Verbatim quotes from Chess.com forum threads. These players are asking for the exact feature we're designing — AI explanations of WHY moves are mistakes, not just labels.
Theme 1: "It tells me WHAT but not WHY"
"For some moves in game review it will say (so and so move) is best. Yet it doesn't give a description."
"At the rating I'm on its severely lacking in any insight to help me improve."
"I don't understand what stockfish wants me to do by certain moves."
"I forked the queen and rook with a knight and it was an 'inaccuracy'. The machine never gives any reasoning, so it is pointless without context."
"Once it said I should have connected my rooks. What does that even mean?"
Theme 2: Misleading or confusing feedback
"You missed an opportunity to win a pawn — but actually that move was setting me up to win a rook."
"The game review script not realizing that a mate is going on and giving random positional tidbits as 'explanation' is a major bug."
"The review bot gave me similarly bad advice. It told me to check the king, when that gave me no advantage."
"A move that set up a major gain of material or even a checkmate is classified as a 'mistake'."
Theme 3: Engine advice isn't human advice
"Are we all supposed to take from these game reviews that we should try to play more like an engine rated 3400 than a human!?"
"This fragments ones ability to understand the individual move in the context of the entire game."
"The new Game Review removes these lessons and scatters the information needed for analysis. Tasks like finding blunders now require multiple steps."
"Anybody find that the game review makes any sense anytime?"