A chess analysis board is a board you can move pieces on freely, with an engine evaluating live, plus variation branches, arrows, and FEN/PGN import. The lichess analysis board is 100% free, needs no account, and is the best option for most people. Use it right by guessing your move and writing your reason down before the engine is on. What no analysis board can do is remember you, so it will never notice that you keep losing the same way.
Interactive lesson — Knight fork: the payoff of seeing one move ahead: Play it on the board below, and make the key move yourself. White to move. The knight on d5 is one leap from disaster — can you already see it? Ne7+ — a fork with check. The knight hits the king and the queen at once. The king is in check, so it has to step aside — it can’t defend the queen. Nxc8 wins the queen. Seeing this one move ahead is the whole game.
Search for a chess analysis board and you get a dozen products but almost no instructions. The tool itself is genuinely simple, and the good ones are free: lichess gives you a full engine board with unlimited variations at 100% no cost, no sign-up, no daily cap. The hard part is the habit. Most players open an analysis board, switch the engine on immediately, scroll through the suggested moves nodding along, close the tab, and learn nothing that survives to the next game. This guide covers what an analysis board is, which free one to use, the five-step routine that makes a session stick, and the one thing an analysis board structurally cannot do for you.
An analysis board is a chessboard with the rules of "whose turn is it" relaxed. You can move either side's pieces, take moves back, and branch off into a different line without losing the one you came from. Running alongside it is a chess engine producing a live evaluation of whatever position is on the screen. Four features make it a study tool rather than a toy:
Most boards add arrows and square highlights, which sound cosmetic but are not: drawing the threat you missed is one of the cheapest memory aids in chess study. Some, like lichess, bolt on an opening explorer and endgame tablebases in the same screen, which turns the board into a small research station.
You do not need to pay for this. The best analysis board available to a club player costs nothing.
The lichess analysis board is free, open source, and needs no account. Stockfish runs locally in your browser, which means the engine is not metered: there is no "3 analyses per day" ceiling, because you are paying with your own CPU. You get unlimited variations, arrows, FEN and PGN import, an opening explorer showing how master and amateur games continued from the exact position on screen, and tablebase-perfect answers once the board is down to seven pieces. You can save any line as a Study and share the link. For the overwhelming majority of players, this is where to start and where to stay.
The Chess.com analysis board is well built and clean, and its real advantage is proximity: if your games are on Chess.com, opening any of them on the analysis board is one click from your archive. The free tier is usable but limited — engine depth and the number of full analyses per day are capped, with the rest sitting behind a Diamond membership. If you already pay, use it. If you do not, the limits will find you on the day you actually want to work. There is a longer comparison in our guide to a free Chess.com game review alternative.
Two routes, and picking the right one saves minutes every session.
Import, when the position already exists somewhere. Copy the FEN string or the whole PGN and paste it into the board's import field. This is exact and takes seconds. Chess.com and lichess both expose the FEN of any position in their game viewers, so moving a position from a finished game to an analysis board is a copy and a paste.
The board editor, when you are building from scratch. Clear the board, drag the pieces you need onto their squares, then set the two things beginners forget: whose turn it is, and castling rights. A position with the wrong side to move can flip from winning to losing, and a stale castling flag quietly invents a resource that does not exist. Check both before you trust the evaluation.
One practical note: set up the position where you were confused, not the position where the game visibly collapsed. Those are usually four or five moves apart. The blunder is the symptom; the moment you stopped knowing what to do is the lesson.
Here is the routine. It takes about fifteen minutes per position and it is the difference between studying and spectating.
Step two is the whole trick. Everything else is logistics. For the wider workflow this routine sits inside, see how to analyze your chess games, and for the single-position skill in more depth, how to analyze a chess position.
Engine scrolling is what almost everyone does instead of analysis. You open the board, the evaluation bar is already on, you click forward, the engine says the move was inaccurate, you say "ah, yes", you click forward again. Twenty minutes later you have read forty engine verdicts and retained none of them.
It fails for a specific reason. Learning needs a prediction to correct. If you never commit to an answer, there is nothing for the engine to correct, so the information passes through you like a weather forecast for a city you do not live in. Recognition feels like understanding and is not — you nod because the move is obviously good once shown, which tells you nothing about whether you would have found it.
The fix is mechanical, not motivational: keep the engine off until you have written your move and your reason. If your board turns the engine on by default, toggle it off before you look at the position. Two seconds of friction converts the whole session. The same principle drives chess visualization training, where the value comes entirely from calculating before you look.
An analysis board is stateless. It evaluates the position in front of it, perfectly and instantly, and then forgets you existed. That has three consequences worth being blunt about.
It has no memory of you. It will never tell you that you have lost eleven games the same way, because it has never seen the other ten. Every session starts from zero, which means the recurring mistake — the one actually capping your rating — is invisible by construction.
It answers, it does not diagnose. The engine tells you the best move here. It does not tell you which kind of position you should stop entering, or whether your problem is calculation, time pressure, or a structural habit. That requires looking across many games, not one.
It cannot prioritise. Forty engine verdicts in a session are forty items with no ranking. Which one costs you the most rating? The board has no opinion.
This is the honest gap, and it is where a different category of tool belongs. Chess DNA imports your own games by Chess.com or lichess username, analyses them with Stockfish 17 at depth 18, and reports the weakness patterns that recur across your whole history, scored on 8 skill dimensions. The published dataset behind it covers 6,288 real games, which is what makes "you lose most rating to one specific pattern" a measurement rather than a guess. It is not an analysis board and does not replace one — you still want lichess open when you want to explore a position. The two tools answer different questions.
See the patterns across your own games → Free, no install, import by username.
Three different jobs, three different tools. Most players use one for all three and wonder why progress stalls.
| Tool | Scope | The question it answers | When to reach for it |
|---|---|---|---|
| Analysis board | One position | What is really going on here, and what would have happened if…? | You were confused in the moment and want to explore branches freely. |
| Game review | One game | Where did this game turn, and what was the worst move? | Straight after a loss, for a fast post-mortem and a shortlist of moments. |
| Pattern analysis | Many games | What do I keep getting wrong, and what should I fix first? | Monthly, to choose what to train instead of guessing. |
A healthy loop runs all three: review the game to find the moments, take the confusing moment to an analysis board to understand it, and check the pattern layer every few weeks to decide what deserves your training time. Our walkthrough of lichess game review covers the middle step, and the chess analysis app comparison covers the third.
A chess analysis board is an interactive board where you can move both sides' pieces freely while a chess engine evaluates every position live. Unlike a game viewer, nothing is fixed: you can take moves back, branch into as many variations as you like, and the board keeps them all in a tree. Most analysis boards also let you set up any position from scratch, import and export FEN and PGN, and draw arrows and highlights. It is the standard tool for asking "what would have happened if I had played this instead?"
Yes. The lichess analysis board at lichess.org/analysis is completely free with no account and no limits, and it is the best free option for most players. Stockfish runs locally in your browser, so evaluation is unlimited rather than metered, and the same page includes an opening explorer with master and community game statistics plus endgame tablebases for positions with seven pieces or fewer. You can also paste a FEN or a full PGN to jump straight to any position you want to study.
There are two ways. The fastest is to paste a FEN string or a PGN into the import field, which reproduces the position exactly and takes a few seconds. The other way is the board editor, where you drag pieces onto empty squares, choose whose turn it is, and set castling rights. Use import when you already have the game and the editor when you are building a position from a book, a puzzle, or memory. Always double check whose move it is, because a position with the wrong side to move gets a completely different evaluation.
No. A game review takes one finished game and walks you through it, labelling moves as blunders, mistakes, or inaccuracies and giving you an accuracy number. An analysis board is a blank workbench: it does not know your game, it just evaluates whatever position you put in front of it. Game review is for the post-mortem of a single game, while an analysis board is for interrogating one position in depth, including lines that never appeared in any game. Serious study usually uses both, in that order.
For free use, lichess. It has no daily caps, needs no account, runs Stockfish in your browser, and bundles the opening explorer and tablebases on the same screen. Chess.com's analysis board is well built and worth using if you already pay for Chess.com and keep your games there, since analysing your own games is one click from the archive. Chess.com's free tier limits how much engine analysis you get per day. Both use Stockfish underneath, so the evaluations themselves do not meaningfully differ.
Because an engine answers a question you did not ask. The common failure mode is engine scrolling: turning the evaluation on immediately and clicking through moves while agreeing with each one. Nothing is retained because you never committed to an answer, so there is nothing to be corrected. The fix is to guess first and write down one sentence of reasoning before the engine speaks. An analysis board also has no memory of you, so it can never tell you that this is the eleventh game you have lost the same way.
