Free Chess Game Review: A Complete Workflow That Rivals Chess.com Premium

What does a free chess game review actually include?

A complete free game review covers four layers of analysis: engine evaluation (move quality scored in centipawns), move classification (Brilliant, Best, Inaccuracy, Mistake, Blunder), opening identification with database statistics, and tactical motif detection for missed shots. ChessDream's free analysis tool provides all four. Chess.com bundles them behind a $14/month Diamond subscription; Lichess offers them free without registration.

ChessDream vs Chess.com Premium vs Lichess — feature-by-feature comparison

The table below compares the three platforms on the analysis features that actually affect rating improvement. Pricing reflects published rates as of May 2026.

The functional gap between free and premium has narrowed sharply since 2023 because Stockfish 16+ runs efficiently in browsers via WebAssembly. The remaining premium-only feature — automated coaching commentary — explains less than 5% of the rating difference between players who do and don't review their games, according to amateur improvement data covered in our chess analysis statistics research.

The 3-pass workflow titled players use for post-game review

Running the engine and reading the squiggly evaluation graph is not analysis. Analysis is the deliberate, structured comparison between what you thought during the game and what was actually best. The 3-pass workflow below is the standard recommended in coaching literature, including IM John Bartholomew's "Climbing the Rating Ladder" series and the methodical approach taught in the endgame mastery guide.

Pass 1 — Engine-free walk-through (5 minutes)

Before turning the engine on, replay the game from move 1 and write down every move where you weren't sure of your plan. For each, jot one sentence: what you thought the position needed. This step is non-negotiable. The whole rating-improvement value of game review comes from comparing your in-the-moment thinking to the engine's verdict — but you need to capture the thinking first, before the engine output anchors you.

Pass 2 — Engine scan at depth 18-22 (3 minutes)

Run ChessDream's analysis at moderate depth. Flag any move classified as Mistake (loss of 1.0-2.0 in evaluation) or Blunder (loss of 2.0+). On a typical 40-move amateur game, this yields 2-5 critical positions. The depth-18-22 scan finishes in roughly 30 seconds and catches every meaningful error a human opponent would have exploited.

Pass 3 — Deep dive on critical positions (10-15 minutes)

For each flagged position, push depth to 28-30 and walk through the principal variation. The goal here is not to memorize the engine's line — it's to articulate the concept the engine sees that you missed. Was it a hanging piece in an unexpected file? An exchange sacrifice for piece activity? A king-safety idea that beats the obvious material grab? One sentence per position. Save those sentences. For deeper engine technique, see our advanced Stockfish analysis guide.

"The point of post-game analysis is not to be told you blundered. You already know you blundered — that's why you lost. The point is to convert one specific blunder into one specific principle you'll carry into the next ten games."

— GM Jonathan Tisdall, Improve Your Chess Now

Why free Stockfish analysis is engine-identical to Chess.com Premium

Stockfish is an open-source chess engine, released under the GNU General Public License. It has been the world's top-rated engine on CCRL since 2016 and has remained dominant through Stockfish 16, 17, and 18 releases. Chess.com licenses the same Stockfish that ChessDream and Lichess run — they cannot make it "more accurate" because the engine is identical at the source level. What premium services package is faster cloud compute (depth-30 in 2 seconds vs. 8 seconds in-browser) and human-readable coaching captions generated by a separate LLM layer.

For amateur game review, the speed difference does not affect insight quality. A depth-22 analysis catches the same blunders as a depth-30 analysis in 95%+ of positions, according to evaluation-stability benchmarks from the Chess Programming Wiki. The remaining 5% involves deeply tactical lines that the engine alone cannot translate into transferable patterns anyway — only structured human review, like the 3-pass workflow above, does that.

Frequently asked questions