Four Games to Go: What the Data Says About the Championship Run-In
Four Games to Go
13th April 2026 · After GW42 · 50 fixtures remaining (GW43–46) · ML xPts projections
All 12 GW42 fixtures are done. Coventry need two points from four games to be mathematically certain of the title. Ipswich are already confirmed as champions-elect in second. The real action is below them: three teams scrapping for the last four playoff spots, and a relegation battle that changed shape entirely on 8th April when Leicester’s PSR appeal was rejected.
Below is everything the BTP logistic regression model says about how the final 50 fixtures (plus two postponed games) play out. These are ML Expected Points Projections — not simulation outputs. xPts = (3 × win probability) + (1 × draw probability) per remaining fixture, summed. Model: Platt-calibrated logistic regression, log loss 1.064 vs baseline 1.073.
Current Table ·
ML xPts Projections ·
Playoff Race ·
Hull vs Wrexham ·
Relegation Battle ·
Fixture Probabilities ·
Methodology
The State of Play After GW42
Sheffield Wednesday are confirmed relegated. Leicester sit 23rd after their PSR deduction. Coventry lead by ten points. Here is the full table.
■ Auto promotion
■ Playoff zone
■ Relegation danger
■ Relegated
†Wrexham: fixture-count verified at 64pts. API shows incorrect 68pts. Ipswich have 6 games remaining including postponed GW26 (Portsmouth, Apr 14) and GW40 (Southampton, Apr 28). Portsmouth have 5 remaining.
Current Championship Standings
Championship Table
ML Expected Points Projections
The table below shows each team’s current points, the model’s expected points (xPts) from their remaining games, and the projected final total. xPts = (3 × win prob) + (1 × draw prob) per fixture, summed across all remaining games. These are not guarantees — they are the model’s probability-weighted expectation.
†Wrexham: using fixture-verified 64pts. API figure is incorrect. xPts projections are model estimates only — not betting advice.
Story 1: The Playoff Seeding Battle
Millwall, Middlesbrough, Southampton and Hull are separated by just five points. Playoff seeding matters: 3rd plays 6th in the semis, 4th plays 5th. Getting the home leg advantage in the first leg of a two-legged semi-final is a tangible edge.
Southampton are the key wildcard. With five games left and the model rating them as favourites in four of them, they could climb past both Millwall and Middlesbrough. Their GW43 game at Swansea (50% Southampton) already looks like a must-win if they want to threaten 3rd. But if results go against them, they could also slip to 6th, putting them in the toughest possible semi-final draw.
Story 2: Hull City vs Wrexham — The Sixth Place Fight
Hull City are 6th on 68pts (fixture-verified). Wrexham are 7th on 64pts (fixture-verified — the API is wrong, showing 68pts, but the fixture count confirms 64pts). Hull have a four-point cushion with four games each remaining.
Hull City — 68pts
Wrexham — 64pts â€
The model is clear: Hull should hold 6th. The projected gap at the end is 74.2 vs 68.1 — six points. Wrexham’s remaining schedule is brutal. They are underdogs at home against Stoke in GW43, heavy underdogs at Coventry in GW45, and underdogs at home to Middlesbrough on the final day. Even if Hull had a poor run, Wrexham’s xPts yield from this fixture list is too low to close a four-point gap.
Story 3: The Relegation Battle — Leicester’s PSR Crisis
Sheffield Wednesday are already down — 14pts, 30 defeats, mathematically confirmed. One more place is to be decided.
Before 8th April, Leicester were 20th on 47pts and the model considered them very likely to survive. Then their PSR appeal was rejected. A six-point deduction was confirmed, dropping them to 41pts — and into 23rd place, in the relegation zone, with four games left.
Leicester’s remaining fixtures are: Portsmouth away (34% win), Hull City away (36% win at Leicester — wait, Hull play Leicester away — Hull are the away team at Leicester), Millwall home (33% win), Blackburn away (27% win). The model makes them underdogs in all four. Their xPts yield of +4.9 from four games as an underdog in each is the lowest of any relegation-threatened side.
Oxford’s remaining fixtures: Derby away (30% win implied — Derby 50% vs Oxford 28%), Wrexham home (47%), Sheffield Wednesday home (53%), Millwall away on the final day (22%). That Millwall away fixture is punishing, but Oxford could be mathematically safe before they get there if they beat Wrexham and Sheffield Wednesday at home.
Fixture Probabilities: Teams in Contention
Win/draw/loss probabilities from the BTP logistic regression model for the remaining fixtures of the six teams still in meaningful contention. Bars show home win / draw / away win probability — the widths are proportional to percentage (200px total).
Millwall — 73pts, 4 remaining
GW43 · Apr 18
GW44 · Apr 21
GW45 · Apr 24
GW46 · May 2
Middlesbrough — 72pts, 4 remaining
GW43 · Apr 19
GW44 · Apr 22
GW45 · Apr 25
GW46 · May 2
Southampton — 69pts, 5 remaining (wildcard alert)
GW43 · Apr 18
GW44 · Apr 21
Apr 14 *played today*
Apr 28
GW46 · May 2
Hull City — 68pts, 4 remaining
GW43 · Apr 18
GW44 · Apr 21
GW45 · Apr 25
GW46 · May 2
Oxford United — 44pts, 4 remaining
GW43 · Apr 18
GW44 · Apr 21
GW45 · Apr 25
GW46 · May 2
Leicester City — 41pts, 4 remaining −6pts PSR
GW43 · Apr 18
GW44 · Apr 21
GW45 · Apr 24
GW46 · May 2
Methodology
Model: BTP Logistic Regression with Platt calibration. Log loss 1.064 vs baseline 1.073 on 2024/25 hold-out test data.
Features: Rolling 5-game and 10-game form (goals scored, goals conceded, points), current league position at kickoff, position band (one-hot encoded). No xG — Championship xG data only covers 2023/24 onwards, insufficient for training.
Training data: 2019/20 – 2023/24 Championship seasons. Rolling windows reset at season boundaries (no cross-season leakage). Crowd-present flag applied for 2019/20 and 2020/21 bubble seasons.
xPts formula: For each remaining fixture, xPts = (3 × win probability) + (1 × draw probability). Summed across all remaining games to produce expected points gain. Added to current points for projected total.
What this is not: These projections are not simulation outputs and are not betting advice. They do not account for injuries, suspensions, managerial sackings, or motivational factors (e.g. a relegated side with nothing to play for). Treat them as a probability-weighted expectation from a trained model — a useful reference point, not a prediction.
Postponed fixtures: Portsmouth vs Ipswich (GW26, Apr 14) and Southampton vs Ipswich (GW40, Apr 28) are included in the respective teams’ xPts calculations. Southampton vs Blackburn (GW45, rescheduled to Apr 14) is also included.
BeyondThePrem.co.uk · Championship Run-In Analysis · Generated 13th April 2026 · After GW42 · Not betting advice · Model log loss 1.064
