Sudoku·4 min read·May 24, 2026

How Melio Sudoku scoring works

Every solo and daily sudoku on Melio gets a numeric score at the end. The leaderboards rank by it. This is exactly how it’s computed.

The formula

score = round( base × timeFactor × mistakeFactor × hintFactor )

Four multiplicative factors, all ≥ 0. The result is always a non-negative integer. Each factor is described below.

The base (difficulty)

The base number is how much a perfect-pace, no-mistake, no-hint solve at that difficulty is worth. Higher difficulty = higher base.

Easy: 100
Medium: 250
Hard: 500
Expert: 1,000
Extreme: 2,000

These are the “mid-pace” scores — what you get for solving in the target time with no mistakes and no hints. They roughly double per difficulty step, which roughly matches how the difficulty actually scales.

Time factor (the big one)

Time factor = targetTime / yourTime, clamped between 0.3 and 3.0.

The target times per difficulty are:

Easy: 5 minutes
Medium: 10 minutes
Hard: 20 minutes
Expert: 30 minutes
Extreme: 50 minutes

Solve in half the target → ~2× multiplier. Take twice as long → ~0.5×. The 0.3-3.0 clamp keeps the formula sane on either extreme: nobody’s solving an Easy in 30 seconds (3× max), and you don’t get negative points for taking an hour on an Expert (0.3× min).

The time factor is the biggest lever. A clean, technique-fluent solver on Hard (~500 target) in 8 minutes scores 500 × 2.5 = 1,250. The same Hard in 30 minutes scores 500 × 0.66 = 330. Same accuracy, very different number.

Mistake factor

Mistake factor = max(0.1, 1 - mistakes × 0.1).

Each mistake costs you 10%. The floor is 10% — so 9 mistakes is the worst it gets; beyond that there’s no additional penalty.

0 mistakes → 1.0×
1 mistake → 0.9×
3 mistakes → 0.7×
5 mistakes → 0.5×
9+ mistakes → 0.1×

Note: in the daily, the puzzle ends after 3 mistakes, so the daily mistake factor never goes below 0.7×. In solo mode there’s no cap on mistakes, so the floor can be reached.

Hint factor

Hint factor = max(0.25, 1 - hints × 0.25).

Each hint costs you 25%. The floor is 25%. With the 3-hint cap per puzzle, the worst case is using all 3:

0 hints → 1.0×
1 hint → 0.75×
2 hints → 0.50×
3 hints → 0.25×

Hints are intentionally expensive. They’re a learning aid more than a casual help — designed to be used when you’re genuinely stuck and want to see the move, not to speed-run a clean solve.

Worked examples

Expert in 18 minutes, 1 mistake, 0 hints base = 1000 timeFactor = 30min / 18min = 1.67 mistakeFactor = 0.9 hintFactor = 1.0 score = round(1000 × 1.67 × 0.9 × 1.0) = 1,503 Expert in 18 minutes, 0 mistakes, 1 hint score = round(1000 × 1.67 × 1.0 × 0.75) = 1,253 Expert in 10 minutes (fast), 0 mistakes, 0 hints score = round(1000 × 3.0 × 1.0 × 1.0) = 3,000 ← clamped at 3.0 Easy in 90 seconds, 0 mistakes score = round(100 × 3.0 × 1.0 × 1.0) = 300 ← also clamped

Pushing a score higher

Three knobs, ordered by impact:

Solve faster. Time is the biggest multiplier (up to 3×). Practiced technique (cross- hatching, pointing pairs, X-wing) cuts solve time more than anything else.
Don’t use hints.Each hint cuts 25%, which is brutal. If you’re grinding for leaderboard rank, take the time to find the move instead.
Slow down on mistakes.Each mistake is “only” 10%, but two mistakes is 19% gone. On Expert that’s 200+ points. Pencil marks reduce mistakes more than caution does.

Why this formula?

We wanted something that:

Rewards both raw speed and difficulty (so an Expert solver and a fast Easy solver can’t directly compare).
Has a clear upside (3× time multiplier) so racing friends feels meaningful.
Has a clear downside (0.1× floor on mistakes) but doesn’t turn into zero or negative — every completion is worth something.
Is simple enough to explain in a guide like this one, so leaderboard placement feels earned, not arbitrary.

Keep reading

Sudoku5 min

Sudoku XYZ-wing: the trivalue pivot variant of Y-wing

Y-wing's slightly bigger cousin. A trivalue pivot {x,y,z} plus two bivalue wings {x,z} and {y,z}, all sharing the digit z. Any cell that sees all three loses z as a candidate. Worked example, when it appears.

Sudoku7 min

Sudoku unique rectangle: the uniqueness-based elimination

Four cells in a rectangle across two rows, two columns, and two boxes, all sharing the same two candidates. If left alone, the puzzle would have two solutions — but a well-formed sudoku has exactly one. Use that to force an elimination.

Sudoku6 min

Sudoku naked pairs and triples: the cell-set elimination

When two cells in a unit share the same two candidates, those two digits are locked into that pair. Everywhere else in the unit can drop both digits. The most-used intermediate technique after pointing pairs.

Sudoku5 min

Sudoku skyscraper: the cousin of X-wing for short-strong patterns

The skyscraper sits between X-wing and the harder fish patterns. Two strong links on a single digit, ending in cells of different heights — and the elimination that falls out. Worked example included.

Browse all guides →

The base (difficulty)

The base number is how much a perfect-pace, no-mistake, no-hint solve at that difficulty is worth. Higher difficulty = higher base.

Easy: 100
Medium: 250
Hard: 500
Expert: 1,000
Extreme: 2,000

Time factor (the big one)

Time factor = targetTime / yourTime, clamped between 0.3 and 3.0.

The target times per difficulty are:

Easy: 5 minutes
Medium: 10 minutes
Hard: 20 minutes
Expert: 30 minutes
Extreme: 50 minutes

Mistake factor

Mistake factor = max(0.1, 1 - mistakes × 0.1).

Each mistake costs you 10%. The floor is 10% — so 9 mistakes is the worst it gets; beyond that there’s no additional penalty.

0 mistakes → 1.0×
1 mistake → 0.9×
3 mistakes → 0.7×
5 mistakes → 0.5×
9+ mistakes → 0.1×

Note: in the daily, the puzzle ends after 3 mistakes, so the daily mistake factor never goes below 0.7×. In solo mode there’s no cap on mistakes, so the floor can be reached.

Hint factor

Hint factor = max(0.25, 1 - hints × 0.25).

Each hint costs you 25%. The floor is 25%. With the 3-hint cap per puzzle, the worst case is using all 3:

0 hints → 1.0×
1 hint → 0.75×
2 hints → 0.50×
3 hints → 0.25×

Hints are intentionally expensive. They’re a learning aid more than a casual help — designed to be used when you’re genuinely stuck and want to see the move, not to speed-run a clean solve.

Worked examples

Pushing a score higher

Three knobs, ordered by impact:

Solve faster. Time is the biggest multiplier (up to 3×). Practiced technique (cross- hatching, pointing pairs, X-wing) cuts solve time more than anything else.
Don’t use hints.Each hint cuts 25%, which is brutal. If you’re grinding for leaderboard rank, take the time to find the move instead.
Slow down on mistakes.Each mistake is “only” 10%, but two mistakes is 19% gone. On Expert that’s 200+ points. Pencil marks reduce mistakes more than caution does.

Why this formula?

We wanted something that:

Rewards both raw speed and difficulty (so an Expert solver and a fast Easy solver can’t directly compare).
Has a clear upside (3× time multiplier) so racing friends feels meaningful.
Has a clear downside (0.1× floor on mistakes) but doesn’t turn into zero or negative — every completion is worth something.
Is simple enough to explain in a guide like this one, so leaderboard placement feels earned, not arbitrary.