Power Loss %: how SquadCheck prices an absence

The list of absences comes from the Premier League’s official Fantasy Premier League feed, which carries a status for every registered player. SquadCheck reads three of those statuses:

• Injured — ruled out.
• Doubtful — the feed gives a chance of playing, such as 75% or 50%.
• Suspended — banned for one or more matches.

Two other statuses are ignored: players who have left the club or gone out on loan, and players who are simply not in the squad. Neither is an absence the club has to plan around. The feed is re-read twice a day, and every half hour around matches. Each absence is attached to the club’s next fixture, and once that match kicks off the record becomes history rather than news — see how return dates are read.

Every player in the squad gets a weight between 0 and 1, built from three ingredients:

The output ingredient was designed for a richer data feed than the one the site reads today. It has room for key passes, dribbles, duels won and shots, and the free official feed carries none of them, so those parts count as zero. In practice a defender’s output is his tackles plus clearances, blocks and interceptions per 90 minutes, and a midfielder’s or forward’s is his goals and assists per 90. The feed has no match ratings either, so every player gets the same neutral value there. The consequence is worth stating plainly: a holding midfielder who rarely scores is valued mostly on his minutes.

Early in a season a player may have started only once or twice, so the weight looks back:

• Five or more starts this season: this season only.
• One to four starts: this season, blended with his previous season at the club.
• No starts yet: his previous season at the club, discounted by a quarter — failing that, the season before, discounted further.

Only matches for his current club count. A player’s record elsewhere is not carried across, which matters for new signings (see the last section).

Power Loss is the weights of the unavailable players added together, divided by the weights of the whole squad, times 100. Nothing is capped or smoothed. Two consequences follow, and both matter when you read the number:

• A doubtful player counts in full. A player given a 75% chance of playing adds his whole weight, the same as one ruled out for months. Power Loss measures how much of the squad is on the list, not how likely the list is to be right on the day.
• A suspension counts the same as an injury. For the match in question, it leaves the same hole.

On the team page the gauge turns from green to amber at 8% and to red at 15%. Those bands are there for reading at a glance; they have not been tested against results.

Each absent player also carries a label — Critical, High, Moderate or Low. It comes from an impact score: the player’s weight, multiplied by two adjustments.

• Role. A player who has started at least 60% of the club’s completed matches this season is a regular starter (×1.0). Between 30% and 60% he is a rotation player (×0.7). Below that, a squad player (×0.4).
• Timing. An absence in the middle of the season, from a player who has been part of the side, counts in full. Early-season absences, recent injuries and long absences count a little less (×0.75 to ×0.9). A player who has not appeared at all this season counts least (×0.6): the team has already been playing without him.

A score of 0.70 or more is Critical, 0.50 High and 0.30 Moderate. Below that a player is Low — unless he is a regular starter, or a rotation player with ten or more starts, who are never labelled lower than Moderate.

Severity decides the order and the label of the list. It does not change the Power Loss percentage, which is always the plain sum of weights.

Some team pages also show an expected-goals (xG) figure moving from one value to another. The first is the club’s average xG per match this season. The second adjusts it using the absent players’ own records: for each absent player with a meaningful role, SquadCheck compares the team’s xG in the matches he started with the matches he did not, and applies part of the difference. The result is kept between zero and twice the starting value.

That comparison controls for nothing — the matches a player missed may simply have been harder, or easier, than the ones he played. So every figure carries a confidence label: High only when three or more absent regular starters each have at least five matches on both sides of the comparison, Medium when at least one does, and Low otherwise. Early in a season it is almost always Low, and should be read that way.

A player who has not yet played for his current club has no weight there, so he cannot be priced. Rather than drop him, the team page lists him as Impact not rated, and he does not move the percentage.

Goalkeepers are scored with the same defensive-action formula as defenders, which fits them poorly. And none of these numbers has yet been backtested against results. Power Loss is a description of the squad, meant to be read next to the names on the list — not a prediction on its own. It is not an input to the match predictions either: those are built from results alone, and the Power Loss of each side is shown beside them rather than inside them. Where absences do change what the site predicts is the lineup — see how the predicted lineups are built.