How scoring works

The short version

Each move has a par
Every move you guess is worth between 4 and 7 points, based on how hard it is to find. A move that’s the only legal one is worth 2.
Your first guess counts
Only your first try on each move is scored. You can keep trying until you find it, or have the trainer show it, but later tries don’t earn points.
Good alternatives earn credit
If Stockfish rates your move as good as the game move, you get full par. Close moves earn 60%, playable ones 30%. Missing a critical move costs a point.
You get a score out of 100
Your points are divided by the points possible, so every game is scored on the same scale. The game’s par tells you what a typical club player would score.

What each move is worth

Like the “Solitaire Chess” column in Chess Life, moves that are harder to find are worth more. Each move starts at 4 points and can go up to 7.

When	What it means	Points
Every move	The starting value.	4
Not obvious	A quick engine look (depth 8) picks a different move, so the game move takes deeper thought to find.	+1
Clearly best	The game move is noticeably better than every alternative.	+1
Only move	Everything else throws away a big part of the advantage. These are the critical moves.	+1
Forced	There’s only one legal move. It still counts, but it’s worth just 2.	2

If the master’s move was clearly worse than the engine’s choice, it stays at 4. You still get full points for guessing it, since the point is to predict what was played.

Credit for other moves

Stockfish’s evaluation of each move is turned into winning chances. Your move is compared with the game move by how much of those chances it gives up.

Your guess	Compared with the game move	Credit	On a 5-point move
The game move	Exactly what was played.	Full	5
Just as good	Gives up 2% winning chances or less.	Full	5
Close	Gives up 6% or less.	60%	3
Playable	Gives up 12% or less.	30%	2
Anything else	Gives up more than that.	None	0
Missed a critical move	On an only-move, a guess that gives up 30% or more.	Penalty	−1
Skipped	You had the trainer show the move, or the clock ran out.	None	0

Where the analysis comes from

Every game in the library was analyzed ahead of time with Stockfish at about a million positions per move, keeping the five best moves in each position. That’s why library games score instantly. If you guess a move outside those five and it could still earn credit, the trainer checks it on the spot.

Games you paste or upload are analyzed in your browser while you play, a little less deeply. Scoring works the same way.

Your score and par

At the end, your points are divided by the points possible and scaled to 100. A 30-move game and a 12-move game are on the same scale.

Score	Estimated strength	Rating
95–100	Senior master	2400+
85–94	Master	2200–2399
75–84	Expert	2000–2199
65–74	Class A	1800–1999
55–64	Class B	1600–1799
45–54	Class C	1400–1599
35–44	Class D	1200–1399
Under 35	Developing player	under 1200

The bands are rough estimates, not an official rating. One game is a small sample, so a few games tell you more than one.

Par for the game

Par is the score a typical club player, around 1800, would be expected to get on this game, playing the side you chose from the move you started on. It’s worked out from the analysis of every position you’ll face. Games full of non-obvious moves have a lower par. Games where many moves are about equally good have a higher one.

Across the library, par runs from the high 40s to the high 70s, and the average game sits around 65. You’ll see the par in the Score panel as soon as a library game starts, so you know what you’re aiming for. It’s an estimate from a model, not an average of real players’ scores.

Two finished games

Here’s how two completed games might look. The guesses are made up, but every point value comes from the trainer’s real analysis of these games.

Botvinnik vs. Capablanca, AVRO 1938

Playing White from move 12 · 30 moves guessed

A strong session. Several guesses weren’t Botvinnik’s moves but scored full points anyway because Stockfish rates them as just as good. The two costly moments came in the famous finish: missing 30.Ba3, the quiet bishop move that sets up the combination, and then 31.Qc7+ instead of the only-move 31.Nh5+, which loses a point.

Move	Played	Your guess	Par	Points	Result
12	Bb2	f3	5	+5	Just as good
13	a4	Ng3	5	+5	Just as good
14	Qd3	Qd3	5	+5	Game move
15	Qc2	Qd2	5	+5	Just as good
16	Rae1	Rae1	5	+5	Game move
17	Ng3	Ng3	4	+4	Game move
18	f3	e4	5	0	Missed
19	e4	e4	4	+4	Game move
20	e5	Qf2	4	+2	Close
21	Qf2	Qf2	4	+4	Game move
22	f4	h4	5	+5	Just as good
23	exf6	exf6	4	+4	Game move
24	f5	h3	4	+1	Playable
25	Rxe1	Qxe1	6	−1	Missed a critical move
26	Re6	Re6	6	+6	Game move
27	fxe6	fxe6	6	+6	Game move
28	Qf4	Qf4	6	+6	Game move
29	Qe5	Qe5	4	+4	Game move
30	Ba3	Qc7+	5	0	Missed
31	Nh5+	Qc7+	6	−1	Missed a critical move
32	Qg5+	Qg5+	6	+6	Game move
33	Qxf6+	Qxf6+	6	+6	Game move
34	e7	e7	4	+4	Game move
35	Kf2	Kf2	6	+6	Game move
36	Kg3	Kg1	4	+4	Just as good
37	Kh4	Kh4	5	+5	Game move
38	Kxh5	Kxh5	6	+6	Game move
39	Kh4	Kh6	4	+4	Just as good
40	g4	g4	4	+4	Game move
41	Kh5	Kh5	4	+4	Game move

80/ 100

Points: 118 of 147 possible
Estimated strength: Expert (2000–2199)
Par for this game: 60 — finished 20 above par

Kasparov vs. Karpov, World Championship 1985, game 11

Playing White from move 14 · 12 moves guessed

A shorter, tougher game. Plenty of reasonable moves picked up partial credit, but the whole game turns on 23.Qxd7!!, the queen sacrifice. Guessing 23.Re8+ there costs a point, and letting the clock run out on the next move earns nothing. The score lands just below this game’s par.

Move	Played	Your guess	Par	Points	Result
14	Bg3	Bb3	5	+5	Just as good
15	Ba2	Qe2	5	+3	Close
16	d5	d5	4	+4	Game move
17	Nxd5	Bxd5	4	+2	Close
18	hxg3	fxg3	4	+2	Close
19	Bxd5	Qd3	4	+2	Close
20	Qa4	Rc2	5	+5	Just as good
21	Rcd1	Red1	4	+2	Close
22	Qg4	Qg4	5	+5	Game move
23	Qxd7	Re8+	6	−1	Missed a critical move
24	Re8+	—	6	0	Skipped (clock ran out)
25	Be4+	Be4+	6	+6	Game move

60/ 100

Points: 35 of 58 possible
Estimated strength: Class B (1600–1799)
Par for this game: 64 — finished 4 below par

Play Botvinnik vs. Capablanca Play Kasparov vs. Karpov

The short version

Each move has a par

Your first guess counts

Good alternatives earn credit

You get a score out of 100