We’re absolutely thrilled to announce the successful conclusion of “The FedCSIS 2025 Challenge: Predicting Chess Puzzle Difficulty” right here on Knowledgepit.ai! This second edition of our unique chess machine learning competition has been a resounding success, thanks to all of you.
This past month, our platform buzzed with activity as participants tackled the challenge of predicting chess puzzle difficulty. We provided significantly larger datasets (over 4.5 million instances!) and new data fields, including 22 precomputed success probability predictions, to give everyone a solid starting point. The goal was clear: build the best models to predict a puzzle’s difficulty, measured by its Lichess Glicko-2 rating, by analyzing initial board positions (FEN) and solution moves (PGN).
“The level of engagement and the quality of submissions in this second edition truly impressed us,” says Dominik Ślęzak – Founder & President of QED Software, the driving force behind this challenge. “It’s inspiring to witness how advanced AI algorithms can tackle the complex nuances of chess puzzle difficulty, a task previously reliant on extensive human testing. This competition truly showcased the power of combining machine learning with the timeless game of chess.”
Meet the Winners!
After intense competition and countless hours of model tuning, we’re incredibly proud to celebrate the teams that demonstrated exceptional skill and accuracy:
First Place
Team Name: ousou
Members: Sebastian Björkqvist (ousou)
Rank in the first contest edition: 3
Second Place:
Team Name: bread emoji
Members: Marco Cognetta (mcognetta), Tyler Woodruff (hhhtylerw), Oleg Filatov (chadius), luke x (arcteryx)
Rank in the first contest edition: 1
Third Place:
Team Name: transformer_enjoyer
Members: Szymon Miłosz (sselterf)
Rank in the first contest edition: 11
A massive congratulations to all our winners! Each of the top three teams has secured free registration for the prestigious FedCSIS conference, taking place in Kraków, Poland, from September 14-17. We can’t wait for them to represent our community there!
