Advice for your team, free Open the app

Does form last in FPL? What nine gameweeks of 2026/27 show

After this article you will know how much a hot or cold run of form should move your transfer plans and how much is just noise. The short version: minutes tend to repeat, points runs mostly do not.

What we tested and how

fplwire tracked FPL points and minutes for every player across GW1 to GW9 of the 2026/27 season, 670 players in total. Nine gameweeks is a small sample from one season, so treat everything below as an early finding, not a law.

See what this means for your own team

Open fplwire for your teamFree. No account, no FPL password.

For this test, form means a player’s average FPL points over his last three gameweeks. A regular starter is defined as someone who played 60 or more minutes in at least two of those three gameweeks.

Does a hot streak carry into the next match?

The main test looked at 967 next-match results from 236 regular starters across GW4 to GW9, counting only cases where the player went on to play 60 or more minutes in that next match. Players were sorted into thirds by their form.

  • Hottest third: 5.92 points a gameweek over their last three, then 4.34 next time.
  • Middle third: 3.41, then 4.01.
  • Coldest third: 1.81, then 3.50.

The gap between hottest and coldest was 4.11 points in the form window. In the very next match that gap shrank to 0.84 points. Hot players did still score more on average, but by less than a single point, not the four-plus point gap their recent form suggested. Look at the size of the moves too: the hottest group fell by 1.58 points and the coldest group rose by 1.69. One limit on this test: it only counts players who went on to play 60 or more minutes, so a cold player who then lost his place is not in these numbers. The minutes section below covers that side.

Does price change the picture?

Splitting the same hottest-versus-coldest comparison by current price shows the small edge for hot players is not consistent:

  • Under £5.0m: 3.56 against 3.21 (343 results).
  • £5.0m to £6.5m: 4.55 against 3.60 (490 results).
  • £6.5m to £9.0m: 4.25 against 4.81, so the cold group actually scored more here (108 results, a small group).
  • £9.0m and over: 5.38 against 4.25 (26 results, too few to lean on).

So the modest overall edge for in-form players does not show up in every price band.

What happens to the very hottest players?

A separate test took the 20 highest scorers over three gameweeks at four different points in the season (GW1-3, GW2-4, GW3-5, GW4-6) and followed them for the next three gameweeks.

  • Overall: 7.84 points a gameweek in the hot spell, falling to 3.79 a gameweek afterwards, against 3.04 for all regular starters in that period.
  • GW1-3: 7.77 then 4.20.
  • GW2-4: 7.92 then 4.00.
  • GW3-5: 8.03 then 3.67.
  • GW4-6: 7.63 then 3.28.

In each window, only 2 to 4 of the original top 20 were still in the top 20 for the following three gameweeks. The group as a whole stayed above average, but it was a different mix of names doing it. This is a textbook case of regression to the mean, the tendency for unusually high or low results to be followed by something more ordinary.

Does one big haul or one blank tell you anything?

A haul here means 10 or more points in a match with 60 or more minutes played. There were 110 such hauls. In the next gameweek, 85% of those players played 60 or more minutes again. Among that group, they averaged 4.43 points, against 4.30 for the same players in their other matches, a difference of just 0.13. Only 9% hauled again (10 or more points), against 6% after a blank, barely different.

A blank here means 0 to 2 points in a match with 60 or more minutes. There were 795 of these. In the next gameweek, 81% of those players played 60 or more minutes again. Those who did averaged 3.78 points, against 3.80 for the same players in their other matches. A single blank told us almost nothing about the next match.

Which signal actually predicts the next match best?

Using rank correlation, a statistic that measures how well one ranking predicts another, where 0 means no link, 1 means a perfect link and a negative number means the link ran backwards, four signals were compared against the 967 next-match results. The 95% range in brackets is where the true figure plausibly sits, given how small the sample is:

  • Form over the last three gameweeks: 0.095 (95% range 0.028 to 0.155).
  • Season average so far: 0.094 (0.025 to 0.153).
  • Last match alone: 0.076 (0.014 to 0.129).
  • Current FPL price: 0.146 (0.071 to 0.212).

A shuffled, meaningless version of form was used as a control and scored -0.011, close to zero as a meaningless signal should be. The ranges for the four real signals overlap, so none is proven better than the others. Price looks strongest, but it has an unfair advantage: this test used today’s price, which has moved since those matches as managers reacted to results. All four signals are weak on their own.

Form’s week-to-week reliability also varied a lot: 0.273 in GW4, -0.004 in GW5, 0.159 in GW6, 0.015 in GW7, -0.067 in GW8 and 0.190 in GW9. The signal comes and goes from one gameweek to the next, it is not a steady dial.

The part of form that does last: minutes

While points form is unreliable, playing time is much more consistent. Looking at the chance a player reached 60 or more minutes in the next gameweek, based on how many of his last three gameweeks he hit that mark, among players who played at least some minutes in those three:

  • 3 of 3 recent matches at 60+ minutes: reached it again 86% of the time (872 cases).
  • 2 of 3: 59% (364 cases).
  • 1 of 3: 38% (372 cases).
  • 0 of 3: 14% (701 cases).

This is a far stronger pattern than anything seen for points. A regular starter stays a regular starter far more reliably than a hot scorer stays a hot scorer.

GW1-5 against GW6-9

Looking further out, 142 regular starters played 60 or more minutes at least three times in both GW1-5 and GW6-9. The correlation between their points per match in the two stretches was 0.072, with a 95% range of -0.09 to 0.25, a range that includes zero, meaning no reliable link was found over those longer stretches either. The top 20 from GW1-5 averaged 6.86 points per match in that spell, then 4.55 in GW6-9, against 4.02 and 3.89 for all 142 players. 8 of those 20 stayed in the top 20.

Individual examples show how mixed outcomes are. Some fell: Pascal Groß of Brighton went from 9.4 points per match in GW1-5 to 3.2 in GW6-9 (counting only matches where he played 60 or more minutes), Jayden Bogle of Leeds from 8.4 to 2.2, James Tarkowski of Everton from 8.6 to 2.8. Some stayed high: Erling Haaland of Man City went from 7.8 to 7.2, Maxim De Cuyper of Brighton from 7.6 to 7.8, Ben White of Arsenal from 6.7 to 7.0. Some rose: Evanilson of Bournemouth went from 3.6 to 8.5, Nicolas Jackson of Aston Villa from 3.3 to 8.2. These are examples only, not a guide to what happens from here, and nothing here explains why any of them rose or fell.

A worked example

Say you are weighing an extra transfer, which costs 4 points, to swap a cold player for a hot one. Both groups in this test played 60 or more minutes in the next match, and the gap between them was 0.84 points. At that rate the hit would take nearly five gameweeks to earn back, and only if the gap lasted and both players kept their places. Form alone rarely pays for that. Other reasons, such as fixtures or a lost starting place, have to do the rest.

What to do with this

  • Treat a hot run as a small nudge, not a strong signal. Across all prices the gap in the next match was under one point, not the multi-point gap it can feel like.
  • Do not chase a single haul or panic over a single blank. Both told us almost nothing about the very next match in this data.
  • Weight playing time heavily. A player who reached 60 minutes in each of his last three gameweeks did so again 86% of the time, far more dependable than any points pattern.
  • Remember regression to the mean. The very best scorers over any three-gameweek stretch tend to come back toward average, and a different set of names usually leads the next stretch.
  • Check team news before each deadline rather than relying on recent points alone. None of these signals is strong enough to use in isolation.

See what this means for your own team

fplwire reads your live FPL team and shows you your captain options, transfer targets, chip timing and mini league standings in one place. Free, and no account needed.

Open fplwire for your team

Related reading