Skip to content
GamePrep
Benchmarks · from our own sessions

What a good score actually looks like

Not an opinion about what counts as good — the distribution of what people really scored, from 1,601 sessions scored on our server. Every row shows how many people it is based on.

Recomputed hourly · 148 players · 16 games with enough data

How to read this, and what it is not

These are first attempts. Each player counts once, on the first time they played that game — before any practice. A distribution over every session would describe our most dedicated regulars rather than a person sitting down cold, and one player here has played a single game seventy times.

This is not a candidate population. These are people who chose to practise for an assessment, which is already a group that differs from everyone who sits one. No employer publishes its cut scores, and nothing here is one — see how we score for what a percentile here does and does not mean.

A zero is a real result. On a memory game it means the first sequence was missed; the bottom decile therefore mixes genuinely low scores with attempts somebody abandoned after one round. We leave them in, because dropping the worst results would quietly make every median look better.

Rows under 20 players say so rather than printing a median that one more person would move, and a practice figure needs 10 players before it appears at all.

Score distribution by game

Read the middle column as the typical first attempt: half of first-timers did better, half worse. Columns run weakest to strongest, so on reaction time — where a lower number is faster — the strongest column holds the smallest millisecond figure.

Game Weakest 10% Lower quarter Median Upper quarter Strongest 10% Based on
Digit Span Memory Test Practice digits recalled 0 5 6 7 9 79 players
205 sessions
Reaction Time Test Practice milliseconds · lower is better 918 758 654 418 357 65 players
207 sessions
Mental Math Test Practice (Numerical Reasoning) correct answers 8 13 16 18 19 56 players
115 sessions
Stroop Test Practice focus score out of 100 41 56 60 69 74 55 players
150 sessions
Mental Rotation Test Practice (Shapedance) rotations spotted 0 2 5 9 10 49 players
105 sessions
Flanker Test Practice (Arrows Game) attention score out of 100 60 69 82 88 92 47 players
96 sessions
Numerosity Game Practice levels solved 4 6 8 10 11 46 players
138 sessions
Pathfinder Game Practice (Path Puzzle) puzzles solved 3 5 5 5 5 44 players
77 sessions
Visual Memory Test Practice (Corsi Block) blocks recalled in order 4 4 5 6 7 39 players
89 sessions
Singularity Game Practice (Odd One Out) search score out of 100 60 70 77 84 86 37 players
61 sessions
Disco Numbers Game Practice speed score out of 100 64 70 76 82 86 37 players
91 sessions
Backward Digit Span Test Practice digits recalled backwards 0 4 5 7 8 33 players
49 sessions
N-Back Test Practice (2-Back Letters) memory score out of 100 0 0 13 63 75 33 players
80 sessions
Inductive Reasoning Test Practice (Next in the Sequence) reasoning score out of 100 40 51 64 73 86 33 players
46 sessions
Task-Switching Test Practice (cut-e Switch Challenge) flexibility score out of 100 55 56 60 71 75 30 players
69 sessions
Numerical Reasoning Test Practice (Data Interpretation) numerical score out of 100 59 63 78 89 93 20 players
23 sessions

Does practice actually move the number?

For players who played the same game at least 3 times: how far their best run ended up above their first, in percentile points, and how many attempts it took to get there.

Reaction Time Test Practice

+10 percentiles

median gain, first run to best · 35 players · best run typically on attempt 2

Digit Span Memory Test Practice

+15 percentiles

median gain, first run to best · 25 players · best run typically on attempt 3

Stroop Test Practice

+30 percentiles

median gain, first run to best · 19 players · best run typically on attempt 3

Numerosity Game Practice

+12 percentiles

median gain, first run to best · 17 players · best run typically on attempt 3

Visual Memory Test Practice (Corsi Block)

+20 percentiles

median gain, first run to best · 15 players · best run typically on attempt 2

Mental Math Test Practice (Numerical Reasoning)

+25 percentiles

median gain, first run to best · 14 players · best run typically on attempt 2

Mental Rotation Test Practice (Shapedance)

+35 percentiles

median gain, first run to best · 14 players · best run typically on attempt 3

Flanker Test Practice (Arrows Game)

+9 percentiles

median gain, first run to best · 13 players · best run typically on attempt 2

Disco Numbers Game Practice

+10 percentiles

median gain, first run to best · 12 players · best run typically on attempt 2

Some of that gain is real skill and some is simply knowing the interface the second time — the two cannot be separated from this data, and anyone claiming otherwise about their own numbers is guessing too. It is still the gap between a cold first attempt and a warmed-up one, which is exactly the gap an assessment catches you in.

Where would your first attempt land?

Play any game free, no account, and your score is added to the same table.

Try a game