Analyzing a study

Analyzing a game means following a single line from start to finish. A study is a tree: fifty branches that all look alike on the page, with nothing to say which one is worth +1.4 and which one −0.9.

Branch analysis answers that question — for every continuation of a move, in one pass.

Running the pass

Two entry points, one gesture:

  • In a study’s toolbar, the Analyze button’s chevron opens Analyze every branch.
  • In Reader mode, the top bar carries the same button.

The engine walks every position of the tree — not just the main line. A study of three hundred positions takes a few minutes; the button shows progress (Branches · 40 %), and Stop analyzing the branches cuts it short at any time.

Reading the result: the study’s point of view

The whole difficulty of a repertoire sits in one detail: +1.4 is excellent for White and disastrous for Black. A study on facing the London System is worked from Black’s side — reading White-oriented numbers there would mean flipping every one of them in your head, line after line.

Régence doesn’t ask that. Every score in a study is written from the point of view of its color: + always means “good for you”. The bar says so: “42 line(s) evaluated · from Black’s point of view”.

What you seeWhat it says
+1.4 at the end of a lineThis continuation leaves you clearly better.
−0.9It leaves you worse.
A red triangle before the scoreA move by your side loses ground in this line.
⚠ 3 not to be learned in the barThree lines of the study illustrate a mistake — they don’t dictate a repertoire.

A small mark also appears on the moves themselves, but only on yours: in a study, a bad move by the opponent is the subject of the lesson, not a mistake to flag.

Ranked crossroads

Once the analysis is done, the move list changes shape at every decision point. Competing continuations are no longer indented blocks stacked on top of each other: they gather in a titled frame — INSTEAD OF 3.Bb5 · 3 optionsin the order the study wrote them. The analysis reorders nothing: the move list is the map of your study, its landmarks stay put. The ranking is read from the numbers.

Each option carries its letter (A, B, C), the color of its arrow on the board, and its score in a column aligned to the right. That column does the work: three continuations compare in a single vertical glance, just like the engine’s best lines.

When an alternative is better rated than the move the study kept, a note says so. It isn’t necessarily an author’s mistake — a line is also chosen for what it is to play, not only for its number — but it’s worth a look.

What it changes for training

The analysis isn’t just a reading comfort: it’s what makes training selective.

  • It knows that a variation the opponent gets wrong is a punishment to work on — and that the same one played by your side is a counter-example not to learn.
  • It therefore feeds the Theory / Punishments / Everything selector on the start screen.
  • It lets the drill set aside the lines the study condemns, and judge your alternative answers instead of refusing them.

Without it, the drill still works — but it asks you everything, including what the author wrote to warn you off.

When the tree changes

Adding or removing moves afterwards breaks nothing: the scores stay on screen, dimmed, with the note (tree changed since). New positions show up as not evaluated — another pass catches them, and only recomputes what’s missing.