Carrick's Seat Rests on One of 220 Simulations — and Man Utd's Board May Still Not Know It

Sitting in Rangpur, cross-referencing Premier League matches from recent seasons with boardroom decision documents, a pattern keeps returning: clubs appoint managers on results and sack them on process. Manchester United's recent history makes this asymmetry so stark it cannot be called accidental. At the end of last season, Manchester United were priced at 10-14% to win the title under Ruben Amorim. He was sacked in January despite his side creating more chances than opponents. Then Michael Carrick was handed a two-year deal on the basis of 2.29 points per match across 17 games. Expected points in that same stretch: 1.59. Results were running far above process. Based on my years watching matches, I have learned that a large gap between expected points (xP) and actual points never persists. Carrick's case was a textbook deviation. One proprietary simulation model found that such a run occurs in just one of 220 simulations. The board awarded a permanent contract on that single occurrence. This season, seven matches in, his side sit on 1.00 points per match and in the relegation zone. Expected points: still 1.58. That contradiction is the story. The problem is not Carrick's coaching but the board's decision process. They sacked Amorim on results while the team ranked top six for chances created. They appointed Carrick on results while the team ranked mid-table for chances created. Same board, same error, opposite directions. How reliable is this analysis? The core simulation was run by Dr. Mehdi Kordi, a sports physiologist and former Dutch Olympic track-cycling coach — not a football analytics specialist. His methodology has not been independently verified. One thing is clear: three independent xP data points (Amorim 1.76, Carrick interim 1.59, Carrick permanent 1.58) all point the same way: the chance-creation engine is essentially unchanged. What does that unchanged chance creation rate tell us? A team generating 1.58 expected points per match projects to about 60 points over 38 games — European-qualification-contender level in the Premier League, not title level. Preseason title odds of 10-14% were market overconfidence. Why the relegation zone, then? Another results-vs-process divergence. The team still creates chances but is not converting them. Under Carrick's interim spell, the side over-scored by 0.56 goals per match and under-conceded by 0.21 — neither sustainable. Regression is now underway. The team is European-level on process but relegation-zone on outcomes. A subtler question: was Carrick's appointment a mistake? On process, his side creates chances at the same rate as Amorim's (1.58 vs 1.76). The difference is finishing. Seventeen matches is a small sample; seven this season is smaller. Neither justifies judgment. From years of watching matches, I know the most dangerous statistic is the one that happens once and is treated as permanent law. Carrick's 2.29 points per match was that number. The board mistook it for a rule. Now the real rule (1.58 xP) is reasserting itself, exposing the flaw in the board's decision. I do not know whether Carrick will succeed at Manchester United. But one thing is certain: his appointment rests on an event that occurs once in 220 tries. If the board does not audit its own decision process, the same error will repeat with the next appointment. The biggest damage in football is done when a club ignores its own data. The question before Manchester United's board is not Carrick's future but whether they will learn to trust their analytics department. If the team regresses toward 1.58 xP, results may return; the flawed decision process will not fix itself. That is the real risk.