A few days ago, I wrote about why I built ChessIQ: I wanted to train the part of chess that happens before calculation.
Not “find the winning move,” but the harder and more common judgment: which side is better, and why? A player may have more space, safer king, stronger piece activity, or an easier position to play—even when no tactic is available.
That core loop has not changed. ChessIQ still asks you to commit to White better, Equal, or Black better before revealing the Stockfish evaluation and AI explanation. But v1.5 improves what happens after that judgment.
A position is not an isolated diagram
The biggest update is the expanded library: ChessIQ now contains 1,200+ real-game positions. Half come from Lichess games; the other half come from international over-the-board games from the early 2000s.
More importantly, many puzzles now include a metadata bar beneath the board. Tournament-game positions show the players, event, and year, while Lichess-sourced positions are identified as online games.
That sounds small, but it changes the learning experience. A position is no longer just an engine snapshot. It came from a real game, with a history and a continuation worth investigating.
From a guess to deeper analysis
v1.5 also adds an export menu so that curiosity does not end at the puzzle.
You can copy the position’s FEN or open the exact board directly in Lichess Analysis or Chess.com Analysis. For positions with an associated PGN, you can also open the full source game on Lichess.
FEN answers, “What can happen from this position?” PGN answers, “How did this position arise, and what happened next?” Both matter for positional training. An engine can tell you that one side is better; the full game can show whether that advantage was actually converted.
I wanted these tools to extend the training loop, not replace it. You still make your positional judgment first. The exports are there for the moments when a position earns more of your attention.
What comes next: Ranked and Rush
Next: Ranked and Rush modes.
Regular ChessIQ play should remain deliberate: take time, assess the position, and build Position Rating through sustained accuracy. Rush will explore the other side of positional skill—fast recognition—with three- or five-minute sessions, a separate leaderboard, and a clear switch from regular rated play.
The design question is how to make speed exciting without rewarding blind guessing. A good fast mode should reward recognition, not replace understanding.
If you are interested in positional chess, chess-training systems, or rating design, I would love your feedback—especially on what would make a speed-based positional mode feel competitive without becoming shallow.
Try ChessIQ at chessiqapp.com, and read the original build post for the data pipeline, theme normalization, and rating-design thinking behind it.
Top comments (0)