Parts 1 and 2 established what was different and why. Part 3 is about what it all means.
The data points in a clear direction. Two managers, the same squad, roughly the same number of matches — and two genuinely distinct versions of Manchester United. Not just different in style, but different in the way individual metrics functioned within each system. The same actions produced different outcomes. The same numbers meant different things.
This final piece pulls those threads together, looks at what each manager’s system actually was, and asks what the contrast between them tells us about United as a club.
Part. 1
“The Numbers Don’t Lie.”
◀◀◀ Click to read!
Part. 2
“The Tactical DNA.”
◀◀◀ Click to read!
[ Amorim’s United ]
High-Intensity, Direct, Volume-Based
Amorim’s system had a clear identity. Press high, win the ball in dangerous areas, move forward quickly through long balls and ground duels, get into the final third, draw fouls, and create enough chances that the volume itself produces goals.
| Press | Amorim | Carrick |
| PPDA | 9.04 | 11.65 |
| Opponent Danger Possession Lost | 24.3 | 18.94 |
The press was the starting point. A PPDA of 9.04 — significantly lower than Carrick’s 11.65 — reflects a side that engaged opponents earlier and more aggressively. The direct output showed up in opponent danger possession lost: 24.3 per game, well ahead of Carrick’s 18.94. Amorim’s press wasn’t just intense in theory; it was producing turnovers in areas where they hurt.
| Attack | Amorim | Carrick |
| Long Balls | 53.05 | 46.82 |
| Ground Duels | 69.55 | 63.82 |
| Fouls Won in the final third | 2.25 | 1.41 |
From there, the attack was direct. Long balls were used at a rate of 53.05 per game — more than Carrick, and that gap was statistically significant. Ground duels were higher. Fouls won in the final third were higher. This wasn’t a team building patiently through the lines. It was a team that wanted to skip the build-up phase and get into the attacking third by force.
| Total Duels | Correlation (Amorim) | Correlation (Carrick) |
| Goal Difference | + 0.477 | . |
| Big Chances to Goal (%) | + 0.490 | . |
The output reflected that approach. Total duels correlated positively with goal difference and big chance conversion. On the days when Amorim’s United fought more, they also played better and scored more. The physicality wasn’t incidental — it was structural. Duels were load-bearing in this system. More fights meant more pressure meant more chances meant more goals.
| Others | Amorim | Carrick |
| Errors Leading to Shot | 0.45 | 0.82 |
| Shots off Target | 6.15 | 4.71 |
| Accurate Passes (%) | 81.184% | 83.3% |
The cost was equally clear. Errors leading to shot increased alongside pressing success. The same aggression that created turnovers also created vulnerability. Shots off target were higher. Pass accuracy was lower. Amorim’s United was a team that operated at a high pitch and accepted the risks that came with it.
The chain looked like this: high press → dangerous turnovers → direct progression → final third entries → volume of chances → goals.
[ Carrick’s United ]
Organized, Transitional, Efficiency-Based
Carrick’s system looked different at almost every level. Lower pressing intensity, higher pass accuracy, deeper defensive shape, fewer duels — and yet better results across most output metrics.
| Press | Amorim | Carrick |
| PPDA | 9.04 | 11.65 |
| Opponent High Turnovers | 3.7 | 5.06 |
The apparent starting point of Carrick’s system isn’t the press — it’s the deliberate absence of one. His PPDA was 11.65, meaning his side pressed less aggressively than Amorim’s in the opponent’s build-up phase. Instead of engaging high, Carrick’s United invited pressure, dropping into a deeper defensive shape and forcing opponents to come to them.
Opponent High Turnovers were higher under Carrick (5.06 vs 3.70) — but this isn’t a pressing metric. Opponents who were drawn into United’s defensive third were losing the ball there. The turnovers were a product of the low block, not of pressing high. What Carrick sacrificed in pressing intensity, he compensated for with defensive organisation and transition speed.
| Defense | Amorim | Carrick |
| Blocked Opponent Shots | 2.95 | 4.41 |
| GK Punches | 0.15 | 0.71 |
| Opponent Danger Possession Lost | Correlation (Amorim) | Correlation (Carrick) |
| Goal Difference | . | + 0.527 |
| Accurate Passes (%) | . | + 0.499 |
The deeper defensive posture showed up clearly in the numbers. Blocked opponent shots were significantly higher under Carrick (4.41 vs 2.95), and goalkeeper punches were dramatically higher (0.71 vs 0.15). Both point toward a side that regularly absorbed pressure and dealt with it closer to their own goal.
But the shape held. When Carrick’s defensive structure was functioning — when opponent danger possession lost was high — goal difference and pass accuracy both moved upward together. A single metric pulling the quality of the entire game with it. The low block wasn’t passive. When it worked, it worked as a complete system.
| Build-up | Amorim | Carrick |
| Accurate Passes (%) | 81.184% | 83.3% |
Pass accuracy at 83.3% — higher than Amorim’s 81.2% — fits the same picture. Carrick’s United retained the ball more carefully, waited for the right moments, and made opponents work harder to get it back. When possession was lost, it tended to be in less dangerous areas.
| Interceptions | Correlation (Amorim) | Correlation (Carrick) |
| Shots Inside Box | – 0.487 | + 0.463 |
| Attack | Amorim | Carrick |
| Big Chances to Goal (%) | 29.074% | 50.744% |
| Big Chances to Goal (%) | Correlation (Amorim) | Correlation (Carrick) |
| xG | + 0.536 | – 0.444 |
| PPDA | . | + 0.405 |
The transition piece is where Carrick’s system separated itself most clearly. When interceptions happened under Amorim, shots inside the box went down — the defensive action ended the threat without generating one. When interceptions happened under Carrick, shots inside the box went up. Winning the ball in midfield immediately became an attacking action. That conversion from defensive moment to attacking opportunity, happening quickly and consistently, is what drove the efficiency numbers.
Big chance conversion at 50.7% sounds like a clinical finishing performance. The correlation data reframes it. Conversion rate was higher on days when xG was lower and when pressing was less intense — meaning the big chances being converted weren’t coming from dominant attacking spells. They were coming from transitional moments, when the shape broke down for the opponent and United punished it with speed and precision.
The chain looked like this: organised mid-block → forced turnovers → quick transition → box entries → fewer but higher-quality chances → efficient finishing.
[ The Same Metrics, Two Different Languages ]
What makes the Amorim-Carrick comparison genuinely interesting — beyond the surface-level style difference — is how consistently the same metrics meant opposite things in each system.
( duels )
( interceptions )
( long balls )
Duels are the clearest example. Under Amorim, more duels meant a better game — higher goal difference, better conversion, cleaner shooting. Under Carrick, more duels meant a worse game — pressing success dropped when duel counts were high, suggesting the team was being pushed around rather than imposing themselves. The metric is the same. The meaning is inverted.
Interceptions follow the same pattern. Higher interception counts correlated with fewer shots inside the box under Amorim, and more shots inside the box under Carrick. The action of winning the ball back produced defensive consolidation in one system and attacking transition in the other.
Long balls complete the picture. Under Amorim, long balls appeared alongside stronger pressing — his side used them as a direct attacking tool, and the press forced opponents into them. Under Carrick, long balls appeared when pressing failed — his side went direct when the mid-block broke down, using the aerial duels as damage limitation rather than as a primary weapon.
This isn’t a case of one manager doing something right and the other getting it wrong. It’s a case of two different systems using the available tools in genuinely different ways. The squad adapted to both — imperfectly, with varying results, but recognisably.
[ Carrick’s United Won This One ]
The data doesn’t leave much room for ambiguity. By virtually every output metric that matters — goal difference, big chance conversion, xG created, opponent turnovers generated — Carrick‘s United outperformed Amorim‘s. That gap is large enough, and consistent enough across enough different metrics, to constitute a clear verdict. But the verdict needs to be framed carefully, because it’s easy to misread what it means.
This is not a claim that Carrick‘s tactical system is superior to Amorim’s as a general proposition. Tactical systems don’t have universal rankings. Amorim’s high-press, direct-progression model — the same system he ran at Sporting — produced a 4-1 win over Manchester City. The system works. What determines whether a system works isn’t the system itself in the abstract; it’s whether the system fits the league it’s operating in, and whether the players available can actually execute it.
On both counts, Carrick‘s approach was the better fit for this version of Manchester United in this particular season. The mid-block organisation, the transitional efficiency, the emphasis on ball retention and quick switches from defence to attack — these are things this squad could do. The correlation data showed it clearly: when Carrick’s system was functioning, everything moved together. Pressing success drove pass accuracy, goal difference, and game quality simultaneously. The pieces connected.
Under Amorim, the connections were different — and harder to sustain. High duels, high pressing intensity, direct long-ball progression — these are the fingerprints of a system that demands a very specific physical and tactical profile from its players. When it worked, it worked well. When it didn’t, the data showed the familiar signs of a squad operating at the edge of its capacity: more errors alongside more pressing success, lower pass accuracy, conversion rates that depended on volume rather than quality.
The conclusion isn’t that one manager is better than the other. It’s that one manager brought a system that fit better, and the numbers reflect that.
[ Players First, System Second ]
The deeper reason Carrick’s tenure produced better results goes beyond tactical fit. It has to do with how each manager related to his players.
Amorim arrived with a defined system and a defined idea of how the game should be played. That’s not a criticism — clarity of vision is a genuine managerial quality. But the cost of rigid tactical vision is that players who don’t fit the mold get forced into roles that don’t suit them, or get left out entirely. The system becomes the priority. The players become the means to an end.
Carrick took a different approach. The evidence — both in the numbers and in how the team played — suggests he started from the players rather than from the system. He identified what each player could do well, built roles around those strengths, and constructed a system from the ground up based on what this particular squad was capable of. The players weren’t being asked to execute someone else’s vision. They were being asked to do what they do naturally, within a framework that gave it structure.
That distinction matters more than it might sound. A player who understands his role and is performing within his natural range of movement and decision-making is going to outperform a player who is technically executing a system he hasn’t fully internalised. The data captures this: Carrick’s United was more cohesive, more consistent, and more efficient — not because the tactical blueprint was inherently superior, but because the players were operating with more clarity and confidence.
This is also where Carrick’s tenure stands apart from some of his predecessors. Getting results is one thing. Getting results while also demonstrating a coherent tactical identity — one built around the players rather than imposed on them — is considerably rarer. It’s what separates a manager who happens to win from a manager who understands why he’s winning. On that measure, Carrick’s record in this half-season is genuinely encouraging.
Closing — Across All 3 Parts
[ What the Data Can and Cannot Tell Us ]
Statistics record outcomes. Every number in this analysis — every duel count, every PPDA figure, every correlation coefficient — is a measurement of something that already happened. The data tells you what occurred. It doesn’t tell you why a manager made the decisions he made, what the training ground looked like, what conversations happened before matches, or what the players understood about their own roles.
What this analysis tried to do — across three parts and two different statistical methods — was work backwards from those outcomes to the systems that produced them. Welch’s t-test identified where the numbers differed meaningfully between the two regimes. Correlation analysis mapped how those metrics moved in relation to each other, separately under each manager. Together, they painted a picture not just of what was different, but of how each system was wired — which metrics were load-bearing, which ones signalled something going right, and which ones signalled something going wrong.
That reverse-engineering process is, I think, the most interesting thing statistical analysis can do in football. The numbers are the effect. The system is the cause. Working from one to the other is never perfect — there are always confounding factors, always things the data can’t see — but when the correlations are consistent and the patterns hold across enough matches, the picture that emerges is real.
What the picture shows, in this case, is that Carrick’s system was the better fit for this United squad in the 2025-26 season. Not because high-press, direct-progression football is inferior to mid-block transition football as a tactical philosophy — it isn’t. But because tactical quality, in practice, is always relative to context. The league, the opponents, the squad’s physical profile, the players’ tactical literacy, the time available to implement a system — all of these shape what a given approach can actually produce.
In this context, with this squad, Carrick’s approach produced more. The data is clear on that.
It also leaves an open question that won’t be answered until next season: what does Carrick’s United look like with more time, more preparation, and more players suited to his system? The 2025-26 data gives us a baseline. If a similar analysis were run on the 2026-27 season, the numbers we’ve identified here as strengths might look like limitations by comparison — because a more developed version of the system, with players who have had a full pre-season to internalise it, should produce different and better correlations.
That’s the hope, anyway. Football analysis at this level is most useful not as a verdict but as a mirror — something a manager, a club, or even a supporter can hold up against what they think they know, and see more clearly. If this three-part series has done that for how you think about what Amorim and Carrick each built at United, then it’s done its job.
The rest is up to Carrick.
( Ruben Amorim’s system )
( michael carrick’s system )
( Same metrics, opposite meaning )
| Amorim | Carrick | |
| Duels | Good Game Signal | Getting Pushed Around |
| Amorim | Carrick | |
| Interceptions | Defensive Act | Attack Trigger |
| Amorim | Carrick | |
| Long Balls | Attacking Weapon | Pressing Failure |
Raw per-match data was collected from Sofascore and MARKSTATS.
Aggregate metrics — including per-game averages, cumulative totals, and derived indicators such as Accurate Passes (%) and Total Duels — were constructed independently by the author.
All statistical analysis, including Welch’s t-test and Pearson correlation, was conducted by the author using this dataset.
Leave a Reply