The maths, the principles, and the questions everyone asks.
Because everyone else’s results affect the strength estimate of teams you played earlier in the season. If a school you beat in March goes on a winning streak in April, their rating rises. The win you got over them in March is now worth a little more, because the system has more evidence they were a strong opponent. The reverse also happens. If a team you beat then loses badly to several others, the win is worth a little less.
We re-price every past match every time we publish, so a team’s rating reflects the most up-to-date picture of who they actually played. Every serious sports rating system in the world works this way. The maths is the same whether the result was on a rugby pitch or a cricket square.
Two things. First, when a higher-rated team beats a lower-rated team narrowly, the system pays out little or nothing. The win was expected. A one-point win against a team rated 20 points below you doesn’t tell us much we didn’t already know.
Second, as explained above, every other result this weekend may have changed the estimated strength of the teams you played earlier in the season. If those teams have, on balance, been revised downward since the last publish, your historical wins over them are now worth slightly less. That re-pricing can outweigh the small bump from this week’s narrow win.
It’s uncommon but expected. And it works the other way too. A team can lose a match this weekend and still see its rating rise, if the teams it has previously beaten have proven stronger than the system originally credited them with.
The fundamentals of the NGR engine are the same across every sport we cover. What changes is how “margin” is calculated and how the bonus threshold is tuned. Here’s how it adapts.
When two teams play, points move between them based on three things:
Team A (rated 0.700) beats Team B (rated 0.500). Same teams, same rating gap, every sport:
Final exchange in every case: Team A +0.0075, Team B −0.0075. Both ratings update, ready for the next round.
Benchmarked against six independent ranking sources across four complete seasons (2022–2025), outperforming ELO, Glicko, and Massey on school rugby data.
Feature by feature against established rating systems.
| Feature | NGR | ELO | Massey | Glicko |
|---|---|---|---|---|
| Historical programme prior | ✓ | ✗ | ✗ | ✗ |
| Opposition quality weighting | ✓ | ✓ | ✓ | ✗ |
| Margin bonus (capped) | ✓ | ✗ | ✗ | ✗ |
| School sport optimised | ✓ | ✗ | ✗ | ✗ |
| Per age group independent | ✓ | ✗ | ✗ | ✗ |
| Accuracy score (2022–25) | 173 | 142 | 171 | 121 |