🎯 Rating Check

Why this exists

DUPR ratings move slowly, and plenty of players are better (or worse) than their number says. Clubs handle this with coach-assigned club ratings — which puts coaches in the awkward spot of judging players by eye and delivering the bad news personally. Rating Check replaces the eyeball with evidence: let the player play rated competition, record the scores, and the games themselves say what level they hold. The coach hands over a measurement, not an opinion.

How it works

It's the Upset Meter's club-calibrated forecast run backwards. The forecast says how a game should score for a given rating gap; Rating Check observes how the game actually scored and solves for the one unknown — the assessed player's level. Each game contributes a reading (blowouts are capped, since an 11–0 only proves "far apart", not how far); the readings average into a performance band. Six games is the practical minimum for a steady answer; ten is better.

The receipts

Before trusting it with anyone's club rating, we gave it a blind test on ourselves. We took 64 club players whose DUPR we already knew, covered up the rating, and made the tool guess it from their game scores alone — 404 real matches. Then we peeked at how close each guess came.

Even the rough guesses behaved: 9 out of 10 landed within about three-quarters of a rating point, and the rare bigger misses were mostly players whose games were all blowouts — an 11–1 proves "mismatched," not by how much, so there was nothing to measure.

Think of it as a scale that weighs to the nearest pound, not the nearest ounce: exactly right for "which bracket does this player belong in," and honest enough to say "borderline" when the question is finer than six games can answer. The band the calculator shows is sized so the player's real level lands inside it about 4 times in 5 — and since the DUPRs we tested against are themselves imperfect, true accuracy is likely a touch better than these numbers.

Running a fair assessment

• 6+ games to 11, all against rated players, everyone playing to win. The model is calibrated on 11-point games.
• Rotate partners. Six games with the same partner rates the pair, not the player. A different rated partner each game isolates the player's own contribution — a rotating round-robin does this automatically.
• Competition near their level. If every game is a blowout in either direction, the honest answer is "outside this bracket's range" — the calculator will tell you.
• Read the band, not the center. The result is what these games showed, stated with honest uncertainty. It's a performance snapshot for club decisions — not an official rating from anyone.

A rating lives in its pool

The band describes performance against the specific pool it was measured in. Games from a women's league produce a women's-playrating — exactly the right number for women's-league decisions, and no promise at all about mixed or men's play. Players transfer between pools differently (power, pace, and agility don't move as one number), so resist applying a one-size adjustment. The honest move is to measure in the other pool: six mixed games produce a mixed rating the same way. Set "measured in" on the calculator so the result card says what it actually measured.

The calculator

One row per game. Ratings are DUPRs of the other rated players — the player being assessed never enters a number. Leave partner and the second opponent blank for a singles game.

Enter at least 4 complete games to get a read (0 so far). Six or more gives a much steadier answer.