power-rating

Compute the NCAA <> NIRA Division I Women's XVs PwrRt.

Usage

const { computePowerRating, data } = require('power-rating');
 
const result = computePowerRating(data);
console.log(result.data); // per-team stats + ratings, sorted as input

Data shape

{
  institute: DARTMOUTH,
  games: [
    { opponent: QUINNIPIAC, pointsFor: 59, pointsAllowed: 0 },
    { opponent: NAVY, pointsFor: 52, pointsAllowed: 7 },
  ],
}

Teams with no games (games: []) are excluded from the output.

Algorithm

1. Per-team stats: record (W/L/D), points for/against, and per-game rates. Points-for is capped at 65/game before averaging, so blowouts don't inflate a team's offensive rate.
2. Offense/defense ratings (AtkRt, DefRt): each team's capped points-for-per-game and points-allowed-per-game are z-scored against the league and re-centered on 0.5. Higher AtkRt = better offense; higher DefRt = better defense (fewer points allowed).
   • If the league has zero spread on either stat (every team identical), that rating defaults to 0.5 for everyone rather than dividing by zero.
3. Team rating (TmRt): AtkRt + DefRt.
4. Weighted win % (wW%): average, across all games, of the opponent's TmRt (win), TmRt - 1 (loss), or TmRt - 0.5 (draw). Beating a strong opponent counts for more than beating a weak one.
5. Power rating (PwrRt): (TmRt + wW%) / 2. The final output ranking.

Output

computePowerRating(teams) returns:

| Field | Description | |---|---| | data | Each team's stats and ratings (wins, losses, draws, pf, pa, pd, pfpg, papg, AtkRt, DefRt, TmRt, wW%, PwrRt) | | TmRt | Map of team name → raw team rating | | meanPfpg, standardDeviationPfpg | League-wide offense stats | | meanPapg, standardDeviationPapg | League-wide defense stats |