How the ratings work

Most teams never play each other directly. NBR doesn’t need them to. We solve the entire web of games at once — a single statistical model finds the one set of ratings that best explains every result across the whole state. Strength flows through chains of common opponents: if your opponents beat the teams that beat other teams, that all feeds your rating. Tournaments are the bridges that connect everyone into one comparable pool.

A blowout says more than a one-run nail-biter, so bigger wins move the needle more — but with diminishing returns. Winning 12–2 counts well above winning 6–5; winning 25–0 counts barely more than 12–2. Running up the score doesn’t pad your rating.

Because every team is solved together from the same web of results, there’s no easy way to inflate a rating. Quality of opponent is what counts — beating weak teams barely moves you, and padding the schedule with easy wins can even lower your rating relative to teams testing themselves against tougher fields. Running up the score doesn’t help (margin is capped), and old results fade as recent games take over. The most reliable way up is simple: play good teams and beat them.

Ratings are recalculated regularly and weight recent games more heavily, so they move week to week as the season unfolds. A team that’s rolling climbs; a team that’s slipping falls. No one stays on top on reputation — you have to keep winning.

Teams with only a handful of games are shown as Provisional and treated cautiously (and kept out of the default ranking) until there’s enough data to be confident. Every team also carries a confidence level based on how much it has played.

Most clubs register a new team each year (e.g. 13U → 14U). When a team links its new season to its old one, its rating carries forward as a starting point rather than resetting — then quickly adjusts as the new season’s games come in.

A rating is only meaningful relative to a connected web of opponents. Early in a season, before tournaments link everyone together, separate groups of teams aren’t directly comparable — expect ratings to settle as the season’s results fill in.

Scores are compiled from publicly available game results and from games entered by our team. Ratings are refreshed on a weekly to semi-weekly basis.