GAME IMPROVEMENT AUDIT · Sugar Drop
A year-old balancing problem found in one audit.
The audit found severe difficulty walls, ranked the levels that needed attention, and gave the team a specific change for each one.
40 days of telemetry across 100 levels.
THE MAIN MOVEMENTS
Before and after, shown side by side.
Attempt-weighted win rate
42.12-point gapLevels audited
Full gameLevels at least 20 points below target
PrioritisedLowest win rates found
Three severe wallsA CLOSER LOOK AT THE DATA
The difficulty curve across all 100 levels.
The full curve revealed repeated difficulty walls across the game.
Actual win rate versus target
The orange line shows where each level should be. The blue line shows what players actually faced. It jumps all over the place, and a few levels drop to a 1-2% win rate.
THE NUMBERS BEHIND THE CHART
A sample of the levels the audit flagged
The gap compares the observed win rate with the target for that position in the 100-level curve.
| Level | Win rate | Target | Gap |
|---|---|---|---|
| Level 6 | 59.79% | 96.60% | −36.81 points |
| Level 9 | 23.33% | 95.40% | −72.07 points |
| Level 38 | 16.22% | 85.20% | −68.98 points |
| Level 47 | 4.55% | 82.80% | −78.25 points |
| Level 88 | 1.48% | 74.40% | −72.92 points |
| Level 90 | 1.52% | 74.00% | −72.48 points |
| Level 94 | 2.05% | 73.20% | −71.15 points |
Forty days of telemetry from 1 July–10 August 2026 across all 100 levels. The audit used this evidence to set the priority and exact change for each problem level.
What was happening
For an early game, the 44.24% win rate was far too low, and the difficulty jumped from level to level. Three late walls had win rates between 1.48% and 2.05%, creating repeated retry points that almost nobody cleared.
What the data showed
The audit found 54 levels at least 20 percentage points below their individual targets. Of those, 25 were more than 50 points below target. The repeated spikes across the curve called for a level-by-level plan.
What we did
We audited all 100 levels, ranked the worst walls, and wrote what each priority level needed: more moves or time, a layout change, a token-target change, or focused validation before changing anything.
A REAL PART OF THE WORK
Example level instructions
| Level | Objective and layout instruction | Moves | Token 2 | Token 3 | Validation |
|---|---|---|---|---|---|
| Level 6 | 8 wrappers; open the corner and centre routes | 25 → 29 | 1,650 | 2,200 | High-priority playtest, win rate, and moves left |
| Level 9 | 7 wrappers; open the lower section and objective routes | 20 → 25 | 1,575 | 2,100 | Critical playtest, failures, and objective progress |
| Level 10 | 10 wrappers; reposition awkward objectives and protect refill paths | 18 → 21 | 1,250 | 1,850 | High-priority playtest, win rate, and token attainment |