Since you’re here...

We hope you will consider supporting us today. We need your support to continue to exist, because good entries are more and more work time. Every reader contribution, however big or small, is so valuable. Support "Chess Engines Diary" even a small amount– and it only takes a minute. Thank you.
============================== My email: jotes@go2.pl



New Chess Engine Testing Policy (Starting Jan 1, 2026)

on


 New Chess Engine Testing Policy (Starting Jan 1, 2026)

As you’ve probably noticed, there’s been a flood of new chess engines lately, many of them AI-generated.
It’s incredibly easy now—we’ve tested it ourselves and managed to build a basic engine in under three minutes. While these engines are quite primitive and lack significant playing strength, the sheer quantity of them is the problem. Testing hundreds or even thousands of such engines is no longer feasible. It would split the CEDR league into too many groups and stretch a single season out for months, which we want to avoid.
Therefore, we are introducing new testing rules starting January 1, 2026. Here is a summary of the changes:

CEDR Chess Engine League

The league consists of 64 engines—mirroring the 64 squares of a chessboard. The competition is structured into four divisions: Leagues 1, 2, 3, and 4. A system of promotion and relegation will be applied at the end of each season.
Top performers from the Qualifying Rounds will advance to the 4th League. To be eligible for the qualifying stage, an engine must meet the following criteria:
a) New engine versions must achieve a score of at least 50% in a preliminary test against our benchmark (either Critter 1.6a or the engine currently ranked 70th in the CEDR rankings).

The Qualifying Rounds will be held within a single group, with the specific tournament format determined by the total number of "promising" engines participating.

Testing New Engines 
New engines will be evaluated in a single tournament that includes a benchmark engine (either Critter 1.6a or the engine ranked 70th in our CEDR rankings). Only engines that achieve a score of 50% or higher against the benchmark will qualify for further testing and potential entry into the league.

We will continue to test chess engines using a time control of 3 minutes + 3 seconds increment per move. We believe this setting minimizes variance and prevents fluke results. Testing will be conducted across multiple Chess GUIs, with a primary focus on Round Robin tournaments.

If you appreciate our work and would like to support us, please consider making a donation. The PayPal link can be found below.












Comments

Post a Comment