PreSnap › How we measure
📐 How we measure
The math behind every number in PreSnap. Football is a numbers game: we want yours to be honest, consistent and useful for practice.
1. What we measure (and what we don’t)
We measure decisions inside the game: whether you picked the right answer, how long it took and in what situation. We do not measure talent, intelligence or fitness. A high rating means you read this game’s plays well; it does not replace a coach or the field.
2. What we record on every play
Every play stores, only on your device:
- right or wrong and your decision time;
- how many options there were (2, 3 or 4);
- the play’s difficulty (see section 4);
- the coverage the defense used;
- whether there was pressure, a disguise or a progression read;
- on coverage-reading plays, what you picked and what was right.
3. Read Rating (0 to 100)
The rating combines three parts, always using your last 60 plays:
- 65% · Adjusted accuracy: how many reads you get right, removing luck and giving more weight to hard plays.
- 20% · Speed: your time compared with the expected time for each type of play. It only counts fully once adjusted accuracy passes 60%, so fast guessing is not rewarded.
- 15% · Recognition: pre-snap reads and play calls, also removing luck.
4. Adjusted accuracy: removing luck and weighting difficulty
With 2 options, guessing is right 50% of the time; with 3, 33%. So adjusted accuracy is (accuracy − luck) ÷ (1 − luck). Example: 75% right with 3 options becomes (75 − 33) ÷ (100 − 33) = 62% adjusted accuracy. Always guessing scores 0%.
Harder plays count more in the average. The weight starts at 1.0 and grows like this (up to 2.0):
- +0.35 per extra read (open receiver is the 2nd or 3rd read);
- +0.30 if the defense disguises and rotates after the snap;
- +0.15 under pressure (blitz or rush);
- +0.15 under 11-on-11 rules;
- +0.20 on defense plays.
- With few plays, the score is pulled toward luck until there is enough data, so it does not overreact up or down.
5. Speed: expected time by type of play
A hot read against Cover 0 must come out fast; a progression to the 3rd read takes longer. So we compare your time with an expected time:
- 1.6 s base;
- +0.7 s per extra read;
- +0.6 s if there is a disguise;
- +0.4 s on read-the-defense plays;
- +0.3 s under 11-on-11 rules;
- −0.3 s under pressure (the ball must come out faster).
- Using 60% of the expected time or less scores the maximum; 180% or more scores zero. Only right answers count.
6. QB Brain
Six ratings from 0 to 99 show where you are strong and what to practice. Each appears only after a minimum number of plays:
- ⚡ Processing (6 right answers): speed compared with the expected time.
- 🧠 Recognition (6 answers): pre-snap reads and play calls, removing luck.
- 🔍 Progression (3 plays): adjusted accuracy when the open receiver is the 2nd or 3rd read.
- 🧊 Poise (5 plays): adjusted accuracy against blitz and rush.
- 🛡️ Ball security (10 game plays): every interception lowers it.
- 🧩 Disguise reading (3 plays): adjusted accuracy when the defense changes after the snap.
- Patience (12 plays): drops when you answer wrong in under 1 second, which is guessing.
- Consistency (8 right answers): whether your time, compared with the expected time, is steady from play to play.
7. Reliability and trends
Every rating shows how many plays it is based on. Up to 15 plays reliability is low, 15 to 40 is medium and above 40 is high.
Ratings vary naturally from day to day. So we only show ▲ or ▼ when a change is clearly bigger (1.5 times) than normal variation for the number of plays; otherwise you see “≈ steady”. When the arrow appears, the improvement is real.
8. Coverage grid
The grid shows your accuracy against each defense (Cover 0 to 4 in NFL style; man, zone, two-deep and blitz in flag). On coverage-reading plays it also shows what you confuse most, for example “Cover 3 read as Cover 1”. It is the most direct map of what to practice.
9. Scale and levels
- Career (Read Rating): Rookie up to 39 · Starter 40 · Pro Player 55 · All-Pro 70 · MVP 80 · Hall of Famer 90.
- QB Brain: Developing up to 49 · Solid 50 to 69 · Strong 70 to 84 · Elite 85 and up.
- Reference: about 75% right with 3 options at a normal pace lands in the Pro Player range; about 90% with fast decisions lands between All-Pro and MVP.
10. Limits
Ratings only measure decisions inside this game. They are not a medical, psychological, educational or cognitive assessment, and the scale is PreSnap’s own, not a scientific standard. Weights and expected times follow common quarterback-training principles and may be tuned as we learn from more data. All data stays on your device.