GoalSoon · 11 August 2026

Bradford finished fourth and Stevenage sixth. On performance they were tenth and twelfth.

League One's play-off six banked 34.8 points more than their performances were worth, the widest that gap has been in the twenty-six seasons we hold — and across 425 clubs that stayed in the division, none of it carries into the next year (r = +0.015). Plus a league that has scored 2.62 goals a game for a quarter of a century while its shot count climbed, the fewest red cards in the archive, and the worst Championship season we have on record arriving in the third tier.

Fourth on the tenth-best season

Bradford City finished fourth in League One last season with 77 points. Stevenage finished sixth with 75, having scored 49 goals in 46 matches — as few as Wigan, who came sixteenth. Both went into the play-offs. Neither had the fourth- or the sixth-best season in the division.

We refit every match of it with a season-long Poisson model — an attack and a defence rating for each of the twenty-four clubs, fitted on that season alone — and asked what each club's performances were worth in points. Bradford's answer was 68.0, which was tenth. Stevenage's was 64.7, which was twelfth. They took 77 and 75.

One warning before the table, because it is the trap in this kind of work. Deserved goals would not be a finding: the way this model is fitted, the sum of a club's expected goals comes out exactly equal to the goals it scored, so that column would read 58 against 58 and mean nothing. Points are different. Points are a non-linear function of goals — a one-goal win and a four-goal win pay the same — so a club that creates and concedes what a 65-point side does can finish on 75. That gap is the subject.

League One 2025/26, what each club took and what it was worth

Exeter went down on 49 points from a season worth 57.1, which would have been fifteenth. Doncaster stayed comfortably up on 60 from a season worth 50.9, with a goal difference of minus nineteen.

Added up, the six clubs in the play-off places took 34.8 points more than their performances were worth. That is the widest that gap has been in the twenty-six seasons we hold. It needs one honest qualifier, though: the table selects for clubs that banked more than they played, so the number is never zero. Across all twenty-six seasons the top six average +13.1 points with a standard deviation of 12.2, which puts last season 1.8 of those out — the largest, but not far enough out to be called strange on its own.

The promotions themselves came out right anyway. Lincoln, Cardiff and Bolton went up, and those were also the three best seasons in the division on performance. Bolton, the one club in the top six that took exactly what it played for, won the play-offs.

None of it carries into the next season

The obvious question is whether banking more than you played for says anything about next year. Running the same fit over every League One season back to 2000/01 gives 623 club-seasons from 14,200 matches, and 425 cases of a club playing the division in consecutive seasons.

The correlation between this season's gap and next season's is +0.015 (t = 0.31). That is not a weak relationship, it is nothing. The forty-four clubs that finished five or more points above their performances averaged −0.2 the following season; the sixty that finished five or more points below averaged −0.5. Both are ordinary. The standard deviation of the gap across all 623 club-seasons is 4.7 points, and only 24 of them are ten or more in either direction.

Five of the ten biggest overshoots in the archive were not in League One the next season, and all five had gone up. Overachieving and getting promoted are frequently the same event, which is a good reason not to treat the gap as a warning by itself. The largest shortfall belongs to Fleetwood in 2021/22, 40 points from a season worth 52.1.

Twenty-six seasons of the division itself

Goals in this league do not move. Across the twenty-six seasons the average is 2.62 a game with a standard deviation of 0.09; the highest was 2.81 in 2004/05, the lowest 2.42 in 2007/08, and last season's 2.63 sits on the mean. Shots, meanwhile, went from 21.0 a game in 2000/01 to 23.8 last season, and corners fell from 11.3 in 2010/11 to 10.1. More shooting, the same number of goals.

What has changed is the refereeing. Yellow cards have gone from 2.95 a game across the first five seasons of the archive to 3.67 across the last five, peaking at 3.97 in 2023/24. Red cards went the other way: 0.192 a game across the first five seasons, 0.137 across the last five, and 61 in 552 matches last season — the fewest of any season here, against 93.6 expected at the rate of the previous twenty-five (z = −3.4). Treat the single season lightly, because it is one number out of several we looked at; the direction across five seasons is the sturdier part.

What arrives for 2026/27

Seven clubs are new. Leicester, Oxford United and Sheffield Wednesday come down from the Championship, and all three took fewer points than their performances were worth: Leicester 52 against 57.1, Oxford 47 against 52.8, Sheffield Wednesday 18 against 26.1. Wednesday's is the lowest points total in any of the twenty-six Championship seasons we hold — two wins in 46, 29 goals scored, 89 conceded, and that is on results alone, before any deduction.

From League Two come Bromley, MK Dons, Cambridge United and Notts County. The order there is worth a look: Bromley won the division with 87 points from a season worth 80.0, while MK Dons finished second on 86 from a season worth 89.5 — the best performance in the fourth tier — and Cambridge third on 82 from 86.7. Two of the three best fourth-tier seasons went up behind the club with the best table position.

What this cannot tell you

There is no xG in this division before 2025/26, and even last season our collector holds it for a little over half the team-matches, so everything above is built on goals and results, not on chance quality. The fit is centred on League One alone — a rating here is not comparable with a rating from the Championship or League Two. And when we held our own projection out season by season it came in at plus or minus 12.0 points over 46 games, against 13.7 for handing every club the league average. The whole spread of our 2026/27 projection is about 21 points, which is under two error bars, so we publish bands and not a 1-to-24 order. Anyone selling you a table of this division to the place is selling you the error bar.

All twenty-four clubs for the new season, with their summer arrivals and departures, minutes retained and squad ages, are on the League One 2026/27 page.