What a Chess DNA report actually measures
Colour, time control, and opening splits only mean something when they beat your own average by enough games. That is the Chess DNA report.
A DNA report is the pattern across a corpus, not a grade of one game. It keeps a split when the bucket is large enough and the score sits clear of your baseline. Noise is dropped on purpose.
Three splits, one baseline
Colour, time control, and the named opening are compared to your overall score in the same import. A 40% Black score is a finding only if your overall score is meaningfully higher and Black has enough games.
Chessigma needs at least five games in a bucket before it will name it. That is a floor for “this slice exists,” not a reason to abandon an opening.
Why this is not an accuracy number
Accuracy and centipawn loss grade moves against an engine. DNA never does that. It only asks whether a slice of your own games has been scoring differently from the rest.
If nothing clears the threshold, the report stays quiet. Inventing a style problem from two blitz losses is how chess sites waste your study time.
Frequently asked questions
Is DNA the same as a game report?
No. A game report looks for one repeated departure in that game. DNA looks across the whole import.
Can I run DNA without an account?
The public preview can read a recent public history. Saving the import and refreshing it needs an account.