Finishing position
The trajectory City were on
Takeover completed 1 Sep 2008. The trend is fitted on 2002/03–2007/08 only, which works out to +0.4 pts a season.
Every Premier League result from 2009/10, the first season covered by the Premier League's charges against City, to 2025/26, replayed three ways. You can keep Manchester City in, strike their matches out, or swap in the mid-table City of 2002–2008 on its old trajectory.
The real table from the real results.
Every Man City match is struck out, so the other 19 clubs play 36 games each.
City's real fixtures are replayed by a modelled City on the 2002/03–2007/08 points trend (49–56 pts a season). Figures are expected points from 20,000 simulations.
Takeover completed 1 Sep 2008. The trend is fitted on 2002/03–2007/08 only, which works out to +0.4 pts a season.
Each bar shows the modelled City win / draw / loss chance against that season's opponent; the real score is on the right.
Data. All 6,460 Premier League results 2009/10–2025/26, plus 2002/03–2008/09 for City's history, from the football-data.co.uk results archive. Tables use points, then goal difference, then goals scored. Points deductions are applied in every scenario: Portsmouth −9 (2009/10), Everton −8 and Nottingham Forest −4 (2023/24).
Without City. City's 38 matches are deleted and the remaining 342 results are tabulated as they stand.
Pre-takeover City. A Poisson team-strength model (attack, defence, home advantage) is fitted to each season's 342 non-City matches to rate every opponent. City keep their average 2002–08 balance of attack to defence, scaled so that their expected points across the real fixture list match the 2002–08 points trend line for that season. Each fixture is then simulated 20,000 times. The table shows expected totals; title, top-four and relegation odds come from the simulations, with everything except City's games held at the real results.
Limits. The other 342 results are fixed, so knock-on effects are ignored: transfers, managers, fatigue and the 2011/12 final day all stay as they were. A linear trend over six seasons is a modest projection, and City's actual 2008/09 (50 pts, 10th) sits right on it. Carried forward to 2025/26 it gives about 56 points, so the further out the season, the more this is a thought experiment.