Is a 0.5 xPts difference worth a transfer?

Not on its own over a single gameweek. A four-point hit needs roughly 4 points of expected gain across the horizon you plan to hold the player, so look at the multi-gameweek sum, not this week's line.

Why is my captain's xPts not doubled in the table?

Most tables show the base projection. Captaincy doubles both the mean and the spread, which is precisely why the armband should be chosen on floor or ceiling rather than on the doubled average.

Concepts · 7 min read

What expected points (xPts) really means in FPL

xPts is the single most misused number in Fantasy Premier League. Understanding what it is — and what it hides — is the difference between using a model and being used by one.

Updated 16 August 2026

The definition

Expected points, written xPts, is the average number of FPL points a player would score across many repetitions of the same gameweek, given his chance of playing, his underlying rates, his fixture and the FPL scoring rules. A 5.0 xPts player is not predicted to score five. He is predicted to average five if the gameweek were replayed a thousand times.

That distinction matters because FPL is played once. You do not collect the average; you collect one sample from the distribution.

What the average hides

Consider two midfielders both projected at 5.2 xPts. The first plays 90 minutes every week for a defensively solid side, picks up regular bonus and rarely blanks below 2. The second is a rotation risk who returns 1 point most weeks and 16 occasionally.

Their means are identical, their usefulness is not. The first protects a lead; the second is how you claw back a 40-point deficit. Reporting only the mean makes them look interchangeable.

Floor, ceiling and standard deviation

A simulation-based model can report the whole distribution. The 10th percentile — the floor — is roughly a bad-but-not-catastrophic week. The 90th percentile — the ceiling — is a good week without being a freak event. The standard deviation summarises how wide the whole thing is.

Read them together with the mean. Floor for safety, ceiling for aggression, and the gap between them as a measure of how much of a coin flip the pick is.

xPts over multiple gameweeks

Summing xPts across the next four or six gameweeks is the right way to compare transfer targets, because a single fixture is dominated by noise. But sum the means only — you cannot add floors and ceilings, since a bad week and a good week partially cancel over a run.

Also discount the far horizon. A projection six gameweeks out carries unknown injuries, rotation and form drift. Weighting the nearer gameweeks more heavily reflects that honestly.

xPts per million

Dividing expected points by price is useful for the middle of your squad, where budget is genuinely constrained, and misleading at the top, where a premium's job is to produce ceiling rather than efficiency. Value metrics systematically favour cheap defenders who will never win you a mini-league.

Use points-per-million to find the enablers that fund a premium, then pick the premium on ceiling.

Common misreadings

Treating a 0.3 xPts gap between two players as meaningful — it is well inside model error
Comparing xPts across tools with different scoring assumptions, especially around bonus
Ignoring the minutes probability attached to the projection
Taking a hit for a 1.5 xPts upgrade, which is a losing trade in expectation over one week
Assuming a high xPts means a high chance of a haul; often it means a high chance of a solid, unspectacular return

Frequently asked

Is a 0.5 xPts difference worth a transfer?: Not on its own over a single gameweek. A four-point hit needs roughly 4 points of expected gain across the horizon you plan to hold the player, so look at the multi-gameweek sum, not this week's line.
Why is my captain's xPts not doubled in the table?: Most tables show the base projection. Captaincy doubles both the mean and the spread, which is precisely why the armband should be chosen on floor or ceiling rather than on the doubled average.