GoalSoon · 1 August 2026

Sheffield Wednesday won one league game last season. Our League One projection cannot hold it against them.

Twenty-seven clubs have dropped from the Championship into League One in ten seasons. How they played on the way down correlates +0.05 with what they do next — no signal at all. Coming up from League Two is the opposite, and costs a third of your points. All 24 clubs for 2026/27, with the error we measured on ourselves.

One win in a season, and the model shrugs

Sheffield Wednesday finished bottom of the Championship in 2025/26 with fifteen points. In the forty-five matches our data records they won once, drew twelve and lost thirty-two. It is the worst season by any club in our English database.

We have just built a projection for League One 2026/27 — all twenty-four clubs, from ten seasons of results — and it puts Sheffield Wednesday in the top group. Not because we rate the squad. Because the data gives us no licence to rate them down.

Here is the whole of it. Over ten seasons, twenty-seven clubs have been relegated from the Championship into League One. Correlate how they did on the way down with how they did the following season, and you get r = +0.05. That is not a weak relationship. It is the absence of one.

Read the middle row as a sentence: a club dropping out of the Championship lands at about 1.67 points a game in League One, and almost nothing about its Championship season moves that number. The three coming down this summer finished 19th, 22nd and 24th, on 1.13, 1.04 and 0.33 points a game. The model separates them by four hundredths of a point.

What does carry is coming up

The bottom row of that table is the interesting one. League Two form transfers better than Championship form does, at r = +0.38, and it transfers with a steep slope: the promoted clubs that dominated stay ahead of the promoted clubs that scraped up. But the level itself takes a large bite. Thirty-five promoted clubs have averaged 1.26 points a game in League One, and this year's four all arrive well above that:

Bromley won a division at 1.91 points a game and the same method expects 1.23 from them here — a drop of roughly a third, which over forty-six games is thirty-one points. That is not pessimism about Bromley. It is what the last thirty-five promoted clubs did.

How wrong this is, stated before the table

We held out one season at a time, learned the transfer rates from the others, predicted the season we had removed, and compared the result against two approaches that need no model at all.

So it beats knowing nothing, and it beats the naive assumption comfortably, and it is still wrong by twelve points across a season. On position it misses by 5.6 places on average, landing within three places 44 per cent of the time and within six 66 per cent.

The entire spread between our highest and lowest projected club is 21 points. The error on any one of them is ±12. A ranked 1–24 table built on those numbers would be a decorative object, so we have not made one — the clubs below are sorted, but they are grouped into three bands drawn at half the measured error, and within a band the order carries no information.

League One 2026/27, twenty-four clubs

The membership is not our guess either: it is read from the fixture list, the clubs that actually have 2026/27 League One matches scheduled.

The part the model cannot see

Everything above is built from results, which means it is blind to the summer it is projecting. So we pulled the squads as well. Across the twenty-four clubs there are 560 registered players, and since 1 May our data records 103 arrivals, 195 departures and 131 returns from loan — with seventy-five of those departures leaving for no club we can name, which at this level usually means a contract that simply ended.

The turnover is not spread evenly, and that is the useful part:

Mansfield have kept 88 per cent of the minutes they played last season and field the oldest squad in the division at 29.2 years. Wigan have kept 45 per cent. Peterborough's squad averages 21.9, more than seven years younger than Mansfield's. A projection that puts Mansfield and Wigan within a point of each other is, quite openly, projecting two different kinds of season and calling them the same.

The full page has all of it: every squad player with last season's appearances, minutes and goals, every recorded arrival and departure by club, and the method in longer form.

League One 2026/27: the 24 clubs, their squads, and what our numbers expect →