xG and xGI

Goals are rare and lumpy, so a handful of matches tells you very little about whether a player is genuinely creating danger. Chance quality accumulates faster than goals do, which makes xGI a steadier read on a player's role in his team than his points column over the same period.

Fantasy Premier League publishes expected goals, expected assists and expected goal involvements per player in the official data, so this is not a private metric any more: it sits next to price, form and ownership.

A player scoring far above his xG is finishing well or getting lucky, and the two are hard to separate in a small sample. A player far below it is getting into the right positions and missing, which historically corrects. Neither is a rule, and elite finishers do beat their xG over long periods, but a wide gap is a reason to check the fixtures and the minutes rather than to buy or sell on instinct.