Sprint Scoring

Every formula the game uses is published here. Nothing about scoring is hidden.

Why Score Against the Field

An absolute return target would reward whoever happened to be holding when a mega-pump fired. Scoring relative to everyone who traded the same hour measures something closer to skill: did you read that hour better than the people who lived through it with you?

It also means a calm hour and a violent hour are worth the same. You cannot farm points by waiting for volatility.

Step 1 — The Field

For each sprint, take the percentage equity change of every wallet that traded during it.

The field benchmark is the median of those returns — the middle value, not the average.

Median rather than mean so that one player catching a mega-pump cannot drag the bar up for everyone else.

If fewer than 5 wallets traded that sprint, the median is meaningless, so the benchmark falls back to 0% (break-even) and you are scored on whether you made money.

Step 2 — The Spread

Measure how spread out the field is, using median absolute deviation:

spread = median( |each return − benchmark| ) × 1.4826

The 1.4826 converts median absolute deviation to a standard-deviation-equivalent scale.

A floor applies:

spread = max(spread, 40)

Without a floor, a tightly clustered field divides by a near-zero spread and turns rounding noise into a perfect score.

Step 3 — Your Sprint Score

score = clamp( (your return − benchmark) ÷ spread , −3, +3 ) × 100

So a sprint score runs from −300 to +300, and 0 means you matched the field exactly.

The ±3 clamp is three standard deviations. Beyond that you are in mega-event territory, where the outcome is luck rather than skill, and the game stops paying more for more luck.

Step 4 — Round Points

Your round score is the shrunk mean of every sprint you were scored in:

n      = number of scored sprints, capped at 10
mean   = average of your sprint scores
points = mean × n ÷ (n + 3)

Three properties this gives you:

Every sprint counts. Not just your best ones. A bad hour genuinely costs you, so consistency beats one spectacular gamble.

Small samples are pulled toward zero. With 3 sprints your multiplier is 0.50; with 10 it is 0.77. One lucky sprint cannot outrank a sustained record.

Playing longer is not an edge. The sample caps at 10, so a 20-hour session earns no more credit than a 10-hour one. Three good hours can beat twenty mediocre ones.

Scored sprints	Multiplier
3	0.50
5	0.63
10	0.77
20	0.77 (capped)

Step 5 — Season Points

Season points are the sum of your round points across the season.

Because a median round scores near zero, simply showing up accumulates nothing. Only performance above the field accumulates — and below-field rounds subtract.

Worked Sprint

Seven wallets traded this hour, returning: −38%, −14%, +2%, +6%, +11%, +18%, +62%.

Benchmark = median = +6%
Distances from benchmark: 0, 4, 5, 12, 20, 44, 56 → median 12 → 12 × 1.4826 = 17.8
Below the floor of 40, so spread = 40

Player	Return	Difference	÷ spread	Score
A	+62%	+56pp	1.40σ	+140
B	+18%	+12pp	0.30σ	+30
C	+11%	+5pp	0.13σ	+13
D	+6%	0pp	0.00σ	0
E	+2%	−4pp	−0.10σ	−10
F	−14%	−20pp	−0.50σ	−50
G	−38%	−44pp	−1.10σ	−110

Note player E: +2% profit, but a negative score. They made money and still finished below the field.