Michael Carrick answers 15 questions you've ALWAYS wanted to ask a Premier League manager! Unpacked

Michael Carrick's wide-ranging interview with Sky Sports Premier League (YouTube Video) pulled back the curtain on what it actually means to manage in the Premier League — from the unrelenting scrutiny of results to the invisible timeline every manager works against before patience runs out. While Carrick's answers spoke generally to the profession, the themes resonate sharply with the current situation at Manchester United, where Rúben Amorim is navigating exactly the kind of pressure Carrick described. predict.game's models quantify that pressure precisely. United sit 11th in the table, a position built on a brutal recent sample: LLWWL across the last five, with just 2 goals scored against 5 conceded in their most recent 3-match window. The underlying numbers are more nuanced than the results suggest — an xG/game of 1.45 against an xGA/game of 1.6 shows a team generating reasonable attacking value but consistently exposed defensively, a net xG differential of -0.15 that aligns with a squad still searching for tactical cohesion under Amorim's demanding 3-4-2-1 structure. That system, built on aggressive wing-back pressing and vertical inside-forward runs, is inherently high-risk when personnel is compromised. The absences of Luke Shaw (calf strain) and Mason Mount (hamstring injury) are not marginal losses — both are central to the double-pivot rotations and half-space overloads that make Amorim's shape function. Shaw's specialist role as an inverted left-sided wing-back is particularly hard to replace, and predict.game's simulation engine flags a measurable increase in expected goals conceded when auxiliary defenders are forced into that slot, consistent with the 1.6 xGA/game figure. For anyone modeling United's next fixtures, this context matters more than the win-loss column alone. The LLWWL form string masks a team whose xG output (1.45) is competitive for a mid-table side, suggesting the underlying process hasn't collapsed even as results have wobbled — a classic signal in predict.game's framework for potential regression toward the mean, in either direction. However, the defensive injury cluster introduces volatility that pushes simulated win probabilities down and draw/loss probabilities up for the next two to three matches, until Shaw or a like-for-like replacement returns to stabilize the back line. Carrick's broader point — that managerial success is judged in results long before process catches up — is precisely the tension predict.game's models are built to isolate. Amorim's system shows promising process metrics (1.45 xG/game) undermined by results and injury attrition, a gap that historically narrows once squad availability normalizes, making the next fixture cycle a critical data checkpoint for United's trajectory.