What is Computer Chess Rating Lists (CCRL)?

Updated

CCRL (Computer Chess Rating Lists) is a volunteer computer chess engine testing and rating group founded in 2006, best known for publishing ELO rating lists for chess engines. It is useful if you want a community-maintained, game-based estimate of engine strength across several time controls and variants — but it is not an official authority, and its ratings are one input among several rather than a definitive ranking.

What CCRL actually is

CCRL describes itself as computer chess enthusiasts who enjoy engine testing. It was formed in 2006 by Graham Banks, Ray Banks, Sarah Bird, Kirill Kryukov, and Charles Smith, and currently active testers include Graham Banks, Gabor Szots, Shaun Brewer, Basti, Andres Valverde, and Mattia Giambirtone.

Two points shape how you should read its output:

  • Testers follow general testing guidelines, but each tester chooses what and how to test. Some use systematic methods, others prefer tournaments. That means the lists are a combined group hobby effort, not a single controlled experiment.
  • CCRL states it does not judge engine originality. Inclusion or exclusion does not imply endorsement or rejection, and viewers must form their own opinions.

What data the site publishes

The site's main sections are About us, Rating Lists and Statistics, Testing Methodology, and Joining CCRL.

Rating lists are updated independently of each other. The lists named on the site are:

List Notes
40/15 Longer time control
Blitz Fast time control
Chess960 Shuffle chess variant
40/2 Longer time control
Archive Historical results
Chess324 Variant
DFRC Variant

Links to these lists sit below the Best Versions lists.

Beyond the ratings themselves, the site offers:

  • Downloadable games by engine, by month, or by ECO code.
  • ELO ratings saved for engines with 100+ games, so low-game engines may not have a stored rating.
  • Engine pages: clicking an engine name reveals its opponents, homepage links, and comparison tools.
  • An openings report showing ECO statistics, draw rates, and White win percentages.
  • Sortable columns on the lists.
  • LOS (Likelihood of Superiority) statistics, which show the probability of an engine outperforming the one directly below it.

How the testing works

The methodology page gives the concrete conditions behind the numbers:

  • Benchmark hardware: Intel i7-4770k.
  • Time control calibration: Stockfish 10 is used to calibrate time controls.
  • Books: generic books, maximum 12 moves.
  • Ponder: off.

Two practical consequences follow. First, ratings are tied to that hardware and calibration setup, so absolute numbers are not portable to your own machine. Second, CCRL itself warns that early ratings may fluctuate, and recommends consulting other rating lists for broader context.

How to use it well

  1. Pick the list matching your interest — for example Blitz if you care about fast games, 40/15 or 40/2 for slower play, or Chess960/Chess324/DFRC if you play variants.
  2. Check the game count before trusting a rating, since ELO is only saved at 100+ games.
  3. Use LOS to judge whether a gap between two adjacent engines is meaningful rather than reading raw ELO differences as exact.
  4. Open an engine's page to see who it played and follow homepage links.
  5. Cross-check against other rating lists, as CCRL advises.

Joining as a tester

If you want to contribute, the site says to sign up on the CCRL forum and contact a member. For security, only testers known to the community or vouched for by trusted members are accepted — so joining is not an open, automatic process.