How we project
Points per minute × expected minutes
Every projection on Rugby Per Minute comes from two numbers: how many fantasy points a player scores for each minute he's on the field, and how many minutes we expect him to play. This page explains both, and how well the method has done when we tested it.
Rate×Minutes±Opposition
The opposition step is small and switches on once the 2027 draw is out.
Why per minute
A points-per-game average mixes two different things: how good a player is when he's on, and how long he stays on. Props usually come off around the 54th minute. Outside backs often play the full game. A prop and a winger with the same average are not the same pick.
Splitting the two lets us handle the part that changes every week, minutes, on its own. A player's rate is fairly stable. His minutes are not.
Average minutes, Super Rugby 2022 to 2026.
| Position | Starting | Off the bench |
|---|
| Prop | 54 | 27 |
| Hooker | 57 | 24 |
| Lock | 69 | 25 |
| Back row | 72 | 24 |
| Halfback | 63 | 20 |
| First five | 73 | 20 |
| Centre | 73 | 22 |
| Outside back | 75 | 22 |
The rate
Early in a season, a player's own numbers are too thin to trust. Three good games can make anyone look like a star. So every player starts the season from a starting point, and his own 2027 numbers take over as the minutes build up.
- Played Super Rugby last season: his rate from that season.
- Stepping up from the NPC: his NPC rate, adjusted to Super Rugby. In our testing, a player's NPC rate predicted his Super Rugby rate almost as well as a full Super Rugby season did.
- Back from overseas: his rate in that league, adjusted for how that league scores at his position.
- Straight out of the under-20s: his rate in under-20 internationals, adjusted to Super Rugby and pulled hard toward the position average. In our testing it was a weaker guide than the NPC, but still better than the average alone.
- None of those: the average for his position.
Every starting point is also pulled part of the way toward the position average, so one hot season on a small sample doesn't carry straight into his projection. Props and locks are pulled much harder when the starting point is their NPC rate, because for tight forwards the NPC barely carries over to Super Rugby.
Goal kicking is part of the rate. In the official game, kicking points go to whoever kicks, so we count them. Our Teams page flags who we expect to have the tee: in our testing, a starting goal kicker kept it the next week 84% of the time.
The minutes
Before team lists are out, minutes come from our Best 23 for each team. A player we have starting gets the usual starting minutes for his position, a bench player gets bench minutes, and anyone outside the 23 gets none.
This is the biggest source of error, and there's no way around it before teams are named. In our testing, about 3 in 10 starters didn't start the following week. So the moment the official team lists come out, the site switches to them and every projection updates.
Opposition
Once the draw is loaded, we adjust each projection a little for the opponent, based on how many fantasy points that team has given up. We tested breaking this down by position (does this team leak points to hookers?) and it did no better than the team-level number, so we keep it simple. The Fixtures page shows the full picture.
What we leave out: form
We don't project from recent form. We tested it: last-three-games averages were the worst predictor we tried, and adding some form on top of the main method didn't help. Rugby fantasy scores swing a lot from week to week, and a hot streak mostly isn't a signal. You'll still see hot and cold flags on the Players page, as context.
How good is it
We tested the method on five Super Rugby seasons, 2022 to 2026: 18,580 player games, each one predicted using only what was known before it was played.
- Starting from last season's rate, rather than this season's numbers alone, cut the average miss by almost a point a game and ranked starters about 20% better. It was the biggest single improvement we found.
- Even so, the average miss is around 9 points a game. Fantasy rugby is noisy: a try, a yellow card or an early substitution moves a score a long way.
- Knowing the actual minutes in advance would have improved the ranking of starters far more than anything else we tried. That's why team-list day matters so much, and why the site updates as soon as lists are named.
Treat projections as a well-tested starting point for your own judgement, not a guarantee.
Prices and scoring
Opening prices for 2027 are our estimates until the game opens, based on each player's average last season. We worked out the official game's price rule from 2026: our version gets 94% of price changes within $100k, which is how the breakevens are calculated.
The official game's scores only cover the 2026 season. For earlier Super Rugby seasons, the NPC and overseas leagues, we score games with our own engine, built to match the official scoring. It lands within about 4 points a game of the official score.