Player prop edges on shots, shots on target and fouls, ranked by expected return.
Bookmakers price player props with less precision than match odds, which is where the edges live. We only publish them in the markets where our own projections have earned the right to be trusted. This page shows shots, shots on target and fouls committed. They are officially recorded stats, priced deeply enough for us to strip the bookmaker's margin out of the odds and compare like for like. Markets whose settled record does not support them are withheld rather than padded onto the page, and any market that starts losing is pulled automatically. Every pick shows the model's probability, the market's margin-free probability and the expected return per unit staked at the price on offer. That return is what the list is ranked by, because the same edge is worth far more on a 4.00 shot than a 1.20 one. Picks require the player to be in the expected XI, since a no-start voids the bet. The model's number also has to agree with the player's recent hit rate before a pick qualifies. For the empirical view of the same markets, last-six-game hit rates rather than the projection, see Form Edges.
A bookmaker's price contains its margin, so the raw price is the bar a bet has to clear. We also strip that margin out — using the shape of the whole priced ladder for a player and stat — to see what the market really thinks, then shrink our own projection toward it, because a large disagreement with the price is usually model error rather than value. The published edge is that blended probability against the price you can actually get.
Because the rest have not earned it. Shots, shots on target and fouls committed are officially recorded and priced deeply enough to analyse properly. Markets like goals and cards were measurably unprofitable and are withheld until their record supports showing them again — that judgement is made on settled results, automatically, not by opinion.
Most bookmakers void a player prop when the named player does not start, so picks require the player to be in the expected or confirmed XI. Our own published record settles the same way — a no-start is a void, not a loss.
Because two percentage points of edge on a 1.20 shot and two on a 5.00 shot are not the same bet. Expected return converts the edge into what it is worth per unit staked, which is the only ranking that treats short and long prices consistently.
Yes, 1+, 2+ and 3+ are projected for each player and stat, and the deeper lines are treated more cautiously: a projection error moves the probability of three or more far more than it moves one or more.