The game that tells you where you lost

Lose a level in most block puzzles and you get one of two screens: a “continue?” offer with a price on it, or a “try again” button. Neither tells you anything. You know the run ended. You do not know when it was already over.

Chess apps solved this years ago. After a game, the engine shows you the move where the evaluation fell off a cliff, and what you should have played instead. That single feature is most of why people got better at chess on their phones.

O’Block 1.0.1 does the same thing for the Journey.

What the review shows

The Journey has an exact solver. It exists because every level ships with a fixed queue of pieces and a set of target cells, so “is this position still winnable?” is a question with a definite answer, and the game already uses it for the hint button: a paid hint is the first move of a shortest winning line, and if no line exists the game tells you for free. We wrote about that here.

When a level is lost, the same solver now walks your run backwards. Position by position, it asks whether a win was still open from there. The first position that answers yes is the turning point: the move you made from it is the one that closed the door.

The loss page then shows that board. Your piece is outlined where you put it. The piece that would have kept the win is solid where it belonged. Underneath: “At move 7 you were 4 moves from a win.”

Retry from move 7

Under that is a second button, next to the usual Retry: Retry from move 7. It puts the board back exactly as it was before the turning move, with the same pieces in the rack and the same queue behind them, and the level goes on. The clock keeps its time.

It is free. Undo during a level costs coins; this does not, because the point of a review is to let you act on it, and a review you have to pay to use is a continue offer wearing a different coat.

What it will not do

The solver has a budget, tens of thousands of positions per move, so it stays fast on a phone. A position it cannot settle inside the budget is skipped rather than guessed at.

And there is one case where the review is withheld entirely: when the solver’s best move at the turning point is the move you actually played. That means the real turning point is later and the search there ran out of budget. Showing you a board that says “you should have done what you did” is worse than showing nothing, so it shows nothing.

Classic and the Daily do not get a review. Both deal pieces at random, so there is no fixed queue to search and no honest answer to “was this winnable”. A review that speculates is not a review.

Why a puzzle game should explain itself

Puzzle players blame themselves. Lose a level three times and the story you tell yourself is that you are bad at it, when the truth is usually that you made one wrong move on turn six every time and never saw it.

Telling someone exactly which move it was is the difference between a game that sells continues and a game that teaches. We would rather make the second one.