Free online Chess server. Play Chess now in a clean interface. No registration, no ads, no plugin required. Play Chess with the computer, friends or random opponents.
Create a game Tournament Simultaneous exhibitions
Chess basics Puzzles Practice Coordinates Study Coaches
Lichess TV Current games Streamers Broadcasts (beta) Video library
Players Teams Forum Questions & Answers
Analysis board Board editor Import game Advanced search
Sign in
  1. Forum
  2. Lichess Feedback
  3. Lichess Analysis not always correct


Sometimes Lichess judges moves as mistakes (or blunders). In this game (below) that is definitely not the case! Here I played as white and made a good move, but Lichess noted it as a mistake.
I know this is not the easiest thing to program, but can you take a look at it? Thanks!

The engine often disagree with itself at different depths.

By heading the black king into the whites territory, that king should have expected trouble. During the opening or middle games, if the king stays in their own territory they tend to have a longer game life. Wait until the end game to bring a king towards the centre.

Example by playing poorly a rook there was a mate:
21. ... Rab8 22. Bxh6 Ra8 23. Rh21#

Placing the king anywhere else, would place the king in a better position than that move 20. ... Ke5.

Horizon effect: Expected moves are different depending on the depth of perception.

Nc3 would have won two pieces, and most of black's pawns soon thereafter. Bxh6 also wins two pieces, but takes longer to do so. It seems Stockfish missed the line where this happens, and the +7.8 evaluation came from a line (d5 Rhf1) where white only wins the one piece and several pawns.