On the Productivity of Manchester United’s Academy: A Comparison

There’s been a great deal of handwringing from various media figures this summer about a perception of Manchester United supposedly departing from their usual philosophy. The ‘Manchester United way’, we are told, is to eschew big money purchases of ready made players, instead focusing on developing talent, with an emphasis on their famous academy. This has been rumbling on all summer, but really reached a peak with United shipping out Manchester born Danny Welbeck for one season of Radamel Falcao on deadline day.

“For so long the standard-bearers of youth football, Manchester United are now just another big corporation in the market.” (Mark Payne, ESPN).

“There was a place for him and I am not too pleased he’s gone to Arsenal. It is sending out the wrong message when local home grown players like that are leaving.” (Eric Harrison, former United coach).

And perhaps most dramatic of all:

“What will happen in the future now, nobody knows, but that thread has been broken now.” (Mike Phelan, former assistant manager).

This all ties in with the recent debates over the lack of young English players breaking through at top clubs, and arguments over the way clubs should be run in terms of their recruitment strategies. These are much bigger debates than I can hope to address, so I will skip over both issues.

I reacted to this outcry with some scepticism- traditionally United’s academy has been very strong, but I couldn’t recall too many players it had pushed through in recent years- Wes Brown, John O’Shea, and Darren Fletcher were the only ones who immediately sprung to mind in addition to Welbeck. This, I thought, was a bit meagre, and I sent an ill-informed and hasty tweet in reaction.

I wanted to compare the productivity of United’s academy to other top English clubs, to determine whether this initial opinion was right. I chose Arsenal and Liverpool for this comparison, as all three are clubs traditionally challenging for at least the top 4 over the course of the Premier League era.

I looked at every season from 1992/93 until 2013/14, totalling all appearances by all players for each of the 3 clubs in the league for any one season. I then worked out what percentage of these total appearances were accounted for by players signed before their 18th birthday (so, in essence, how much playing time did these players get over the season, expressed as a percentage). I chose to include players signed at 17 or younger because the vast majority of these players have not made their debut for another club, were not signed for the first team, and still represented some kind of ‘gamble’, where academy coaching would be needed to realise their potential. There are exceptions to this, which I will address a little later.

First, an average of these figures for all 3 clubs, by season:

Screen Shot 2014-09-07 at 10.00.21 (2)I had actually expected a sharper decline, but its quite steady between 20% and 30% for the last decade, with a greater number before 2000 as foreign players became more commonplace.

Do United better this average?

Screen Shot 2014-09-07 at 10.04.51 (2) Indeed they do, and the difference is pretty great for the most part. What we are largely seeing in this figure is the influence of the ‘Class of ‘92’, including David Beckham, Gary and Phil Neville, Ryan Giggs, Nicky Butt, and Paul Scholes. None of these players remain, however, and United have moved closer to the average in the past 4 seasons. At the peak, in 2000/01, 48% of all appearances came from players signed before their 18th birthday. That’s a lot. This includes players such as Wes Brown (28 apps), Ronnie Wallwork (12 apps), and Luke Chadwick (16 apps), as well as a relatively injury free season for all of the players mentioned above.

How do Arsenal and Liverpool do?

Screen Shot 2014-09-07 at 10.08.27 (2)Liverpool nearly hit 50% in 1998/99, with Jamie Carragher, Michael Owen, Steven Gerrard, Robbie Fowler, and Steve McManaman featuring heavily among others, and these players contributed to Liverpool’s jump above the average for 3 seasons. Largely they’ve been operating below the average, but have been pushed up by the likes of Jon Flanagan and Raheem Sterling more recently.

Arsenal have largely placed below the average too, since just before Wenger’s tenure began. They have seen an increase in more recent years too though-bettering the average since 2007/08. Whilst the development of players such as Jack Wilshere, Wojciech Szczesny, and Kieran Gibbs has contributed to this, there’s another type of player contributing to it too.

Remember when I said the vast majority of players signed at under 18 were not for the first team, and hadn’t made their debuts yet? There are exceptions. Arsenal have had 4 in recent seasons- Gael Clichy, who already had minutes at Monaco, Theo Walcott and Alex Oxlade-Chamberlain from Southampton, and Aaron Ramsey from Cardiff.

Manchester United’s Lee Sharpe had already played for Torquay, whilst Jamie Redknapp, Ronnie Whelan, and Steve Harkness had all featured for other clubs prior to joining Liverpool. How do these graphs look after excluding these players?

Screen Shot 2014-09-07 at 10.13.17 (2)With the average recalculated to exclude these players, United are still the leaders. Predictably, Arsenal are the ones who suffer.

So, how good is United’s academy?

Even after the influence of the Class of ’92 began to fade, Manchester United were still giving significant minutes to academy players such as John O’Shea, Wes Brown, Rafael, Jonny Evans, Darren Fletcher, Tom Cleverley, Danny Welbeck, and Adnan Januzaj. Whether this approach will continue remains to be seen (you somewhat sense Louis van Gaal’s selection of players such as Tyler Blackett is reluctant at this point), but I was certainly wrong in my initial opinion. This, of course, does not account for academy players who have gone on to have successful careers either in the Premier League, Championship, or abroad, such as Ryan Shawcross, Danny Higginbotham, David Healy, Kieran Richardson, Guiseppe Rossi, Paul Pogba, Frazier Campbell, Danny Simpson, and dozens more. This article suggests United have made around £140 million from academy sales alone. Considering many of these players move for big fees before making their debut for United, that’s a pretty glowing endorsement.

Follow One Man Club on Twitter


Visualising Minutes Played and Scoring Rates: This Season’s Top Scorers

I previously wrote a piece looking at some ways to visualise goal scoring rates, producing a few different figures with the aim of including as much information as possible about minutes played. The problem with statements such as ‘Player X has scored Y number of goals in Z number of goals’ is that this can be misrepresentative, particularly to players who are frequently used as substitutes or are withdrawn early. Any attempts to visualise scoring rate (i.e. how a player’s goal tally grows over a season) must include some information about minutes played.

I’ve been experimenting with some different techniques to get as much information as possible into these figures, and have knocked up some graphs looking at scoring rates, minutes played, and strength of opponent in any given game. Minutes played is now represented by a colour gradient- the darker the colour, the more minutes played in a particular game. Conversely, the lighter points mean the player did not play, or was on the pitch for a short amount of time. I’ve split the Premier League table after Newcastle in 9th place to denote strength of opponent, choosing this divide because it has been in place for much of the past few months. Fixtures against Top 9 teams are illustrated by the triangle markers, and those against Bottom 11 by the circles.

When the points remain level, this means no goals were scored in that particular game (for instance, with the example below, Luis Suarez did not play in game 5, but played in game 6 and scored twice, so the total rises).



We can see current top scorer Luis Suarez has played 90 minutes or thereabouts in almost all of his games since returning from his biting ban, and has frequently scored against the current top 9.

Sturridge was unfortunate with injuries earlier in the season, and missed a good chunk of games. This chunk has been sandwiched by some very consistent goal-scoring, and Sturridge has also scored well against the current top 9.


Yaya Toure’s goal tally has been boosted in recent weeks, but his goal-scoring exploits are generally fairly consistent. Again, he plays the vast number of available minutes for Manchester City, as one of their key players.


Its testament to Sergio Aguero’s great record in the first 15 games that he still sits among the top scorers after missing so many minutes for City. He’s been pretty much absent since around game 16, and even before then was often withdrawn early by Pellegrini.


Finally, its everyone’s  (now injured) wildcard pick for England’s World Cup squad, Jay Rodriguez. His form in recent weeks in particular has been fantastic, going from 5 goals after 15 games to 15 after 32.

Follow One Man Club on Twitter

High-scoring teams in the Premier League

Liverpool’s goalscoring this season has been nothing short of incredible. Liverpool have scored 88 goals in 32 Premier League games thus far this season, and if they continue at their current scoring rate they will break Chelsea’s record of 103 goals from 2009/10.

How do they measure up to other high-scoring teams in past Premier League seasons? Liverpool have scored 4 or more goals in 10 of their 32 games so far; is this typical of high-scoring teams, or do others rely more on consistency by scoring 2 or 3 goals more frequently? Obviously, with the season not yet finished, things could change, but how do things look at this point?

The following figures plot the frequency with which each team has scored a number of goals (0, 1, 2, 3, etc.) in a single game.  For example, Liverpool have failed to score in 5 games this season, and 5 goals in 3 games. The most immediate comparison to make is to fellow high scorers from this season, Manchester City.


Manchester City have scored 1, 2, 3 goals in a game frequently, whilst scoring 5 and 6 in a game twice each to give their goal total a big boost. Liverpool are weird in comparison. They’ve scored twice in a game just 3 times, seemingly more likely to score 3 or 4.

How do Liverpool compare to the record holders, Chelsea?


Chelsea appeared much more likely to score just once or twice, but their eventual goal totals received a massive boost from scoring 7 goals in 3 different games, and they sealed the record with an 8-goal haul on the final day. These two aren’t really similar- Ancelotti’s side was solid, but their record was skewed by those 4 games. The same can’t be said of Liverpool this season.

Looking further back, how do Liverpool compare to Manchester United and Manchester City in 2011/12, who scored 89 and 93 goals respectively.



They look a little more like City than United, but not too much similarity. United are very consistent, whereas City were more likely to score 5 or 6.

Going even further back, how do they look compared to Arsenal’s highscoring seasons in 2002/03 and 2004/05? (85 and 87 goals respectively).


Arsenal 04/05 looks a little like Liverpool, but both are more likely to score 1 or 2.

Finally, lets look at the marauding United sides of the late 90’s/early 00’s, with their famous ‘score more than you’ approach. These sides didn’t have great defences, but compensated by scoring well above average: 97 in 1999/2000 and 87 in 2001/02. This sounds a little like someone we know…. how do Liverpool compare here?


Utd 99At the risk of angering Liverpool fans with the comparison, Liverpool look quite a look like these United sides. None had a great defence, but scored a huge number of goals to compensate, and these United sides scored 3 or 4 goals in a game very regularly. It is worth noting that Manchester United finished 3rd in 2001/02, despite their goalscoring feats.

There is more than one way to reach the heights Liverpool are reaching this season: you can be a solid side who benefits from a few demolition jobs, such as Chelsea, or more consistently put 3 or 4 past opponents, like a few others in this comparison, who may not have such a strong defence.

Follow One Man Club on Twitter