Expected Goal Involvement Form

Five bars, one match each

Every bar is a single match. Its height is set by the expected goal involvements a player produced that day, expected goals plus expected assists added together. Five matches, five bars, read left to right in the order they were played.

That is the whole mechanic, and it matters more than it looks. A season average collapses five different performances into one number and throws away how they were distributed. Form keeps the shape. Two players can share an identical five match total and look nothing alike once the bars are next to each other. One rises match on match. The other is carried entirely by a single afternoon.

Three shapes, three different stories

Bars climbing left to right. A player worth 0.1 involvements two matches ago and 0.4 in the most recent one is getting into better positions or seeing more of the ball. It shows up here before anywhere else. Goals and assists lag this by design, since a chance has to be taken or created before it converts. A climbing shape is the earliest read available on a player heating up.

One tall bar, four flat ones. A single big match sitting well above the rest is a hot performance, not a change in level. Treat it as noise, not a trend. The following two or three matches settle it: if the bars build on that peak, the level has genuinely shifted. If they drop straight back down, it was one good day against one accommodating opponent.

Five bars at a steady, unremarkable height. Nothing here will catch the eye scanning for a spike, and that is the point. A consistently mid-height set of bars describes a dependable rate: the player is producing roughly the same threat every week. Unglamorous, but real, and often a more useful signal than a run with one huge outlier.

Read form before the goals and assists columns, not instead of them

Form exists to be checked first, not to replace what comes after it. It is a leading indicator: it moves before the actual output does, because chance quality changes before finishing catches up with it. A player climbing here is a name worth watching over the next few matches. A player flat here with a hot scoring streak is running ahead of their underlying level, and the two numbers should be read side by side.

Neither reading has the final say. The season total and the per 90 rate still carry more weight, because five matches is a small sample. Form flags the change. The larger numbers confirm whether it stuck.

The misreadings, corrected

"Five high bars means the player is now elite." Five matches is a form and role signal, not a settled level. Establishing a true rate takes hundreds of shots and chances across a full season, not one good week.

"Form bars replace the season average." They complement it. The bars show shape, the season figure and the combined xGI total show scale. Read both together rather than picking one.

"A flat run of bars means the player is out of form." It can equally mean a settled, reliable level of output. Low variance is not automatically bad. Judge the height of the bars first, and only read a flat run as a slump if that height has actually dropped from where it normally sits.