[Ongoing] The Ultra-Bullet Evolution - PlentyChess Gauntlet

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

goni-K26
Posts: 185
Joined: Mon Mar 09, 2026 4:16 pm
Full name: Agron Kovaci

Re: [Ongoing] The Ultra-Bullet Evolution - PlentyChess Gauntlet

Post by goni-K26 »

PlentyChess Gauntlet Tournament Report (v8.0.0 vs v7.0.37)
Tournament Settings & Environment
• Tournament Type: Gauntlet (60 games per opponent, 2400 total games)
• Time Control: 9s + 0.1s (Ultra-Bullet / 5% of Standard 3m+2s Blitz)
• Hardware: Intel(R) Core(TM) i5-6500 @ 3.60GHz (Single Thread testing)
• Resources: 16 MB Hash, 5-man Syzygy Tablebases, Arena Chess GUI
Key Performance Highlights

Code: Select all

Total Games:  2400
Wins:         799
Draws:        1428
Losses:       173
Overall Score:63.0%
Elo Gain: PlentyChess 8.0.0 performed at a 2877 Performance Elo against a 2773 average opponent field strength. This results in a massive +104 Elo Delta over the field and brings its established rating to 2875 (+24 Elo gain).


PlentyChess 8.0.0 vs PlentyChess 7.0.37 Head-to-Head Breakdown

While PlentyChess 8.0.0 lost a tiny bit of ground overall compared to version 7.0.37 (-1.00% total score difference due to a high draw rate against specific engines), it showed massive evolutionary leaps against top-tier opponents:

The Stockfish Breakthrough: Version 8.0.0 showed immense progress against the strongest engines. It scored +9.0% better against Stockfish 18, +3.7% better against Stockfish 17, and +3.8% better against Alexandria 9.0.0 compared to version 7.0.37.

Biggest Improvements: The most notable performance jumps were against Velvet 8.1.0 (+4.7%), Horsie 1.1 (+4.3%), and Tarnished 5.0 (+4.0%).

Regression / Tough Matchups: Version 8.0.0 struggled significantly compared to v7.0.37 against RubiChess 20240112 (-12.8%), Viridithas 19.0.1 (-9.0%), and Obsidian 14.0 (-8.8%), where it settled for more draws/losses than its predecessor.
Conclusion
PlentyChess 8.0.0 proves to be a highly efficient, stable, and lethal engine in ultra-fast time controls. The optimization for single-thread micro-blitz environments clearly shines through, especially in its ability to squeeze better net results against the absolute top of the field like Stockfish 18.
Image