See every player and multi-gameweek runs with projections across the season.
Fixture gain ranks at Spearman 0.072 over three games and 0.058 over five. That is why runs are graded Kind, Average or Tough and never ranked by a gain figure.
Expected output ranks at 0.52 against realised npxG, so it is the default sort. It is the one quantity here that orders players reliably.
Colour is projected xGI per fixture, and every cell prints that projection beside its unit. Nothing else on the board borrows these colours.
How accurate is this? See the methodology and track recordThis early in the season projections lean on last season. Trust the fixture grade more than the absolute figure while player data is thin: a player's own scoring rate largely cancels out of the fixture comparison, so it survives a small sample far better.
Players who appeared in none of the league's observed pre-season matches have a 40% readiness reduction applied to their gameweek 1 to 3 minutes: players who skipped the summer build-up have historically played well short of the minutes their record implied. Those players are labelled "No pre-season" under their name on the board. It is missing pre-season, not a late return, that has historically cost players minutes, so a player labelled "Tournament return" instead has had no projection changed. We can only assess clubs whose pre-season matches Wyscout recorded with a lineup, and a player at a club we could not assess never carries either label.
How accurate is this? See the methodology and track record
