EFL Championship · 2026/27
Updated Sep 1, 2026The Ninjas' Championship
Where the model says all 24 finish — automatic-promotion title, top 6 (promotion playoff) and relegation odds for the 2026/27 season. A pure preseason projection until the first whistle; real results take over from there.
- 1
Southampton4726%77%—76 - 2
Middlesbrough3620%70%—74 - 3
Millwall3613%60%1%72 - 4
Swansea4108%47%1%70 - 5
Bristol City475%37%2%67 - 6
Derby County444%34%3%67 - 7
Birmingham464%34%3%67 - 8
Sheffield Utd464%32%3%66 - 9
Norwich433%29%4%66 - 10
Charlton393%27%4%65 - 11
Wrexham323%28%5%65 - 12
QPR373%27%5%65 - 13
West Brom362%24%5%64 - 14
Watford451%15%9%61 - 15
Stoke431%15%9%61 - 16
Blackburn451%11%13%60 - 17
Portsmouth43—7%18%58 - 18
Preston40—7%20%57 - 19
Wolverhamptonpromoted47—6%22%57 - 20
West Hampromoted45—4%31%55 - 21
Cardiffpromoted33—3%31%54 - 22
Lincoln Citypromoted44—3%33%54 - 23
Boltonpromoted44—3%34%54 - 24
Burnleypromoted31—2%43%52
Where they finish
The table isn't a ranking — it's a cloud. Each row is a club, each column a finishing spot across 20,000 simulated seasons. Brighter = more likely; gold is the automatic-promotion title, green a top 6 (promotion playoff) place, red the drop.
SOUMIDMILSWABRIDERBIRSHENORCHAWREQPRWBAWATSTKBLAPORPNEWOLWHUCARLINBOLBURHow the Ninjas call it
The same engine behind our Premier League table: every club carries an attack and defense strength learned from last season's results, regressed halfway to the league average (one season overstates the gaps). It cannot see this summer's transfers. We have no free points-total market for this league yet, so these are an honest last-season prior. When we find one, this board gets the same treatment as the Premier League and we will say so. Those feed two independent Poisson distributions per match — home/draw/away — and we simulate the full season 20,000 times to read off how often each club wins the automatic-promotion title, finishes top 6, or goes down. Promoted clubs inherit last season's relegation-zone profile. Once the season kicks off, real results replace the simulated ones.