Does momentum exist in football? What the data says

Listen to any broadcast and you will hear it inside ten minutes: this team has the momentum, they are on a roll, you can feel the shift. It is one of football’s most durable ideas, and one of its least examined. Before we hand a run of results the power to explain the next one, it is worth asking what the data actually shows. Mostly, it shows that we are very good at narrating noise.

A streak is what randomness looks like

Start with the uncomfortable maths. A fair coin has no memory and no confidence, yet flip it a couple of dozen times and you will see clumps: four heads in a row, then a scatter, then five tails. Our eyes read those clumps as a pattern, a hot hand, a cold spell. There is nothing there. The coin is exactly as likely to come up heads after a run of five as after a single one.

Football results carry the same trap. Across a season, most winning and losing streaks are statistically consistent with a team of a fixed underlying quality meeting the ordinary variance of a low-scoring sport. A run does not need a cause to exist. It only needs a schedule and a bit of luck, and it will appear on its own, ready to be explained by whoever is holding the microphone.

Why form often predicts a slowdown

Here is the part that surprises people. Where momentum does seem to have predictive power, it frequently points the wrong way. A team on a hot streak has usually been overperforming its underlying numbers: scoring a shade more than its chances warranted, converting half-chances, riding a run of clean sheets that the expected goals against did not merit.

That overperformance is the thing that regresses. Regression to the mean is not a curse that punishes success; it is simply what happens when the luck that topped up the results runs out and a team drifts back toward its true level. So “they are flying, back them to keep flying” is often exactly backwards. The streak was the overperformance, and the overperformance is the thing most likely to fade.

Where a real signal does hide

None of this means every good spell is an illusion. The honest position is that momentum, as usually described, is weak, but there are concrete, measurable reasons a team can genuinely change level for a stretch. It is worth separating those from the vibe.

Fatigue and fixture congestion.Legs are real. A side deep into a run of midweek games, or one that has just had a fortnight’s rest, is measurably different from its baseline, and the schedule is knowable in advance.

Injuries and suspensions. Losing a key defender or getting a talisman back is a change to the actual squad, not a mood. It moves the underlying numbers, which is why it belongs in a model rather than in a narrative.

A genuine tactical fix or a new manager. Sometimes a side really does solve something: a shape that stops the goals leaking, a role that unlocks a striker. The tell is that it shows up in the performance data, not just the results column.

Confidence. This one is real and also the easiest to overstate. It is plausible that a settled, in-form group plays with more freedom. It is also nearly impossible to measure cleanly, and it is the first thing reached for when the simpler explanations (luck, fixtures, personnel) would do. Treat it as a small effect wearing a large hat.

The honest verdict

Momentum in the loose, commentary sense (a mysterious force that makes the next win more likely because the last one happened) is largely a story we tell after the fact. It is over-narrated because a good story needs a cause, and “that was mostly variance” is a poor line of commentary. The signal that does exist lives in things you can point to: tired legs, a missing centre-half, a change that shows up in the chances. Underlying performance is the durable thing; the results streak is the shadow it casts, plus noise.