What held up
Bookmaker-derived clean sheet probabilities beat the official FDR at predicting defensive returns in 31 of 38 gameweeks. Set-piece takers outscored their xG models by a consistent margin. And bench boosts played on double gameweeks averaged 21 points against 9 for those played on singles.
What didn't
Form is shorter-lived than anyone wants to believe. A five-gameweek hot streak added almost nothing to a player's next-gameweek projection once fixtures were accounted for. 'Form vs fixtures' isn't a debate — fixtures win, and it isn't close.
Early wildcards also underperformed: managers who held until after the first international break gained an average of 14 points over those who pulled the trigger in GW2–4.
What we changed in the engine
This season's model weights expected minutes more aggressively, discounts short-term form harder, and blends rest-day and travel fatigue into the fixture difficulty split. Every one of those changes came directly from the review above.
Adam FPL
Guest Analyst
Written by the FPL King editorial network - tactical FPL analysis, captaincy edges, transfer strategy and fixture-led planning.
FPL manager chat
Discussion
Loading discussion…