What xPts means.
Expected points are the mean of the model's simulated outcomes for a player over a selected Gameweek horizon. The model uses expected minutes, player rates, team strength and fixture context.
An average is useful. The range is the decision.
FPL expected points are a starting point, not a promise. FPL Prism puts the expected score beside the likely floor, median, ceiling and outcome bands so managers can see the trade-off.
Why two players with similar xPts differ.
Two players can average the same score while one has a much wider spread. The p10, median and p90 values show whether the average comes from a stable role or a boom-or-bust path.
- Bust probability: 0–2 points
- Middle bands: 3–5 and 6–9 points
- Haul probability: 10+ points
Sharpe-style scores add context.
The Sharpe-style score compares expected points with simulated volatility. It can highlight a steadier pick when raw xPts alone would favor a more variable player.
Forecasts are checked against results.
The Modelbook archives frozen Gameweek projections and official FPL points. It reports mean absolute error, RMSE, signed bias and the share of active players within two points.
Useful answers before the deadline.
Are FPL expected points the same as actual points?
No. Expected points are a forecast before the matches. Actual points are only known after the fixtures are played.
What are p10 and p90 in FPL projections?
They are the 10th and 90th percentile outcomes in the simulation. Roughly speaking, most simulated results fall between them, although no range is a guarantee.
Why can a low-xPts player still be useful?
A player with a tighter range, strong minutes confidence or a better squad fit can be preferable to a higher-xPts player with more downside or concentration risk.
Where does the FPL Prism data come from?
The model uses public FPL player, fixture and event feeds, with historical player priors and transparent validation in the Modelbook.