The rules, in the order you meet them

  1. Players are numbers. The director adds numbers from 1 to 999 to each age group. No name, birth date, photo, school or contact is asked for, and the director keeps the number-to-name sheet on paper.
  2. Evaluators score by link. Each evaluator gets a private link and taps a score from 1 to 5 (or 1 to 10) for each criterion. The standard criteria are Technique, Movement, Game sense, Effort and Coachability. An evaluator sees only their own scores, never another evaluator's, and never a rank.
  3. Unit scores. Each score becomes a number from 0 to 1: the lowest score is 0 and the highest is 1. On a 1 to 5 scale, a 4 is 0.75.
  4. Each evaluator's average for a player. The unit scores are averaged, with each criterion's weight (1 for every criterion unless the club sets its own). It counts only if the evaluator scored at least half the criteria, rounded up: 3 of the standard 5. A player with fewer is ignored for that evaluator, not counted as low.
  5. Evening out harsh and generous evaluators. Within an age group, each evaluator's averages are compared with that evaluator's own mean and spread (a standard score, sometimes called a z-score). Someone who gives everyone 4s and 5s and someone who gives 2s and 3s end up on the same footing. When an evaluator has fewer than 5 players or gives almost everyone the same score, the whole group's mean and spread are used for them instead, and the Board says so.
  6. The shown score. The evened-out values are averaged across the evaluators who scored the player, then shown as 50 plus 10 times that average, to one decimal. 50 is the average evaluator's middle; 60 is about one spread above it.
  7. Spread. How far apart the evaluators were on a player: 10 times the gap between the highest and lowest evened-out value. A spread of 15 or more is flagged in amber, because the evaluators saw that player very differently.
  8. Rank. Within the group, by shown score; ties go to the player seen by more evaluators, then the higher raw average, then the lower number. A player nobody has scored is listed below as Not scored. A withdrawn player is left out of every average and listed apart.
  9. Leaving an evaluator out. The director can leave one evaluator's scores out (or turn their link off and keep their scores). Leaving someone out recomputes every score in the group.
  10. The family report. Once scoring is closed, the director publishes the reports. A family sees, for their player only: the average score per criterion on the rubric's scale, how many evaluators scored the player, the club's next steps, and if the club chooses, the team and the rank. Never an evaluator's name or single score, the shown score, the spread, or any other player.
  11. Limits and keeping. Up to 2 tryouts at a time with 60 players and 6 evaluators each. A tryout becomes read-only after 30 quiet days past its date, and is deleted 365 days after its date (with 7 days to restore first), family reports included. Reopening scoring after a publish hides the reports until the next publish, and each report then says when scoring was reopened.

Evening out removes some of the difference between harsh and generous evaluators. It cannot remove bias, and it does not make a selection fair or right: the scores are the views of the club's evaluators on one day, and team decisions are the club's.

Worked example

Two evaluators, A and B, score five players on the standard 5 criteria, giving each player the same score on every criterion. A is generous with the top and harsh with the bottom; B scores everyone lower. A's averages are 1, 0.75, 0.5, 0.25 and 0, with a mean of 0.5 and a spread of 0.354, so A's evened-out values are 1.41, 0.71, 0, -0.71 and -1.41. B's averages are 0.5, 0.5, 0.25, 0 and 0, evened out to 1.12, 1.12, 0, -1.12 and -1.12.

The worked example
PlayerA gaveB gaveScoreSpreadRank
15362.73.01
24359.14.12
33250.00.03
42140.94.14
51137.33.05

Player 1's score is 50 + 10 × (1.41 + 1.12) / 2 = 62.7, and the spread is 10 × (1.41 - 1.12) = 3.0. Player 1's family sees each criterion as 4.0 out of 5 (the average of 5 and 3), scored by 2 evaluators.

What it is not