Phase and colour
The same category behaves differently as White and as Black, or in endgames versus middlegames. The split is reported rather than averaged away.
Recurring mistakes
Find the Mistakes You Repeat
One mistake may be isolated. When the same decision fails across several games, it becomes a clear training target. ChessLogix finds those patterns and shows which ones cost you most.
Free Finding, counting and training patterns costs nothing
- No Chess.com or Lichess password — a public username is all it needs.
- The learning platform is free — AI Move Explanation and Meta-Coach are an optional membership.
- Every recommendation comes from your own games and Stockfish analysis.
- Each recurring decision is named, not left as a red label on a move.
- Priced in win probability, so a rare expensive habit can outrank a common cheap one.
- Counted, so you can see the difference between one bad afternoon and a pattern.
Why Single-Game Review Plateaus
Reviewing one game tells you what went wrong in that game. It cannot tell you whether the same thing went wrong last week, and that is the question that decides what you should study.
So most review ends up chasing the last painful loss. The game you remember is the dramatic one, and the decision quietly costing you the most points over a month is rarely the one you remember at all.
What the Report Gives You
Every significant swing is tagged with a decision category. The categories are then counted, priced, and ranked across everything you have imported.
- Named patterns with their win-probability cost and how many times each occurred.
- Error types underneath, counted separately — the mechanism, not the mood.
- Training Focus is the hand-off: each tag maps to the exercises that address it.
One Pattern, With Everything Behind It
Open a pattern and it shows its working: how often the decision came up, what it typically cost you, whether recent games are going better or worse than older ones, how many openings and colours it turns up in, and the clearest example from your own games.
- How often it happens and what it typically costs, side by side.
- Recent games against older ones, so "this is getting worse" is something you can see rather than something you suspect.
- The example position, then Train this pattern.
How a Pattern Is Found
-
Critical moves
Engine analysis marks where the evaluation moved materially against you.
-
Categories
Each swing is tagged by what the position offered and what you played.
-
Repeated pattern
Tags are counted across your history and split by phase, colour and speed.
-
Priority
Categories are ranked by how often they appear and what they cost.
-
Training
The top category hands its own positions to puzzles or opening prep.
Import games you played after the training and the same counts are worked out again. If the habit turns up less often than it used to, you can see that directly.
Split the Way the Fix Is Split
A repeated decision is not automatically a habit. It may be a preparation gap, a calculation problem, a clock problem, or something positional — and the report splits along the lines that tell those apart.
Time control
A tendency that only appears in bullet is usually a clock problem rather than a chess problem. Separating speeds is what makes that visible.
Opening spread
A pattern confined to one opening is preparation. The same pattern across six openings and both colours is something else, and the spread is printed either way.
Frequently Asked Questions
How many games do I need?
One game is enough for explanations. Seeing what repeats takes more — ten to twenty games in the same time control is usually where real habits start to stand out from ordinary bad days. Every figure is shown with the number of games behind it, so you can judge it for yourself.
What does "cost" mean here?
How much the mistake changed your chances of winning that game. That is why a habit can rank first on only a handful of occurrences — each one was expensive. Chances of winning are used rather than raw engine numbers because losing half a pawn in a balanced game matters far more than losing half a pawn in one already decided.
Is finding my recurring mistakes free?
Yes. Detecting patterns, counting them across your games, ranking them by cost, opening a pattern to see the positions behind it and sending it into training are all free. Meta-Coach — the AI layer that reads across everything, picks one priority and writes out the reasoning — comes with the monthly AI membership. You do not need it to find your patterns or to train them.
Does it work with Chess.com games?
Yes. Chess.com and Lichess accounts both import directly, and PGN upload covers everything else. Patterns can be computed per account, which is useful if you play seriously on one platform and casually on another.
Is a pattern the same as a weakness?
Not exactly. A pattern is a recurring decision that costs win probability. Whether it is a knowledge gap, a calculation habit, a clock problem or a positional misunderstanding depends on context — which is why the report splits by phase, colour and time control instead of handing you a single label.
Does ChessLogix diagnose my chess psychology?
No. Reports describe recurring decisions on the board and how often they recur; whether you were short of time, distracted, or had never seen the idea is not something the position records. Some pattern names carry the shorthand players already use with each other and do sound psychological — the underlying finding is always a board situation and a count, and both are shown.
Can I tell whether training worked?
Yes. Import the games you played after the training and see whether the habit you were working on turns up as often as it used to. That is the whole reason for tracking patterns rather than just reading a report and moving on.
Import Enough Games to See a Pattern
A single game cannot tell you what repeats. Import a batch from Chess.com or Lichess and let the counts build up — that is where the ordering comes from. Finding, counting and training your patterns is free; Meta-Coach is the optional AI layer that reads them for you and picks one to start with.