High accuracy, but a lost game? Make sense of Game Review
A good percentage does not explain the whole game. Connect accuracy to the result, evaluation graph and critical position, then retry a decision in MoveForge Game Review.
Why can you lose a game with good accuracy?
You finish a game, open the analysis and see a respectable accuracy score. Yet the result is a loss. That is not necessarily a contradiction. Accuracy summarizes the assessment of your moves; the result records how this particular game ended. A long stretch of good play does not undo a later mistake.
When using Game Review in MoveForge, look beyond the percentage. Connect it to the result, the evaluation graph and the position on the board. The goal is not to defend your statistics. It is to find a decision you can handle better next time.
Result: how did the game end?
A win, loss or draw records the outcome. It does not tell you who held the advantage earlier or whether the winner took every opportunity. Winning does not erase your mistakes, and losing does not invalidate your good decisions.
Accuracy: how were the moves assessed?
It is an analysis summary, not a percentage of winning positions, a rating or a promise of victory. Read it alongside specific moments in the game. The number alone does not show how difficult your decisions were.
Three situations a percentage cannot explain on its own
Imagine three different losses. Each calls for a different training takeaway, even if the final summaries look similar.
- Good play, one decisive mistake. You maintain a playable position for a long time, then leave a piece hanging. Examine that decision rather than dismissing everything you did well before it.
- The problem began earlier. You enter a difficult ending and defend sensibly, but your opponent never gives up the advantage. Look for the decision that created the problem, not just the last moves before defeat.
- The clock decides. Your position remains promising, but you lose on time. The evaluation of the board does not explain your whole time-management problem. Check how the game ended in its record or game information.
These are situations to investigate, not automatic diagnoses of every loss. Establish what actually happened in your game first.
Read the evaluation graph for meaningful changes
First check that you are examining your own moves and the correct color. Then find a stretch where the position changed significantly. Return to the board just before that change and follow what happened next.
The largest numerical swing is not always the most useful training moment. Losing your remaining chances in an already lost position is different from making the first mistake that forced you onto the defensive. Equally, do not stop at your first error: your opponent may have offered you another chance later.
An engine evaluation describes a position, not just its material balance. An advantage can come from an attack or active pieces. It does not automatically mean the position is easy for a human to win.
A mistake label is the start of an explanation
Select one important move highlighted in the review. Examine the position before and after it, then play through the opponent's reply suggested by the engine. Do not stop at the first arrow. Follow enough of the line to see a concrete consequence.
That might be a lost piece, an open line toward your king or a difficult ending. If you cannot identify the consequence, you still have a question to investigate, not a finished lesson.
It also helps to separate the tools' roles: Stockfish supplies evaluations and lines; the application uses that analysis to present move classifications. A label helps you locate a moment worth checking. It does not replace understanding what happened on the board.
A takeaway that is hard to train
“I made a blunder.” “The engine preferred another move.” “My accuracy could have been higher.” Each describes an analysis result, but none tells you what to check before your next move.
A takeaway you can use at the board
“Before capturing a pawn, check my opponent's forcing checks.” “After moving a defender, check what is left unprotected.” Write one such reminder in your own notes, based on the position you examined.
Retry the decision on the same board
At a key moment in Game Review, use the retry action to make the decision again. This gives you a chance to test your own idea instead of only watching the best continuation.
Before moving, identify your opponent's threat, choose a candidate and predict the strongest reply. After the attempt, compare your idea with the analysis. Notice not just the move, but whether you anticipated its consequences. Return to the review when the attempt is complete.
If you have already seen the solution, treat the attempt as practice rather than an independent test. The next time you open this game, revisit the position and try to explain the idea yourself before checking the analysis.
Compare similar games, not isolated percentages
For your own review, choose a few games at the same time control and analyze them under comparable conditions. Instead of only asking whether accuracy changed, check whether the same problem returns: missed checks, undefended pieces or an ill-judged exchange.
Do not treat accuracy percentages from different websites as a single shared scale. Calculation methods and analysis settings can differ. A different score does not, by itself, prove that an analyzer is malfunctioning.
The practical question is: can I now find a better move in a similar situation and explain why it works? That is a more concrete training goal than improving the summary alone.
A 10-minute exercise after analysis finishes
Choose one completed game and import its PGN into Board Lab. Wait for Game Review to finish. These ten minutes are for your work with the results; they do not include engine calculation time.
- 2 minutes: check the result, your color and the evaluation graph. Choose one moment you want to understand.
- 4 minutes: replay the mistake and the opponent's response. Identify a specific consequence.
- 3 minutes: retry the decision and explain your chosen move.
- 1 minute: write one reminder for your next game.
This is a suggested short session, not an application time limit. A difficult position may need longer. Finishing with a well-defined question is better than pretending you understand everything.
Common questions about accuracy and Game Review
Does high accuracy mean a high rating?
There is no simple conversion. One game does not describe a player's overall strength. The type of position and the difficulty of the decisions also matter, not just the percentage in the summary.
Do I need an account to review my game?
No. You can paste a PGN into MoveForge and complete a full Game Review, including a local retry, without signing in or paying. Your personal game history and opening repertoire are separate account-linked tools.
Do I need to correct every highlighted move?
Not in one session. Start with a moment whose consequences you understand and that gives you a useful reminder. You can return to the others later.
Take a better decision away from analysis, not just a number
A useful session does not have to end with a higher percentage or every variation explored. It should leave you able to explain what you missed, which reply mattered and what you will check next time.
In MoveForge, connect the summary to the board and the board to your own attempt. Turn your mistakes into better moves, starting with one real decision from your game. Work with completed games; do not use engine analysis as assistance during a game against another person.
What is hardest to understand after analyzing a game?
- How accuracy relates to the result
- The evaluation graph
- Why a move was marked as a mistake
- What to work on in my next session