How do you work out how likely a player is to score?

Two numbers. Start with how many goals the team is expected to score in the fixture, then multiply by the share of the team's goals this player usually provides. That gives the player's personal scoring rate for the match, and the chance of at least one goal follows from it. A striker taking 30% of a team expected to score 1.6 lands at roughly a 38% chance of scoring.

What does a 40% chance to score mean?

It means that if the same match were played many times, the player scores in about four of every ten. It does not mean he probably scores; 40% fails more often than it succeeds. It also does not mean the number was wrong when he blanks. A single game can never prove a probability right or wrong, which is why an honest model grades itself over many calls, not one.

Why is no player ever certain to score?

Because the ceiling is set by the team, not the player. Even a striker taking an enormous share of his team's goals can only score if the team scores, and teams are routinely held to one or none. Run the sum for the best striker in the world in a friendly fixture and it tops out around 50%. Anything sold as more certain than that is a story, not a number.

Do penalty takers have a better chance of scoring?

Yes, meaningfully. Penalties are converted at a high rate and they are added to a taker's share of goals rather than replacing any of it. Taking the penalties is worth more to a player's scoring chance than most differences in form, which is why a mid-priced taker can carry a better number than a more talented teammate who does not take them.

the scorers

Who's most likely to score, and by how much

Turn a team's goals into one player's chance of scoring and the answer is humbler than the hype: the best striker in the world, in a good fixture, is still odds-against. The sum, and how to read it.

21 July 2026 · 6 min read · James Frewin

Cover illustration for Who's most likely to score, and by how much

Check it withChatGPT Perplexity Claude

“He’s scoring tonight.” Every group chat has someone who says it with total conviction, and once in a while the game makes them a prophet. The honest version of that sentence is a number, and the number is easy to build: how many goals should the team score, and what share of them does this player take? Two inputs, one multiplication, and an answer that is humbler than the hype.

The sum

Start with the team. A model prices a fixture as expected goals for each side, say 1.6 for a good attack against an ordinary defence. Then ask what fraction of the team’s goals this player usually provides. A main striker might take 30%. Multiply and you have his personal rate for the night: 0.48 of a goal. Goals being rare events, the chance of at least one follows from that rate, and 0.48 works out at roughly 38%.

That is the entire trick behind every “anytime scorer” price you have ever seen. The bookmaker’s number and the model’s number are both built from these two parts: the team’s goals, and the player’s slice of them.

Elite striker, soft home fixture2.0 × 35%

50%

First-choice striker, ordinary game1.5 × 30%

36%

In-form attacking midfielder1.4 × 18%

22%

Full-back who “always pops up”1.2 × 8%

Illustrative, not a tip sheet. Team expected goals times the player’s share gives a personal rate; the percentage is the chance of at least one goal at that rate. Bars scaled to a 50% ceiling, which is where reality tops out.

Nobody is ever likely to score

Run the sum for the most feared striker in the world, in form, at home to a struggler, and it tops out around 50%. The best case in football is a coin flip.The ceiling is set by the team, not the player: he can only score if the team scores, and even great teams are held now and then. So “he always scores against us” is not an observation, it is memory doing what memory does, keeping the hits and losing the misses. The number never said likely. It said more likely than anyone else on the pitch, which is a different and much smaller claim.

A 40% chance fails six times in ten. When it lands, that is not the model being right. When it misses, that is not the model being wrong.

What actually moves the number

Three things, in order. The fixture: a softer opponent lifts the whole team’s expected goals, which lifts every attacker’s chance at once. The penalties: spot kicks are converted at a high rate and belong to whoever takes them, so the armband on penalties is worth more than most differences in raw talent. And the minutes: the sum assumes a full game, so a player likely to get an hour carries roughly two-thirds of his full-game chance, and a probable substitute far less. Form, the thing the chat argues about most, moves the number least.

How the engine sets the shares

Touchline’s open-source engine builds scorer weights the same honest way: half from a player’s verified share of his team’s recent goals, half from a prior, so one hot month cannot hijack the number and one quiet month cannot bury a good player. The blend is deliberate. Raw recent form is loud and mostly noise; the prior is slow and mostly signal; the truth sits between them.

Probabilities, not picks

None of this tells you who to back, and it is not meant to. It tells you what the confident voice in the chat is actually claiming, and gives you a defensible number to put next to it. If you want your own call on the record, tag it on a match page and let the leaderboard grade you over a season, the same way the receipts grade the model.