Simbelmyne - UCI engine Author:Sam Roelants
Rating Chess Engines Diary CEDR=2474
v.1.10.0:
What's new
Version 1.10 is a pretty huge grab-bag of changes. Lots of re-factors that make
the codebase easier to maintain, a lot of search improvements, and — for
the first time in several versions — a big round of improvements to the
evaluation function. I think Simbelmyne is slowly but surely getting to the
point where a NNUE-based evaluation becomes inevitable, but let's see how far
we can take her, first!
No gauntlet just yet, so I don't have a very accurate guess at the playing
strength, but self play tests say +215 STC and +200 LTC.
Added features
🔍 Search
Add pawn based correction history (46+-13.62) (#261)
Non-pawn based correction history (19.73+-8.64) (#298)
Material based correction history (6.99+-4.69) (#300)
Don't clear countermoves between searches (2.96+-2.36) (#263)
Negative extensions (2.29+-1.59) (#262)
Multicut (3.84+-3.07) (#264)
Triple extensions (3.30+-2.64) (#265)
Remove History aging (6.85+-6.05) (#268)
Remove delta pruning (3.51+-4.55) (#283)
⚖️: Evaluation
Pawn cache table (18.57+-8.21) (#282)
Bonus for knight/bishop shelters (5.26+-3.88) (#286)
Bonus for safe check options (12.52+-6.66) (#284)
Bonus for unsafe check options (15.32+-7.39) (#285)
Penalty for bad bishops (16.14+-7.65) (#287)
Square rule for passed pawns (7.40+-4.83) (#294)
Endgame scaling (11.37+-6.17) (#296)
Bonus for free/unhindered passed pawns (8.73+-5.39) (#295)
Include attacked stop square in free passed pawn condition (6.36+-4.43) (#299)
Bonus for protected passed pawns (3.73+-2.97) (#297)
🐛 Bugfixes
Pass conthist table to move picker by value, instead of copy (11.69+-7.72)
(#256)
⌛ Time management
Better hard/soft time limits (22.08+-9.26) (#272)
Node based scaling of soft time limit (9.24+-5.53) (#273)
best_move stability scaling of soft time limit (9.83+-5.80) (#270)
Misc
Big refactor of the SPSA tuning architecture (#254, #255)
Big refactor of the search history (#257)
Big eval refactor (#274, #275, #278, #280, #281, #288, #292)
Normalized eval using a logistic WDL model (#259)
Shrink TT entries (16.38+-7.96) (#260)
Use TT eval in Quiescence search (4.81+-3.60) (#267)
Make mv.is_quiet() return true for all non-tacticals (24.67+-10.53) (#249)
See the respective PRs for self-play results where relevant
Choosing a binary
This release comes with precompiled binaries for all major platforms. Because
the engine benefits tremendously from more modern CPU instruction sets, there
are binaries compiled for major instruction sets, following the x86-64
microarchitecture levels
as a naming scheme.
Realistically, on modern hardware (< 10 years old), you should be okay to use the
binaries labeled V3. If the engine crashes within the first seconds, try V2,
and so on.
Simbelmyne 1.8.0 - results:
| Rustic-Knight 1.0.0 | 7/7 | +7 | 7 Games |
| Camel 1.6.0 | 6.5/7 | +6 | 7 Games |
| Simplex 0.9.8 JA | 6/6 | +6 | 6 Games |
| Spaghet Pesto 1.0 | 6/6 | +6 | 6 Games |
| Beast 15 MPV | 0.5/6 | -5 | 6 Games |
| Clover 6.2 | 0/6 | -6 | 6 Games |
| Lynx 1.5.0 | 4.5/5 | +4 | 5 Games |
| Princhess 0.17.0 | 2/5 | -1 | 5 Games |
| Odonata 0.9.0 | 1.5/5 | -2 | 5 Games |
| Wasp 7.00 | 1/5 | -3 | 5 Games |
| Obsidian 12.29 | 0/5 | -5 | 5 Games |
| Clarity 7.0.0 | 1/4 | -2 | 4 Games |
| Stockfish dev-20240608 | 1/4 | -2 | 4 Games |
| Steel Fighter 2 | 0.5/4 | -3 | 4 Games |
| Little Beast 15 | 0.5/4 | -3 | 4 Games |
| Lc0 0.31.0-rc3 | 0.5/4 | -3 | 4 Games |
| Titan 1.1.0 | 0.5/4 | -3 | 4 Games |
| SF-PRO 10.06.2024 | 0/4 | -4 | 4 Games |
| CorChess dev-20240606 | 0/4 | -4 | 4 Games |
| PlentyChess 2.0.0 | 0/4 | -4 | 4 Games |
| Marauders 3.6 | 0/4 | -4 | 4 Games |
Comments
Post a Comment