What is form rate?

You will see the number all over the site: a Win form of 4.00 going into a derby, or BTTS sitting at 3.50 on a fixture page. The score looks like it should favour recent matches. It does not: every match in the window counts exactly the same, whether it was played last week or two months ago.

How form rate works on FootyMetrics

Form rate is shown as the Win, O2.5 and BTTS boxes on team and fixture pages. It is not recency-weighted. The form rates explained page states the design directly: the scale runs from 0.00, meaning a condition never happened in the window, to 5.00, meaning it happened in every match, and each match in the window is worth a fixed, equal amount, 1.00 over a 5-match window, 0.50 over 10, 0.33 over 15. A win in the first match of the window and a win in the last match of the window add exactly the same amount to the score. There is no decay and no extra weight for the most recent game.

The player-level equivalent is hit rate, covered in reading a player trend card and how to use the Player Props Finder. A hit rate of 6/8 means the player cleared the line in 6 of their last 8 qualifying matches, and again, every match in that window counts the same regardless of how recent it was.

A concrete case shows what that flatness hides. Take a player whose trend card reads 6/10 on a shots on target line, and suppose the four most recent of those ten matches all landed. The card reports 60%, full stop. A recency-weighted number built from the same ten matches would sit noticeably higher, because the 4/4 run is the newest data and would carry the most weight. Both numbers are honest answers to different questions: the flat 60% says how often the line was cleared across the window, the weighted one leans toward what the player is doing right now.

What recency weighting means

Instead of treating every match in a window the same, a recency-weighted stat gives more recent matches more influence over the final number, usually through an exponential decay formula: a match's weight shrinks the further back in time it happened. The standard reference is the 1997 Dixon-Coles paper, and many football models built since, including ones bookmakers price from, keep some version of its time decay.

The point of doing this is to separate two different questions: what has this team or player done on average across the whole window, and what are they doing right now. Those two answers can diverge, especially after a manager change, an injury return, or a run of tougher or easier fixtures than usual. A flat average or a flat hit rate answers the first question. A recency-weighted stat is built to answer the second.

Why recency weighting matters conceptually

A flat hit rate and a recency-weighted one can tell different stories from the exact same set of matches. Take a player who cleared a line in 3 of their first 5 matches in a 10-match window, then cleared it in all 5 of their last 5. A flat hit rate reports 8/10, 80%, with nothing in that number to flag the shift partway through. A recency-weighted version would push higher still, because the better, more recent matches count for more than the weaker, earlier ones. The same logic runs in reverse: a player who started hot and has since gone cold looks better in a flat hit rate than their current form suggests, because the flat number has not caught up with the recent dip.

This is exactly why FootyMetrics' window setting exists even without weighting. Choosing Last 5 instead of Last 10 or Last 15 is a blunt way to lean toward recent form, at the cost of a smaller sample: a 5-match window reacts fast to a genuine change but can also overreact to three results either way, while a 15-match window is steadier but slower to reflect a real shift. Recency weighting is the more precise version of that same trade-off. Instead of picking between a short window and a long one, every match counts, just not equally.

One thing a weighted number would not do is hand you value. Bookmaker models weight recent form too, so by the time a hot streak is obvious enough to show up in any form stat, some of it is already in the price. Whether a form number, flat or weighted, points at a bet worth taking is a separate question, covered in hit rate is not value.

The caveat: a short run is still a short run

Weighted or flat, a form number built on five matches is still built on five matches. A player or team can look completely different depending on whether you catch them mid hot streak or mid cold spell, and neither streak is guaranteed to hold once more matches are played. See variance and sample size for the general reasoning on why a small number of matches carries far less weight than a larger one, and treat any form number, FootyMetrics' or otherwise, as one input alongside the actual match history behind it, not a verdict on its own.