Every ranked team carries a single number — its power rating. This page explains exactly how that number is calculated from a season’s results, what moves it, and what it can and cannot tell you.
A base of 1500, plus four parts calculated from the season so far
Rating = 1500 + Win % part + Margin part + Schedule part + Efficiency part| Part | Calculation | Typical size |
|---|---|---|
| Win % | 160 × (win % − 0.5) | −80 (winless) to +80 (unbeaten) |
| Margin | 3 × average margin per game | +60 for a team winning by 20 a game |
| Schedule | 40 × (opponents’ win % − 0.5) | −20 to +20 |
| Efficiency | 10 × (points for ÷ points against − 1) | Usually under ±5 |
Margin moves the rating most. Each point of average winning margin is worth 3 rating points, so a side winning by 20 a game gains about 60 — nearly as much as an unbeaten record (80). Margins are not capped, so a blowout counts in full.
Worked example. A team that is 6–1, averages 80–62, and has played opponents who win 55% of their games rates 1500 + 57 + 54 + 2 + 3 = 1616.
A draw counts as half a win. Ratings are rounded to a whole number and held between 800 and 3000.
Recalculated from the whole season, not adjusted game by game
When a round is marked complete, every team’s rating is worked out again from all of that season’s completed and forfeited games so far. Abandoned games do not count. The order the games were played in makes no difference.
Each completed round stores the rankings as they stood, which is what powers the rank-change arrows, form lines and biggest-movers views.
Strength of schedule is the average win percentage of the teams you have played, and is worth at most 20 points either way. Beating a strong side helps mainly through the win and the margin, not through a bonus for the opponent.
Upcoming fixtures show a win probability from the two teams’ ratings, using the standard formula 1 / (1 + 10^(−gap ÷ 400)). That formula was built for a different rating scale: here a 100-point gap can mean one team is winning by 30-plus points a game more than the other, yet it reads as only 64%. Treat it as a direction, not a forecast, especially for big mismatches.
Built for one competition’s season at a time
How competitions are ordered against each other — never part of a rating
Weight = Prestige × Tier × Level × AgeEvery competition we rank gets a weight from its classification. Competitions are listed heaviest first, and a new visitor’s homepage opens on the heaviest one that has rankings. Weight never feeds a team’s rating: each competition is rated only from its own games.
Every competition’s classification and weight is listed on the competitions page.
The rating is only as good as the results behind it
Results come from each competition’s official source. A result with no confirmed score is not imported, and 0–0 is refused unless the game was abandoned, because sources use 0–0 to mean “no score entered”.
Every rating is recalculated from the stored game record. There is no manual adjustment layer — the same results always produce the same ratings.
When a result is corrected, the rankings for the rounds it affects are regenerated from the corrected record rather than patched by hand.
How to read and interpret a power rating
1500 is a team that has won half its games, with an even points margin, against an average schedule. At the end of 2026 the top-rated sides in GPS, AIC and the CBSQ Opens sat between about 1610 and 1670. Compare ratings only within the same competition and season.
A ladder counts wins. A power rating also counts how much you won or lost by, who you played, and how efficiently you scored, so a team losing close games to strong opposition can rate above a team with a better record built on narrow wins over weak sides.
Because the rating is recalculated from the whole season. A narrow win can pull a team’s average margin down, their opponents’ results change their strength of schedule, and rivals may have improved more that round.
Not with these ratings. Each competition’s season is rated only from its own games, so there is no shared scale — a 1600 in one competition is not a 1600 in another. Comparing competitions needs games that link them, such as tournaments where schools from different leagues meet, and that is not part of the published ratings.
No. Every season is rated from its own games only, starting from the first one played.
When a round is marked complete, after all of its results are in. Each completion stores a snapshot for that round, which is what allows rank changes and rating progression to be tracked across a season.