THE SCIENCE BEHIND THE ASSESSMENT

Every Test is scored against handicap-banded thresholds (scratch, 5, 10, 15, 20, 25+). Those thresholds were built from published shot-outcome distributions for each band and refined with coach-run Tests at Skill IQ academies.

In progressSample size, handicap range tested, age/gender split, and number of facilities in the validation cohort: [TBD — pending validation study]

Reliability answers a simple question: if you take the same Test twice in a week without changing anything, how much does the number move? Until that study reports, every reliability band you see in the app (± strokes, "low / moderate / high") is a statistical estimate from your own shots — a Wilson interval per category rolled into a standard error on the test score, widened when few sessions are on file. It is not a published reliability coefficient.

In progressIntraclass correlation coefficient (ICC), standard error of measurement, and minimal detectable change: [TBD — pending validation study]

A test score should track handicap index closely enough to be a useful benchmark, while still separating two golfers with the same handicap and opposite skill profiles.

In progressPearson r between test score and handicap index, with confidence interval and n: [TBD — pending validation study]

The Transfer Gap chart already compares your test score to your logged on-course rounds player-by-player. The population-level version of that relationship is what turns "you gained 3 strokes on the Test" into "you should expect to shoot roughly this much lower".

In progressRegression of test score on 20-round scoring average, R², and standard error of the projection: [TBD — pending validation study]

Structure. Full Test = 72 shots: 12 driving, 18 approach, 18 short game, 24 putting. Scoring Zone Test = 36 shots: 12 short game, 24 putting, scored on a 36 baseline.

Scoring. Test Score = baseline + strokes lost − strokes gained. Point values per outcome are fixed for every player; only the thresholds (carry distance, fairway width, proximity in feet, putt distance bands) scale by handicap band. That is what makes a 15-handicap's "good approach" different from a scratch player's while keeping the score on one common scale.

Measurement. Driving and approach require measured carry and proximity, so only a certified coach can run them. Short game and putting are scored by where the ball finishes, which a golfer can judge unaided — that is why the solo test stops at the scoring zone.

Ability vs. performance. Today's Test Score is one session's result. Estimated Ability is an exponentially weighted average of your sessions (60-day half-life), so a single bad putting day moves ability far less than it moves performance.

Level boundaries are a coaching convention for talking about progress, not a validated classification. Cut points will be re-fit to the population once the validation study reports.

We do not claim the Test predicts tournament results, that it replaces launch-monitor fitting, or that a handicap-matched benchmark captures every way two golfers can differ. Outcome-based cohorts ("golfers who score in your range") are on the roadmap and are labelled as such in the app.