52. Kxh2 {-6.92/16 0} exf1=N+ {+8.22/19 0} 53. Kg2 {-6.90/14
0} Nxg3 {+8.54/18 0} 54. Kxf2 {-6.86/13 0} Ngxf5 {+8.93/18 0} 55. exf5
{-6.83/1 0} Qf4+ {+9.50/18 0} 56. Qxf4 {-6.64/8 0} gxf4 {+9.81/18 0} 57.
Kg2 {-6.43/1 0} Ke7 {+10.26/18 0} 58. Kf2 {-6.29/1 0} Kd6 {+12.45/17 0} 59.
Kg2 {-6.57/2 0} Nxf5 {+14.39/18 0} 60. Kf3 {-6.51/5 0} Ke5 {+14.63/18 0}
61. Kg2 {-8.79/21 0} Nd4 {+15.70/18 0} 62. a3 {-9.05/16 0} Nxb3 {+21.51/20
0} 63. Kh3 {-10.72/21 0} Kd4 {+25.43/19 0} 64. Kg4 {-9.60/15 0 Arena
Adjudication} 0-1[/pgn]

Game 2: Stockfish 17 vs RubiChess 20221203

[pgn][Event "Testing - Stockfish 17 x64(AVX2-5Pct)"]
[Site "Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM"]
[Date "2024.09.08"]
[Round "31"]
[White "Stockfish 17 x64(AVX2-5Pct)"]
[Black "RubiChess 20221203 x64(AVX2-5Pct)"]
[Result "0-1"]
[BlackElo "2709"]
[ECO "E71"]
[Opening "King's Indian Defense"]
[Time "19:31:02"]
[Variation "Makagonov Variation"]
[WhiteElo "2880"]
[TimeControl "9+0"]
[Termination "adjudication"]
[PlyCount "225"]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. h3 O-O {-0.48/19 0} 6. Be3
{+0.64/19 0} e5 {-0.41/20 0} 7. d5 {+0.67/15 0} Na6 {-0.38/20 0} 8. g4
{+0.59/17 0} Nc5 {-0.34/20 0} 9. f3 {+0.68/19 0} Ne8 {-0.34/20 0} 10. h4
{+1.07/17 0} f5 {-0.08/19 0} 11. g5 {+1.21/17 0} a5 {-0.25/20 0} 12. Qd2
{+1.46/19 0} a4 {-0.50/20 0} 13. O-O-O {+1.43/15 0} Qe7 {-0.64/20 0} 14.
Bh3 {+1.60/18 0} b6 {-0.92/19 0} 15. Qc2 {+1.76/18 0} Qf7 {-0.72/19 0} 16.
Kb1 {+1.84/17 0} Bd7 {-1.28/21 0} 17. a3 {+1.86/18 0} Rb8 {-1.27/20 0} 18.
Nge2 {+1.83/18 0} fxe4 {-1.86/20 0} 19. fxe4 {+2.38/17 0} Bxh3 {-1.91/22 0}
20. Rxh3 {+2.35/16 0} Qd7 {-1.93/23 0} 21. Rg3 {+2.53/16 0} Bh8 {-2.12/23
0} 22. Rh1 {+2.08/21 1} Ng7 {-1.48/19 0} 23. Bxc5 {+2.15/16 0} bxc5
{-1.65/20 0} 24. Qxa4 {+2.36/18 0} Qc8 {-1.72/19 0} 25. Rh2 {+2.39/18 0}
Nh5 {-1.82/18 0} 26. Re3 {+2.32/17 0} Rf4 {-1.89/17 0} 27. Ka2 {+2.38/18 0}
Ra8 {-1.74/20 0} 28. Qd1 {+2.35/16 0} Rg4 {-2.04/19 0} 29. Nc1 {+2.33/18 0}
Rf4 {-2.18/21 0} 30. N1e2 {+2.32/17 0} Rg4 {-2.35/22 0} 31. Nb5 {+2.41/17
0} Bg7 {-2.12/21 0} 32. Nec3 {+2.37/17 0} Rf4 {-2.38/20 0} 33. Ne2
{+2.32/14 0} Rg4 {-2.32/22 0} 34. Qh1 {+2.43/16 0} Qd7 {-2.42/20 0} 35. Rb3
{+2.39/18 0} Rf8 {-2.13/19 0} 36. a4 {+2.39/12 0} Ra8 {-2.88/20 0} 37. Ra3
{+2.47/12 0} Ra5 {-2.97/21 0} 38. Qf3 {+2.58/16 0} Nf4 {-2.80/20 0} 39. Qc3
{+2.43/18 0} Ra6 {-3.01/17 0} 40. Qf3 {+2.45/10 0} Ra5 {-3.24/19 0} 41. Ng3
{+2.36/18 0} Bf8 {-2.98/20 0} 42. Qd1 {+2.31/17 0} Be7 {-2.37/17 0} 43. Nf1
{+2.34/14 0} Nh5 {-3.71/19 0} 44. Nxc7 {+2.48/11 0} Rg1 {-3.54/18 0} 45.
Ne6 {+2.16/15 0} Nf4 {-3.38/18 0} 46. Rf2 {+2.23/7 0} Qb7 {-3.55/19 0} 47.
Qc2 {+2.23/1 0} Qb4 {-3.40/18 0} 48. Nc7 {+2.07/1 0} Qe1 {-1.19/18 0} 49.
Raf3 {+3.04/7 0} Rh1 {-1.32/19 0} 50. b3 {+1.83/1 0} Rxh4 {-0.81/16 0} 51.
Re3 {+1.05/16 0} Qb4 {0.00/17 0} 52. Ne6 {+1.06/16 0} Ra8 {-0.65/18 0} 53.
Nd2 {+1.10/11 0} Rh1 {-1.21/19 0} 54. Nf3 {+1.22/1 0} Qb8 {-1.59/19 0} 55.
Rh2 {+1.95/1 0} Rf1 {-1.12/20 0} 56. Rc3 {+1.61/1 0} Bf8 {-0.96/18 0} 57.
Re3 {+1.95/1 0} Qc8 {-1.07/17 0} 58. Nd2 {+1.93/1 0} Rg1 {-1.29/16 0} 59.
Nf3 {+1.91/3 0} Rg4 {-0.51/18 0} 60. Nxf4 {+0.83/1 0} exf4 {+0.13/14 0} 61.
Rd3 {+0.13/17 0} Bg7 {-1.42/17 0} 62. Qe2 {+0.26/15 0} Rg3 {-0.55/18 0} 63.
Rh4 {+0.36/14 0} Qd7 {-1.02/20 0} 64. Rxf4 {+1.07/13 0} Qh3 {-1.07/18 0}
65. Rh4 {+1.09/14 0} Rg2 {-1.05/18 0} 66. Rxh3 {+0.83/17 0} Rxe2+ {-0.38/18
0} 67. Kb1 {+0.87/14 0} Rf8 {-0.85/19 0} 68. Kc1 {+0.93/14 0} Rxe4
{-0.74/15 0} 69. Kd1 {+0.90/12 0} Ba1 {-1.05/14 0} 70. Rg3 {+1.45/11 0} Re7
{-1.69/14 0} 71. Rg1 {+1.35/19 0} Ref7 {-2.47/19 0} 72. Nd2 {+1.58/14 0}
Be5 {-2.32/19 0} 73. Ne4 {+1.69/12 0} Rf1+ {-2.02/19 0} 74. Rxf1 {+1.40/13
0} Rxf1+ {-2.29/18 0} 75. Ke2 {+1.35/10 0} Rb1 {-2.56/17 0} 76. Re3
{+1.35/2 0} Rb2+ {-2.19/16 0} 77. Kd1 {+1.35/1 0} Rb1+ {-2.58/17 0} 78. Ke2
{+1.35/1 0} Rb2+ {-2.10/17 0} 79. Kd1 {+1.35/1 0} Rb1+ {-1.54/17 0} 80. Kc2
{+1.35/1 0} Rb2+ {-2.15/17 0} 81. Kc1 {+1.25/17 0} Rg2 {-2.34/18 0} 82. Kd1
{+1.29/18 0} Rb2 {-2.24/19 0} 83. Ke1 {+0.96/20 0} Rb1+ {-2.27/21 0} 84.
Kd2 {+0.96/12 0} Rb2+ {-2.16/21 0} 85. Kc1 {+0.96/13 0} Rg2 {-2.10/20 0}
86. Rf3 {+0.96/1 0} Kg7 {-2.13/16 0} 87. Kd1 {+0.96/1 0} Rb2 {-1.92/16 0}
88. Nf6 {+0.96/1 0} Rb1+ {-2.27/17 0} 89. Kd2 {+1.28/1 0} Rb2+ {-1.24/17 0}
90. Kc1 {+0.73/3 0} Rg2 {-2.26/19 0} 91. Nd7 {+0.87/17 0} Rxg5 {-1.15/19 0}
92. a5 {+1.02/13 0} h5 {-1.73/20 0} 93. a6 {+1.55/14 0} Rg1+ {-1.35/20 0}
94. Kc2 {+1.62/15 0} Bd4 {-0.94/20 0} 95. b4 {+0.73/18 0} cxb4 {0.00/15 0}
96. Nf8 {+0.40/13 0} Be5 {-0.90/17 0} 97. Ne6+ {+0.40/1 0} Kh7 {-0.95/18 0}
98. Nf8+ {+0.38/3 0} Kh6 {+0.72/17 0} 99. a7 {+0.07/1 0} Ra1 {+4.61/16 0}
100. Rf7 {-0.04/6 0} Ra3 {+4.55/16 0} 101. Rh7+ {-2.33/20 0} Kg5 {+4.55/1
0} 102. Ne6+ {-2.11/15 0} Kg4 {+4.83/18 0} 103. Rb7 {-2.46/14 0} b3+
{+4.95/18 0} 104. Kd2 {-2.54/1 0} h4 {+6.02/18 0} 105. c5 {-2.27/1 0} dxc5
{+6.43/16 0} 106. Nxc5 {-2.14/2 0} h3 {+6.58/16 0} 107. Rb4+ {-2.13/1 0}
Bf4+ {+6.59/17 0} 108. Rxf4+ {-1.84/1 0} Kxf4 {+13.79/15 0} 109. d6
{-3.97/1 0} Rxa7 {+15.11/16 0} 110. d7 {-1.97/1 0} Rxd7+ {+17.85/14 0} 111.
Kc3 {-13.09/18 0} b2 {+31.75/17 0} 112. Kc2 {-199.91/26 0} Rc7 {+298.96/25
0} 113. Kb1 {-199.95/14 0 Arena Adjudication} 0-1[/pgn]

Tournament Setup (Stockfish 13)
________________________________________________________________________________
Top
goni-K26
Posts: 217
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Re: Stockfish Family Gauntlet Series — Bullet

Post by goni-K26 »

Testing Stockfish 17.1

Stockfish 17.1 — same time-trouble signature, and one game where it plays itself into a mate it never saw coming

Following up on the 17.0 gauntlet report — same 9s+0.1s ("5Pct") control, same aging i5-4570 — with a fresh 2,500-game field against 25 opponents. Overall SF17.1 scores a combined 65.4% (891 wins / 119 losses / 1,490 draws across 2,500 games), comfortably dominant as expected. But five losses in this batch show the same search-depth collapse pattern I flagged in the 17.0 report, and one in particular is the sharpest example yet.

The standout: vs Motor 0.7.0
Stockfish is confidently ahead most of the game (+0.9 to +1.7). Then, in the final three moves:

Code: Select all

47. Kg3   {+1.99/1}
48. Kh2   {+1.56/1}
49. Qxf4  {+1.64/1}
    Qg1#   <- checkmated on the board
Depth 1 for all three moves, eval still saying "comfortably ahead" (+1.64) on the very last one — and then mate, delivered on the board, not an adjudication. This is the shortest collapse window I've seen from this pattern (3 plies) and the cleanest possible illustration of the cost: it's not "slightly worse," it's total blindness to an immediate mate while still believing it's winning.

Supporting cases from the same batch
Four other losses in this run show the same shape, varying in how extreme:

vs Berserk 13.0 — an unusual variant: moves 36–39 report the identical eval, +4.75, four times in a row, all at depth 1, before violently flipping to -2.22 the instant real search resumes:

Code: Select all

36. Ng4 {+4.75/1}   37. Nf2 {+4.75/1}   38. Ke1 {+4.75/1}   39. Kf1 {+4.75/1}
40. Rd8 {-2.22/1}   <- reverses hard, still depth 1
Four moves returning an unchanged score at depth 1 looks less like "shallow search" and more like the engine wasn't meaningfully re-searching the position at all for a stretch.

vs RubiChess 20230918 — long depth-1 runs from roughly move 47 on, same shape as the 17.0 report; ends in adjudication after the position quietly falls apart.

vs RubiChess 20240817 — a loss, but a slower grind rather than a sharp cliff; no clean depth-1 signature here, worth noting as a case that doesn't fit the pattern.

vs PlentyChess 4.0.1 — different mechanism worth flagging separately: the eval collapse at move 26 (Rxb6+, -2.25) happens at depth 24 — a genuinely deep search, not a shallow one. This looks like a real evaluation miscalculation on a forced trade, not a clock problem, so I'd treat it as a separate issue from the time-trouble cases above rather than folding it into the same bucket.

Takeaway
The time-trouble signature from the 17.0 report reproduces cleanly in 17.1 on the same hardware/TC, and the Motor game is about as unambiguous a demonstration as you'll get — an engine mated while its own last reported eval says it's ahead. Happy to share full PGNs for any of these on request.
[/code]

Participants

Code: Select all

+----+--------------------------------------+------+
| No | Name                                 |  Elo |
+----+--------------------------------------+------+
|  - | Stockfish 17.1 x64(AVX2-5Pct)        | 2885 |
+----+--------------------------------------+------+
|  1 | Alexandria 8.0.0 x64(BMI2-5Pct)      | 2847 |
|  2 | Obsidian 14.0 x64(BMI2-5Pct)         | 2831 |
|  3 | Caissa 1.21 x64(BMI2-5Pct)           | 2827 |
|  4 | Obsidian 15.0 x64(BMI2-5Pct)         | 2820 |
|  5 | Berserk 13.0 x64(BMI2-5Pct)          | 2818 |
|  6 | RubiChess 20240817 x64 NN(BMI2-5Pct) | 2817 |
|  7 | Alexandria 7.1.0 x64(BMI2-5Pct)      | 2816 |
|  8 | PlentyChess 5.0.0 x64(BMI2-5Pct)     | 2811 |
|  9 | Caissa 1.18 x64(AVX2-5Pct)           | 2806 |
| 10 | PlentyChess 4.0.1 x64(BMI2-5Pct)     | 2805 |
| 11 | Caissa 1.19 x64(BMI2-5Pct)           | 2801 |
| 12 | PlentyChess 2.0.0 x64(BMI2-5Pct)     | 2796 |
| 13 | RubiChess 20240112 x64 NN(AVX2-5Pct) | 2796 |
| 14 | Starzix 6.0 x64(AVX2-5Pct)           | 2793 |
| 15 | Obsidian 12.0 x64(BMI2-5Pct)         | 2785 |
| 16 | Motor 0.8.0 x64(AVX2-5Pct)           | 2782 |
| 17 | RubiChess 20230918 x64 NN(AVX2-5Pct) | 2765 |
| 18 | Horsie 1.0 x64(AVX2-5Pct)            | 2763 |
| 19 | Integral 7.0 x64(BMI2-5Pct)          | 2759 |
| 20 | Motor 0.7.0 x64(AVX2-5Pct)           | 2759 |
| 21 | Integral 5.0 x64(BMI2-5Pct)          | 2757 |
| 22 | Clover 8.1 x64(AVX2-5Pct)            | 2757 |
| 23 | Velvet 8.1.0 x64(AVX2-5Pct)          | 2756 |
| 24 | Viridithas 16.0.0 x64(AVX2-5Pct)     | 2742 |
| 25 | Quanticade 1.0f x64(BMI2-5Pct)       | 2700 |
+----+--------------------------------------+------+
Tournament Standings after 2.500 Games vs 25 Teams

Code: Select all

+------+--------------------------------------+--------+-------+----------+-----------+-------+--------+
| Rank | Name                                 | Points | Games | Live Elo | Start Elo |   +/- | Score% |
+------+--------------------------------------+--------+-------+----------+-----------+-------+--------+
|    1 | Stockfish 17.1 x64(AVX2-5Pct)        | 1636,0 |  2500 |   2901,0 |    2885,0 | +16,0 |   65,4 |
|    2 | Obsidian 14.0 x64(BMI2-5Pct)         |   43,5 |   100 |   2840,7 |    2831,0 |  +9,7 |   43,5 |
|    3 | Alexandria 8.0.0 x64(BMI2-5Pct)      |   41,5 |   100 |   2822,6 |    2847,0 | -24,4 |   41,5 |
|    4 | PlentyChess 5.0.0 x64(BMI2-5Pct)     |   41,0 |   100 |   2822,9 |    2811,0 | +11,9 |   41,0 |
|    5 | Caissa 1.21 x64(BMI2-5Pct)           |   40,0 |   100 |   2813,2 |    2827,0 | -13,8 |   40,0 |
|    6 | Obsidian 15.0 x64(BMI2-5Pct)         |   39,5 |   100 |   2810,0 |    2820,0 | -10,0 |   39,5 |
|    7 | Berserk 13.0 x64(BMI2-5Pct)          |   39,0 |   100 |   2806,2 |    2818,0 | -11,8 |   39,0 |
|    8 | RubiChess 20240817 x64 NN(BMI2-5Pct) |   38,5 |   100 |   2802,3 |    2817,0 | -14,7 |   38,5 |
|    9 | Caissa 1.19 x64(BMI2-5Pct)           |   37,0 |   100 |   2791,9 |    2801,0 |  -9,1 |   37,0 |
|   10 | Starzix 6.0 x64(AVX2-5Pct)           |   36,5 |   100 |   2788,5 |    2793,0 |  -4,5 |   36,5 |
|   11 | Alexandria 7.1.0 x64(BMI2-5Pct)      |   36,5 |   100 |   2786,4 |    2816,0 | -29,6 |   36,5 |
|   12 | Obsidian 12.0 x64(BMI2-5Pct)         |   36,0 |   100 |   2785,1 |    2785,0 |  +0,1 |   36,0 |
|   13 | Horsie 1.0 x64(AVX2-5Pct)            |   35,5 |   100 |   2782,0 |    2763,0 | +19,0 |   35,5 |
|   14 | Caissa 1.18 x64(AVX2-5Pct)           |   34,5 |   100 |   2771,4 |    2806,0 | -34,6 |   34,5 |
|   15 | Integral 7.0 x64(BMI2-5Pct)          |   33,5 |   100 |   2766,0 |    2759,0 |  +7,0 |   33,5 |
|   16 | PlentyChess 4.0.1 x64(BMI2-5Pct)     |   33,5 |   100 |   2763,5 |    2805,0 | -41,5 |   33,5 |
|   17 | Motor 0.8.0 x64(AVX2-5Pct)           |   33,0 |   100 |   2761,2 |    2782,0 | -20,8 |   33,0 |
|   18 | RubiChess 20230918 x64 NN(AVX2-5Pct) |   32,5 |   100 |   2757,9 |    2765,0 |  -7,1 |   32,5 |
|   19 | RubiChess 20240112 x64 NN(AVX2-5Pct) |   32,5 |   100 |   2756,3 |    2796,0 | -39,7 |   32,5 |
|   20 | PlentyChess 2.0.0 x64(BMI2-5Pct)     |   31,0 |   100 |   2744,3 |    2796,0 | -51,7 |   31,0 |
|   21 | Quanticade 1.0f x64(BMI2-5Pct)       |   31,0 |   100 |   2742,9 |    2700,0 | +42,9 |   31,0 |
|   22 | Clover 8.1 x64(AVX2-5Pct)            |   30,5 |   100 |   2742,0 |    2757,0 | -15,0 |   30,5 |
|   23 | Motor 0.7.0 x64(AVX2-5Pct)           |   30,0 |   100 |   2738,0 |    2759,0 | -21,0 |   30,0 |
|   24 | Viridithas 16.0.0 x64(AVX2-5Pct)     |   27,5 |   100 |   2717,9 |    2742,0 | -24,1 |   27,5 |
|   25 | Velvet 8.1.0 x64(AVX2-5Pct)          |   25,0 |   100 |   2698,0 |    2756,0 | -58,0 |   25,0 |
|   26 | Integral 5.0 x64(BMI2-5Pct)          |   25,0 |   100 |   2698,0 |    2757,0 | -59,0 |   25,0 |
+------+--------------------------------------+--------+-------+----------+-----------+-------+--------+
HEAD-TO-HEAD MATRIX

Code: Select all

+----+--------------------------------------+------+------+-------+-------+-------+
| No | Opponent                             | Wins | Loss | Draws | Games | Score |
+----+--------------------------------------+------+------+-------+-------+-------+
|  1 | Obsidian 14.0 x64(BMI2-5Pct)         |   20 |    7 |    73 |   100 | 56,5% |
|  2 | Alexandria 8.0.0 x64(BMI2-5Pct)      |   22 |    5 |    73 |   100 | 58,5% |
|  3 | PlentyChess 5.0.0 x64(BMI2-5Pct)     |   24 |    6 |    70 |   100 | 59,0% |
|  4 | Caissa 1.21 x64(BMI2-5Pct)           |   33 |   13 |    54 |   100 | 60,0% |
|  5 | Obsidian 15.0 x64(BMI2-5Pct)         |   25 |    4 |    71 |   100 | 60,5% |
|  6 | Berserk 13.0 x64(BMI2-5Pct)          |   28 |    6 |    66 |   100 | 61,0% |
|  7 | RubiChess 20240817 x64 NN(BMI2-5Pct) |   30 |    7 |    63 |   100 | 61,5% |
|  8 | Caissa 1.19 x64(BMI2-5Pct)           |   32 |    6 |    62 |   100 | 63,0% |
|  9 | Starzix 6.0 x64(AVX2-5Pct)           |   33 |    6 |    61 |   100 | 63,5% |
| 10 | Alexandria 7.1.0 x64(BMI2-5Pct)      |   29 |    2 |    69 |   100 | 63,5% |
| 11 | Obsidian 12.0 x64(BMI2-5Pct)         |   31 |    3 |    66 |   100 | 64,0% |
| 12 | Horsie 1.0 x64(AVX2-5Pct)            |   33 |    4 |    63 |   100 | 64,5% |
| 13 | Caissa 1.18 x64(AVX2-5Pct)           |   37 |    6 |    57 |   100 | 65,5% |
| 14 | Integral 7.0 x64(BMI2-5Pct)          |   38 |    5 |    57 |   100 | 66,5% |
| 15 | PlentyChess 4.0.1 x64(BMI2-5Pct)     |   37 |    4 |    59 |   100 | 66,5% |
| 16 | Motor 0.8.0 x64(AVX2-5Pct)           |   37 |    3 |    60 |   100 | 67,0% |
| 17 | RubiChess 20230918 x64 NN(AVX2-5Pct) |   41 |    6 |    53 |   100 | 67,5% |
| 18 | RubiChess 20240112 x64 NN(AVX2-5Pct) |   38 |    3 |    59 |   100 | 67,5% |
| 19 | PlentyChess 2.0.0 x64(BMI2-5Pct)     |   40 |    2 |    58 |   100 | 69,0% |
| 20 | Quanticade 1.0f x64(BMI2-5Pct)       |   42 |    4 |    54 |   100 | 69,0% |
| 21 | Clover 8.1 x64(AVX2-5Pct)            |   44 |    5 |    51 |   100 | 69,5% |
| 22 | Motor 0.7.0 x64(AVX2-5Pct)           |   43 |    3 |    54 |   100 | 70,0% |
| 23 | Viridithas 16.0.0 x64(AVX2-5Pct)     |   48 |    3 |    49 |   100 | 72,5% |
| 24 | Velvet 8.1.0 x64(AVX2-5Pct)          |   53 |    3 |    44 |   100 | 75,0% |
| 25 | Integral 5.0 x64(BMI2-5Pct)          |   53 |    3 |    44 |   100 | 75,0% |
+----+--------------------------------------+------+------+-------+-------+-------+
Note: Scores are reported from Stockfish's perspective against each opponent.
________________________________________________________________________________
Download Pgn Games(Zip)

Tournament Info

Code: Select all

Event:                      Stockfish 17.1 x64(AVX2-5Pct)
Site/Hardware:              Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM
TimeControl:                9s + 0.1s Inc.
Participants:               26
Total Games:                2.500

Tourney Start:              30/03/2025 19:10:27 
Tourney End:                01/04/2025 10:03:46 
Tourney Duration:           1d 14:53:19
Avg Game Duration:          00:00:55

White Wins:                 697(27,9%)
Black Wins:                 313(12,5%)
Draws:                      1490(59,6%)
Decisive Games (W/B):       1010(40,4%)
White/Black Win Ratio:       222,7%
White Scoring Probability:    57,7%

Longest Decisive Game:      Caissa 1.18 x64(AVX2-5Pct) vs Stockfish 17.1 x64(AVX2-5Pct)  0-1  (276 plies)
Shortest Decisive Game:     Stockfish 17.1 x64(AVX2-5Pct) vs Motor 0.8.0 x64(AVX2-5Pct)  1-0  (52 plies)
Longest Draw:               RubiChess 20240112 x64 NN(AVX2-5Pct) vs Stockfish 17.1 x64(AVX2-5Pct)  1/2-1/2  (300 plies)
Shortest Draw:              Stockfish 17.1 x64(AVX2-5Pct) vs Caissa 1.19 x64(BMI2-5Pct)  1/2-1/2  (60 plies)

Games with 30-60 moves:     1.284(51,4%)
Games with 61-90 moves:       980(39,2%)
Games with +90 moves:         229(9,2%)
Avg. Moves per Game:           63 (126) Plies.
Round 3: Stockfish 17 vs Motor 0.7
[pgn]
[Event "Stockfish 17.1 x64(AVX2-5Pct)"]
[Site "Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM"]
[Date "2025.03.30"]
[Round "3"]
[White "Stockfish 17.1 x64(AVX2-5Pct)"]
[Black "Motor 0.7.0 x64(AVX2-5Pct)"]
[Result "0-1"]
[BlackElo "2759"]
[ECO "D07"]
[Opening "Queen's Gambit Refused, Chigorin Defense"]
[Time "20:20:58"]
[Variation "Exchange Variation, Costa's Line"]
[WhiteElo "2885"]
[TimeControl "9s+0.1s"]
[Termination "normal"]
[PlyCount "98"]

1. d4 d5 2. c4 Nc6 3. cxd5 Qxd5 4. e3 e5 5. Nc3 Bb4 6. Bd2 Bxc3 7. Bxc3
exd4 8. Ne2 Nf6 {-1.62/16 0} 9. Nxd4 {+0.23/15 0} O-O {-1.21/16 0} 10. f3
{+0.35/18 0} Re8 {-1.20/14 0} 11. Kf2 {+0.33/14 0} Qc5 {-0.83/14 0} 12. Qc1
{+0.49/19 0} Ne5 {-0.82/14 0} 13. Be2 {+0.43/16 0} Nd5 {-1.28/15 0} 14. Rd1
{+0.45/19 0} Nxc3 {-1.16/15 0} 15. Qxc3 {+0.48/15 0} Qe7 {-1.29/14 0} 16.
e4 {+0.47/16 0} h5 {-1.24/15 0} 17. Rac1 {+0.50/18 0} c6 {-0.68/13 0} 18.
Rd2 {+0.51/16 0} a5 {-0.85/13 0} 19. a4 {+0.51/19 0} h4 {-1.09/16 0} 20. f4
{+0.90/17 0} Ng6 {-1.37/15 0} 21. f5 {+0.92/14 0} Ne5 {-1.10/16 0} 22. h3
{+0.77/22 0} Qg5 {-1.75/16 0} 23. Qe3 {+0.73/15 0} Qf6 {-1.45/15 0} 24.
Rcd1 {+0.81/20 0} b6 {-1.27/16 0} 25. b3 {+0.72/15 0} Nd7 {-0.49/15 0} 26.
Bd3 {+0.57/17 0} Ne5 {-2.16/15 0} 27. Bc2 {+0.70/15 0} Ra7 {-2.03/17 0} 28.
Nf3 {+0.79/17 0} Nxf3 {-2.01/16 0} 29. gxf3 {+0.85/22 0} Rae7 {-2.60/16 0}
30. Qf4 {+0.74/15 0} Qc3 {-2.26/16 0} 31. Qxh4 {+1.48/13 0} Rd7 {-3.23/15
0} 32. Qg5 {+1.39/15 0} Qc5+ {-3.04/16 0} 33. Kg3 {+1.42/15 0} f6 {-3.52/17
0} 34. Qf4 {+1.41/13 0} Rxd2 {-3.45/15 0} 35. Qxd2 {+1.57/14 0} b5
{-3.74/15 0} 36. Bb1 {+1.51/20 0} Qe5+ {-3.64/17 0} 37. f4 {+1.59/19 0} Qe7
{-3.52/15 0} 38. Qd6 {+1.61/15 0} Qa7 {-3.39/15 0} 39. Qd4 {+1.65/13 0} Qe7
{-3.18/15 0} 40. Qd6 {+1.61/18 0} Qa7 {0.00/28 0} 41. Qd4 {+1.67/13 0} Qe7
{0.00/29 0} 42. Bc2 {+1.60/18 0} Kh8 {-2.99/16 0} 43. Kh4 {+1.41/19 0} Rg8
{-3.07/15 0} 44. Qd6 {+1.56/12 0} Qa7 {-2.66/13 0} 45. Rd4 {+1.77/10 0} Qf7
{-2.45/15 0} 46. Rd2 {+1.17/9 0} g5+ {-1.18/13 0} 47. Kg3 {+1.99/1 0} Qg7
{0.00/16 0} 48. Kh2 {+1.56/1 0} gxf4 {+1.95/12 0} 49. Qxf4 {+1.64/1 0} Qg1#
{+199.99/21 0} 0-1[/pgn]

Tournament Setup (Stockfish 13)
Stockfish 17
________________________________________________________________________________

Gauntlet test report — detailed anomaly review, v17.1
________________________________________________________________________________
Top
goni-K26
Posts: 217
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Re: Stockfish Family Gauntlet Series — Bullet

Post by goni-K26 »

Testing Stockfish 18(dev)

Stockfish 18 (dev, 2026-08-02) — dominant field result, and a third consecutive version showing the same time-trouble signature

Continuing the series from the 17.0 and 17.1 reports — same 9s+0.1s ("5Pct") control, same i5-4570 rig — this time against the top 30 field, 1,260 games. SF18d3 is thoroughly dominant: 70.8% score, Live Elo 2900.4 (+17.4 over its 2883 start), a clear +38.8 Elo margin over 2nd place (Reckless, 2861.6). No surprise there — but one loss in this batch, against an opponent rated 132 points below it, shows the clearest version yet of the pattern flagged in the previous two reports.

vs Halogen 16.0 (start 2751, 132 points below SF) — a gradual ramp-down, not a sudden cliff

Unlike the earlier examples where search depth cliffs suddenly, this one shows a steady decline across four consecutive moves, bottoming out exactly at the fatal move:

Code: Select all

45. Qd2  Rd8  {+3.13/7}
46. d7   Qh3+ {+3.13/4}
47. Ke2  Rf6  {+4.00/6}
48. Qc3  a5   {+6.41/1}   <- depth 1, Black believes it's winning big
49. Rd3  Qe6  {-0.73/3}   <- reversal begins
50. Qc7  Kh7  {-3.40/3}
51. Qxd8 Qxa2+{-6.55/17}  <- depth recovers, position already lost
Black's own eval climbs steadily through the middlegame (+2.83 at move 43, +3.13 at move 45) and then spikes to +6.41 at move 48 — on a single ply of search. The very next move, with White's Rd3, the position reverses instantly and never recovers; the game ends in a forced mate a dozen moves later. The eval swing from "+6.41, winning" to "-0.73, in trouble" happens in exactly one ply, and it's a direct consequence of the depth collapsing to 1 at precisely the wrong moment.

Why this one's worth flagging separately
The previous two reports (17.0, 17.1) mostly showed sudden depth cliffs — healthy search one move, depth 1 the next. This game shows the ramp itself: 7 → 4 → 6 → 1 across four straight moves, which looks like a clock genuinely running down rather than a one-off search anomaly. Combined with the earlier reports, this is now the third separate version/build (17.0, 17.1, 18-dev) reproducing essentially the same failure mode on the same hardware/TC pairing — which points toward something structural in how time gets allocated at this specific control rather than a per-version regression.

Context
Worth repeating: this is one loss out of 1,260 games in a field SF dominates comprehensively. Flagging as a reproducible, version-independent edge case, not a strength concern.
________________________________________________________________________________

Participants

Code: Select all

+----+-----------------------------------+------+
| No | Name                              |  Elo |
+----+-----------------------------------+------+
|  - | Stockfish 18d3 x64(AVX2-5Pct)     | 2883 |
+----+-----------------------------------+------+
|  1 | Reckless 0.9.0 x64(AVX2-5Pct)     | 2880 |
|  2 | PlentyChess 8.0.0 x64(BMI2-5Pct)  | 2872 |
|  3 | Alexandria 9.0.0 x64(BMI2-5Pct)   | 2844 |
|  4 | Obsidian 16.0 x64(BMI2-5Pct)      | 2816 |
|  5 | Berserk 14.0 x64(AVX2-5Pct)       | 2810 |
|  6 | Hobbes 3.0 x64(AVX2-5Pct)         | 2810 |
|  7 | Pawnocchio 2.0.1 x64(AVX2-5Pct)   | 2810 |
|  8 | RubiChess 20240817 x64(BMI2-5Pct) | 2806 |
|  9 | Horsie 1.1 x64(AVX2-5Pct)         | 2800 |
| 10 | Astra 7.0.0 x64(BMI2-5Pct)        | 2800 |
| 11 | Komodo Dragon 3.3 x64(AVX2-5Pct)  | 2794 |
| 12 | Caissa 1.25 x64(BMI2-5Pct)        | 2793 |
| 13 | Viridithas 20.0.0 x64(AVX2-5Pct)  | 2793 |
| 14 | Tarnished 6.0 x64(AVX2-5Pct)      | 2780 |
| 15 | Raphael 4.2.0 x64(BMI2-5Pct)      | 2778 |
| 16 | Starzix 6.0 x64(AVX2-5Pct)        | 2775 |
| 17 | Stormphrax 8.0.0 x64(BMI2-5Pct)   | 2760 |
| 18 | Integral 7.0 x64(BMI2-5Pct)       | 2757 |
| 19 | Halogen 16.0 x64(BMI2-5Pct)       | 2751 |
| 20 | Motor 0.9.0 x64(AVX2-5Pct)        | 2750 |
| 21 | PZChessBot 7.1 x64(AVX2-5Pct)     | 2734 |
| 22 | Renegade 1.3.1 x64(BMI2-5Pct)     | 2730 |
| 23 | Velvet 8.1.1 x64(AVX2-5Pct)       | 2730 |
| 24 | Tcheran 13.0 x64(AVX2-5Pct)       | 2729 |
| 25 | Koivisto 9.0 x64(AVX2-5Pct)       | 2720 |
| 26 | Clarity 8.0.0 x64(BMI2-5Pct)      | 2712 |
| 27 | Clover 9.1 x64(AVX2-5Pct)         | 2700 |
| 28 | Revenge 4.0 x64(BMI2-5Pct)        | 2698 |
| 29 | RofChade 3.1 x64(AVX2-5Pct)       | 2690 |
| 30 | Cinder 0.5.2 x64(AVX2-5Pct)       | 2647 |
+----+-----------------------------------+------+
Tournament Standings after 1.260 Games vs 30 Teams

Code: Select all

+------+-----------------------------------+--------+-------+----------+-----------+-------+--------+
| Rank | Name                              | Points | Games | Live Elo | Start Elo |   +/- | Score% |
+------+-----------------------------------+--------+-------+----------+-----------+-------+--------+
|    1 | Stockfish 18d3 x64(AVX2-5Pct)     |  891,5 |  1260 |   2900,4 |    2883,0 | +17,4 |   70,8 |
|    2 | Reckless 0.9.0 x64(AVX2-5Pct)     |   18,5 |    42 |   2861,6 |    2880,0 | -18,4 |   44,0 |
|    3 | PlentyChess 8.0.0 x64(BMI2-5Pct)  |   17,5 |    42 |   2849,5 |    2872,0 | -22,5 |   41,7 |
|    4 | Pawnocchio 2.0.1 x64(AVX2-5Pct)   |   17,0 |    42 |   2812,8 |    2810,0 |  +2,8 |   40,5 |
|    5 | Berserk 14.0 x64(AVX2-5Pct)       |   16,5 |    42 |   2808,8 |    2810,0 |  -1,2 |   39,3 |
|    6 | Astra 7.0.0 x64(BMI2-5Pct)        |   15,5 |    42 |   2795,4 |    2800,0 |  -4,6 |   36,9 |
|    7 | Alexandria 9.0.0 x64(BMI2-5Pct)   |   15,0 |    42 |   2815,1 |    2844,0 | -28,9 |   35,7 |
|    8 | Obsidian 16.0 x64(BMI2-5Pct)      |   15,0 |    42 |   2800,2 |    2816,0 | -15,8 |   35,7 |
|    9 | Hobbes 3.0 x64(AVX2-5Pct)         |   14,0 |    42 |   2789,0 |    2810,0 | -21,0 |   33,3 |
|   10 | Caissa 1.25 x64(BMI2-5Pct)        |   14,0 |    42 |   2779,7 |    2793,0 | -13,3 |   33,3 |
|   11 | Starzix 6.0 x64(AVX2-5Pct)        |   14,0 |    42 |   2769,7 |    2775,0 |  -5,3 |   33,3 |
|   12 | RubiChess 20240817 x64(BMI2-5Pct) |   13,5 |    42 |   2782,9 |    2806,0 | -23,1 |   32,1 |
|   13 | Viridithas 20.0.0 x64(AVX2-5Pct)  |   13,5 |    42 |   2775,7 |    2793,0 | -17,3 |   32,1 |
|   14 | Raphael 4.2.0 x64(BMI2-5Pct)      |   13,5 |    42 |   2767,4 |    2778,0 | -10,6 |   32,1 |
|   15 | Halogen 16.0 x64(BMI2-5Pct)       |   13,5 |    42 |   2751,9 |    2751,0 |  +0,9 |   32,1 |
|   16 | Motor 0.9.0 x64(AVX2-5Pct)        |   12,5 |    42 |   2743,4 |    2750,0 |  -6,6 |   29,8 |
|   17 | Komodo Dragon 3.3 x64(AVX2-5Pct)  |   12,0 |    42 |   2764,4 |    2794,0 | -29,6 |   28,6 |
|   18 | Clover 9.1 x64(AVX2-5Pct)         |   12,0 |    42 |   2709,0 |    2700,0 |  +9,0 |   28,6 |
|   19 | Horsie 1.1 x64(AVX2-5Pct)         |   11,5 |    42 |   2763,8 |    2800,0 | -36,2 |   27,4 |
|   20 | Tarnished 6.0 x64(AVX2-5Pct)      |   11,5 |    42 |   2752,7 |    2780,0 | -27,3 |   27,4 |
|   21 | Renegade 1.3.1 x64(BMI2-5Pct)     |   11,0 |    42 |   2719,6 |    2730,0 | -10,4 |   26,2 |
|   22 | Stormphrax 8.0.0 x64(BMI2-5Pct)   |   10,5 |    42 |   2733,4 |    2760,0 | -26,6 |   25,0 |
|   23 | RofChade 3.1 x64(AVX2-5Pct)       |   10,5 |    42 |   2690,8 |    2690,0 |  +0,8 |   25,0 |
|   24 | Integral 7.0 x64(BMI2-5Pct)       |   10,0 |    42 |   2727,7 |    2757,0 | -29,3 |   23,8 |
|   25 | Tcheran 13.0 x64(AVX2-5Pct)       |    9,0 |    42 |   2703,2 |    2729,0 | -25,8 |   21,4 |
|   26 | Clarity 8.0.0 x64(BMI2-5Pct)      |    9,0 |    42 |   2692,8 |    2712,0 | -19,2 |   21,4 |
|   27 | Cinder 0.5.2 x64(AVX2-5Pct)       |    8,5 |    42 |   2646,3 |    2647,0 |  -0,7 |   20,2 |
|   28 | Koivisto 9.0 x64(AVX2-5Pct)       |    8,0 |    42 |   2689,8 |    2720,0 | -30,2 |   19,0 |
|   29 | PZChessBot 7.1 x64(AVX2-5Pct)     |    7,5 |    42 |   2694,3 |    2734,0 | -39,7 |   17,9 |
|   30 | Velvet 8.1.1 x64(AVX2-5Pct)       |    7,5 |    42 |   2691,9 |    2730,0 | -38,1 |   17,9 |
|   31 | Revenge 4.0 x64(BMI2-5Pct)        |    6,5 |    42 |   2664,2 |    2698,0 | -33,8 |   15,5 |
+------+-----------------------------------+--------+-------+----------+-----------+-------+--------+
HEAD-TO-HEAD MATRIX

Code: Select all

+----+-----------------------------------+------+------+-------+-------+-------+
| No | Opponent                          | Wins | Loss | Draws | Games | Score |
+----+-----------------------------------+------+------+-------+-------+-------+
|  1 | Reckless 0.9.0 x64(AVX2-5Pct)     |    7 |    2 |    33 |    42 | 56,0% |
|  2 | PlentyChess 8.0.0 x64(BMI2-5Pct)  |    9 |    2 |    31 |    42 | 58,3% |
|  3 | Pawnocchio 2.0.1 x64(AVX2-5Pct)   |   10 |    2 |    30 |    42 | 59,5% |
|  4 | Berserk 14.0 x64(AVX2-5Pct)       |   10 |    1 |    31 |    42 | 60,7% |
|  5 | Astra 7.0.0 x64(BMI2-5Pct)        |   11 |    0 |    31 |    42 | 63,1% |
|  6 | Alexandria 9.0.0 x64(BMI2-5Pct)   |   13 |    1 |    28 |    42 | 64,3% |
|  7 | Obsidian 16.0 x64(BMI2-5Pct)      |   13 |    1 |    28 |    42 | 64,3% |
|  8 | Hobbes 3.0 x64(AVX2-5Pct)         |   14 |    0 |    28 |    42 | 66,7% |
|  9 | Caissa 1.25 x64(BMI2-5Pct)        |   16 |    2 |    24 |    42 | 66,7% |
| 10 | Starzix 6.0 x64(AVX2-5Pct)        |   15 |    1 |    26 |    42 | 66,7% |
| 11 | RubiChess 20240817 x64(BMI2-5Pct) |   17 |    2 |    23 |    42 | 67,9% |
| 12 | Viridithas 20.0.0 x64(AVX2-5Pct)  |   17 |    2 |    23 |    42 | 67,9% |
| 13 | Raphael 4.2.0 x64(BMI2-5Pct)      |   17 |    2 |    23 |    42 | 67,9% |
| 14 | Halogen 16.0 x64(BMI2-5Pct)       |   17 |    2 |    23 |    42 | 67,9% |
| 15 | Motor 0.9.0 x64(AVX2-5Pct)        |   19 |    2 |    21 |    42 | 70,2% |
| 16 | Komodo Dragon 3.3 x64(AVX2-5Pct)  |   18 |    0 |    24 |    42 | 71,4% |
| 17 | Clover 9.1 x64(AVX2-5Pct)         |   18 |    0 |    24 |    42 | 71,4% |
| 18 | Horsie 1.1 x64(AVX2-5Pct)         |   19 |    0 |    23 |    42 | 72,6% |
| 19 | Tarnished 6.0 x64(AVX2-5Pct)      |   21 |    2 |    19 |    42 | 72,6% |
| 20 | Renegade 1.3.1 x64(BMI2-5Pct)     |   21 |    1 |    20 |    42 | 73,8% |
| 21 | Stormphrax 8.0.0 x64(BMI2-5Pct)   |   22 |    1 |    19 |    42 | 75,0% |
| 22 | RofChade 3.1 x64(AVX2-5Pct)       |   22 |    1 |    19 |    42 | 75,0% |
| 23 | Integral 7.0 x64(BMI2-5Pct)       |   23 |    1 |    18 |    42 | 76,2% |
| 24 | Tcheran 13.0 x64(AVX2-5Pct)       |   26 |    2 |    14 |    42 | 78,6% |
| 25 | Clarity 8.0.0 x64(BMI2-5Pct)      |   25 |    1 |    16 |    42 | 78,6% |
| 26 | Cinder 0.5.2 x64(AVX2-5Pct)       |   25 |    0 |    17 |    42 | 79,8% |
| 27 | Koivisto 9.0 x64(AVX2-5Pct)       |   27 |    1 |    14 |    42 | 81,0% |
| 28 | PZChessBot 7.1 x64(AVX2-5Pct)     |   27 |    0 |    15 |    42 | 82,1% |
| 29 | Velvet 8.1.1 x64(AVX2-5Pct)       |   28 |    1 |    13 |    42 | 82,1% |
| 30 | Revenge 4.0 x64(BMI2-5Pct)        |   29 |    0 |    13 |    42 | 84,5% |
+----+-----------------------------------+------+------+-------+-------+-------+
Note: Scores are reported from Stockfish's perspective against each opponent.

Tournament Info

Code: Select all

Event:                      Testing Stockfish 18d3 x64(AVX2-5Pct)
Site/Hardware:              Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM
TimeControl:                9s + 0.1s
Participants:               31

Tourney Start:              03/08/2026 20:35:39 
Tourney End:                04/08/2026 17:03:04 [RUNNING] 
Tourney Duration:           20:27:25
Avg Game Duration:          00:00:57
Total Games:                1.260

White Wins:                 391(31,0%)
Black Wins:                 198(15,7%)
Draws:                      671(53,3%)
Decisive Games (W/B):       589(46,7%)
White/Black Win Ratio:      197,5%
White Scoring Probability:   57,7%

Longest Decisive Game:      Stockfish 18d3 x64(AVX2-5Pct) vs Tcheran 13.0 x64(AVX2-5Pct)  1-0  (290 plies)
Shortest Decisive Game:     Stockfish 18d3 x64(AVX2-5Pct) vs Clover 9.1 x64(AVX2-5Pct)  1-0  (48 plies)
Longest Draw:               Raphael 4.2.0 x64(BMI2-5Pct) vs Stockfish 18d3 x64(AVX2-5Pct)  1/2-1/2  (300 plies)
Shortest Draw:              Stockfish 18d3 x64(AVX2-5Pct) vs Clover 9.1 x64(AVX2-5Pct)  1/2-1/2  (68 plies)

Games with 30-60 moves:     600(47,6%)
Games with 61-90 moves:     522(41,4%)
Games with +90 moves:       134(10,6%)
Moves per Game:              64 (129) Plies.
________________________________________________________________________________

Halogen 16.0 vs. Stockfish 18(2026.08.02):
________________________________
[pgn]
[Event "Testing Stockfish 18(2026.08.02) x64(AVX2-5Pct)"]
[Site "Intel(R) Core(TM) i5-4570 3.20GHz 8GB RAM"]
[Date "2026.08.03"]
[Round "6"]
[White "Halogen 16.0 x64(BMI2-5Pct)"]
[Black "Stockfish 18(2026.08.02) x64(AVX2-5Pct)"]
[Result "1-0"]
[BlackElo "2883"]
[ECO "D15"]
[Opening "Slav Defense"]
[Time "23:10:15"]
[Variation "Chameleon Variation"]
[WhiteElo "2751"]
[TimeControl "9s+0.1s"]
[Termination "adjudication"]
[PlyCount "126"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 a6 5. e3 {+0.24/16 1} e6 {-0.37/19 0}
6. Qc2 {+0.30/16 1} c5 {-0.29/15 0} 7. cxd5 {+0.25/14 0} exd5 {-0.29/17 0}
8. g3 {+0.47/17 1} c4 {-0.26/19 0} 9. Ne5 {+0.45/14 0} Nc6 {-0.28/21 0} 10.
Bg2 {+0.51/15 0} Bb4 {-0.21/19 0} 11. O-O {+0.14/15 0} O-O {-0.23/18 0} 12.
Nxc6 {+0.12/17 1} bxc6 {-0.25/17 0} 13. Bd2 {+0.08/15 0} h5 {+0.07/17 0}
14. b3 {+0.16/15 0} h4 {+0.02/15 0} 15. bxc4 {+0.14/14 0} Bxc3 {+0.20/18 0}
16. Bxc3 {+0.06/13 0} Re8 {+0.23/15 0} 17. gxh4 {+0.02/16 0} Ng4 {+0.31/16
0} 18. h3 {-0.13/16 0} Nh6 {+0.27/15 0} 19. cxd5 {-0.18/15 0} Qxh4
{+0.56/14 0} 20. dxc6 {-0.14/13 0} Bxh3 {+0.59/15 0} 21. f3 {-0.41/14 0}
Bxg2 {+0.40/17 0} 22. Qxg2 {-0.53/14 0} Re6 {+0.50/16 0} 23. Be1 {-0.49/13
0} Qh5 {+0.26/21 1} 24. e4 {-0.22/15 0} Rxc6 {+0.29/16 0} 25. Qh2 {-0.13/15
0} Qg5+ {+0.28/15 0} 26. Qg2 {-0.36/14 0} Qh5 {+0.29/18 0} 27. Qh2
{-0.19/15 0} Qg5+ {+0.29/16 0} 28. Qg2 {0.00/13 0} Qd8 {+0.26/14 0} 29. Qd2
{+0.08/13 0} Rg6+ {+0.52/13 0} 30. Kf2 {-0.30/11 0} Qf6 {+0.59/14 0} 31.
Qd3 {-0.41/14 0} Rd8 {+0.72/16 0} 32. Rd1 {-0.24/13 0} Qf4 {+0.91/18 0} 33.
Ke2 {-0.95/15 0} Nf5 {+0.65/18 0} 34. Bf2 {-0.85/15 0} Rg2 {+0.62/11 0} 35.
Ke1 {-0.78/16 0} Rb8 {+1.31/17 0} 36. Rd2 {-0.71/14 0} Nh4 {+0.99/15 0} 37.
Ke2 {-0.87/12 0} Qg3 {+1.15/16 0} 38. Qe3 {-0.57/13 0} Qc7 {+1.63/10 0} 39.
d5 {-0.71/13 0} Ng6 {+1.46/12 0} 40. Rfd1 {-0.59/13 0} Nf4+ {+1.37/12 0}
41. Kf1 {-1.09/16 0} Rg6 {+2.22/17 0} 42. d6 {-1.50/15 0} Qc4+ {+2.61/15 0}
43. Re2 {-1.80/14 0} Nxe2 {+2.83/16 0} 44. Qxe2 {-1.89/12 0} Qe6 {+2.84/15
0} 45. Qd2 {-1.31/13 0} Rd8 {+3.13/7 0} 46. d7 {-2.25/14 0} Qh3+ {+3.13/4
0} 47. Ke2 {-2.03/15 0} Rf6 {+4.00/6 0} 48. Qc3 {+0.90/13 0} a5 {+6.41/1 0}
49. Rd3 {+0.87/15 0} Qe6 {-0.73/3 0} 50. Qc7 {+2.74/13 0} Kh7 {-3.40/3 0}
51. Qxd8 {+6.49/14 0} Qxa2+ {-6.55/17 0} 52. Ke3 {+6.59/13 0} Qb1 {-6.60/15
0} 53. Qxf6 {+7.03/15 0} gxf6 {-6.65/10 0} 54. d8=Q {+6.92/14 0} Kg7
{-6.39/7 0} 55. Qd4 {+6.74/14 0} Kg6 {-6.61/7 0} 56. Rd1 {+7.30/12 0} Qb3+
{-6.38/7 0} 57. Qd3 {+7.35/13 0} Qb6+ {-6.64/6 0} 58. Ke2 {+7.57/14 0} Qb8
{-6.89/5 0} 59. e5+ {+24.73/14 0} Kg7 {-9.72/10 0} 60. Rg1+ {+M9/15 0} Kh8
{-8.57/3 0} 61. Qc4 {+M6/23 0} Qb2+ {-M5/65 0} 62. Kd3 {+M5/31 0} Qa3+
{-M4/10 0} 63. Kc2 {+M4/44 0} Qa2 {-M3/8 0 Arena Adjudication} 1-0[/pgn]

Download Pgn Games(Zip)

Tournament Setup (Stockfish 13)
Stockfish 17
Stockfish 17.1
________________________________________________________________________________

Gauntlet test report Stockfish 18 dev(2026/08/02)
________________________________________________________________________________
Top

Return to “Computer Chess Club: Tournaments and Matches”