How accurate are the predictions?
Every prediction is saved before the deadline and checked against what players actually scored. Nothing is changed afterwards, and the misses are shown too.
This season
- Predictions missed by 2.1 points per player on average (players who played).
- The 10 players rated highest each week scored 2.5 on average, against 2.9 for everyone who played.
- Correlation between predicted and actual points: 0.22 (1 is perfect, 0 is guessing).
- The best £100m XI, picked before each deadline, was predicted 68 points over 1 gameweek and scored 54.
Gameweek by gameweek
| GW | Avg miss | Top 10 scored | Everyone | Best XI pred. | Best XI scored |
|---|---|---|---|---|---|
| GW6 | 2.1 | 2.5 | 2.9 | 67.8 | 54 |
“Avg miss” is the average gap between predicted and actual points for players who played. Open a gameweek for every player’s prediction and score.
Why predictions miss
Football is noisy: a goal is worth 4 to 6 points and one can decide a gameweek. Late team news, red cards and substitutions also move points after the deadline. What matters is that the players rated highest score more than the rest, week after week.
More
Questions
How accurate are FPL Checker’s predictions?
Over 1 gameweek this season, they missed by 2.1 points per player on average, and the 10 players rated highest each week scored 2.5 on average against 2.9 for everyone.
Are the predictions changed after the deadline?
No. They’re saved before each deadline and left as they were, so the comparison is fair.