Data & video analysis on the cultured midfielder
The name David Beckham invokes strong emotional reactions in people. Particularly in the UK, he’s either one of your childhood heroes or he’s an overindulged, overpaid prima donna. There is no middle ground. Or at least these more sensible people are quieter on Twitter.
He’s massively overrated or underrated. People aren’t entirely sure.
But we have a firm opinion either way. It’s important to draw your lines, and stick to them, even if the only evidence we have is half remembered games from 20 years ago.
How reliable is one’s opinion of players from 20 years ago? I know mine can’t be particularly reliable, I was at primary school when Beckham was at his peak.
I thought it would be fun to look at some data to provide some evidence for his ability, and then add some colour to that with some video analysis.
As for the second question in the blog title – “could he have played another position?” – we’ll get to that later.
The obvious problem is that data from the 90’s and early 00’s is largely unavailable. Thanks to Fbref, I was able to pull some goals and assists data, some minutes played data and some (very limited) shooting data.
The question I wanted to answer with this data was, how does David Beckham compare to some of the best midfielders of his era and some of the PL midfielders we are familiar with now?
Selecting the players to look at
Important to note here that this part wasn’t particularly scientific. The selection process was incredibly manual and consisted of a mixture between my own memory and Ballon Do’r nominees in the late 90’s/early 00’s… to get in some of the top players who played in his era, in roughly similar positions. I wanted players who often played wide in a midfield 4… but chucked in a few also just for fun (e.g. Veron – to be fair he played wide in a 4-4-2 on occasion earlier in his career). To find similar (ish) current Premier League players, I initially filtered by high crossing volume to highlight players who put the ball into the box often from wide areas. I then gave this list a once over to remove vastly different players e.g. excluding players like Sterling and Salah – who operate as wide forwards, rather than midfielders. Kevin De Bruyne is in as he’s a player Beckham has been compared to for his final ball and devilishly whipped crosses. And, like Zidane, is a player who plays centrally, but often pops up in wide areas. Gylfi Sigurdsson is in for his set piece prowess. I thought this might appeal to the “Beckham was just a set piece merchant” brigade.
I’m sure there are some incredibly obvious wide midfielders I’ve missed, and some names that you may think shouldn’t be in there. But it gives us a decent idea at least, and some useful reference points.
Data was scraped from Fbref for each of the selected players and then combined into one table in Excel. Each player’s best season (with respect to goals and assists per 90 minutes) was selected along with the season before and after to form a 3-season average. The rationale behind this was that it would give some view of consistency, or at least avoid overrating any one-season wonders. At times, a bit of common sense had to be used. David Ginola had 3 really consistent seasons data wise, had I used the previous methodology to the letter, one of those would have been missed and I would’ve included a season where he only made 20 starts and his numbers were way down. Denilson’s time frame spanned four seasons, with the 2000-2001 season, which the Brazilian missed in its entirety through injury, excluded. Harvey Barnes and Dwight McNeil only have a 2-season average as, without stating the obvious, they have only been in the Premier League for 2 seasons, and Jack Grealish’s numbers are entirely from this season.
First up, is a single-axis scatter which plots each players’ combined Goals and Assists per 90 minutes (G+A p90).
I have to say, I did not expect to see Beckham so high here. He is the second-best player in the data set in this regard, behind only the brilliant, and probably underrated himself, Robert Pires.
0.78 G+A p90 is really, really good. For context, that’s better than Manchester City superstar Kevin De Bruyne’s 0.71, which may come as somewhat of a surprise.
A word for Theo Walcott here, 0.74 G+A p90 and 3rd best in the data set. For reference, this was between the ages of 21-23, too. May deserve some more respect for what is a fantastic career.
The boxplot you see overlayed on the chart displays the two extremes in the data (Redmond and Pires) as well as the 25th, 50th and 75th Percentiles. For more on how to read boxplots, click here: https://flowingdata.com/2008/02/15/how-to-read-and-use-a-box-and-whisker-plot/
This does not mean Nathan Redmond is a bad player by the way. By any stretch. This is a data set containing a number of Ballon D’or nominees, remember!
The next chart is a table displaying how many minutes each player played of the total available, in their respective timeframe.
The takeaways here are:
- Beckham is 11th in the data, showing a level of durability and availability for selection which is a really underrated asset. You might be a great talent but getting on the pitch consistently is important.
- Ashley Young was a monster physically between 2006 and 2009, playing 90% (!) of the total available league minutes
- Joe Cole had pretty good outputs but struggled to ever fully cement his place in that Chelsea team
Worth noting that Aldo Duscher broke Beckhams foot in April (remember that rush to get him fit for the World Cup!) so naturally his Minutes Played percentage would have been even higher without that.
For completeness, a full data viz can be seen below. I’ve also included a treemap of each players Goals per Shot on Target. In short, this looks at: when a player takes a shot on target, how often does he score? Beckham profiles fairly well here (joint 11th) suggesting he was a reasonably efficient finisher. This data is not particularly insightful – it misses context of whether the shot was from a good position or whether it was a scuffed effort from 30 yards straight at the goalkeeper. It does however give an idea of his shot conversion compared to the other analysed players and provides a bit of colour to a very limited data set, hence its inclusion here.
So, data has given us an outline of Beckham’s output and it’s a very promising start. Hopefully watching some of him in action can fill in the gaps.
WATCHING THE TAPES
There were plenty of games to choose from but in order to see how he performed in a big game against a good side – I selected a Champions League Quarter Final, at home to Bayern, from 2000. For those of you who calculate early noughties seasons by David Beckham haircuts, we’re in the buzzcut era.
This is a strong Bayern side with some great 90’s throwback names in their side.
Before we start on his performance, it is obvious throughout the game what a megastar he is. It seems like every break in play there’s a closeup of Beckham; mopping his brow, jogging back into position, or really doing nothing at all of any consequence. Between his footballing ability and his celebrity status, you may be inclined to remember one more than the other, depending on your disposition.
Out of Possession
Bayern see a lot of the ball early on so initially what strikes you about Beckham is his fitness levels and work rate off the ball. He’s a supreme athlete. He doesn’t appear to tire throughout the game, and you imagine he would be right up there in terms of sprints completed – if that data were available. He does a terrific job in pressing Bayern’s midfielders, making recovery runs and getting himself into positions to make tackles or force turnovers. Defensively, United allow Bayern’s defenders to have the ball and then look to press once it goes into midfield. Beckham looks to have the athletic tools to play in a modern, high press system – the likes you might see Pep or Jurgen Klopp employ.
What was interesting was how often United would go back to front quickly, playing quite direct football. In style, they are certainly more akin to a Klopp team than a Guardiola team, and it is effective.
He’s not a natural defender, he commits a fair few fouls in the game with some slightly awkward tackling technique. He shows great energy and desire to get in the position to make tackles though, something many coaches will tell you is more than half the battle when it comes to naturally offensive minded players.
Where Beckham is perhaps lacking is out and out pace. He moves well without being particularly fast over long distances. He is agile however, and quick of thought and foot over the first few (and arguably most important) yards.
In Possession – Technical ability/passing
His use of the ball is superb. Both the execution of his passing technique and his vision to spot the right pass are very high level. He’s able to find the feet of forwards consistently with forward, line-breaking passes, played in at pace. Some of his forward passes are demonstrated below.
What did stand out was how often Beckham found himself in central areas. Sometimes this was due to taking a set piece on the other side of the pitch but often, his deployment in the centre of the pitch was deliberate and enabled him to have more influence on the game. Consistently, throughout the match, Beckham received the ball in that right “half space” or inside right channel. It’s an intriguing proposition, we know he can hurt the opposition with his crossing from wide areas but he can also come inside the pitch and show off his passing range. There isn’t much opportunity for Beckham to display his crossing in this game, in truth. Though I have to say I don’t think this is hindering our understanding of Beckham’s game. I think my Grandma, whose enthusiasm for football matches mine for doing the washing up, could probably tell you Beckham could cross the ball well.
The following clip captures Beckham’s performance nicely, I think. Great energy to recover in defensive transition and force a turnover and two crisp forward passes into the feet of teammates.
His first touch looks clean but there is little opportunity to analyse his ability to dribble, something I was keen to investigate further.
Overall, it’s a pretty good performance from clearly a very good player. It did get me thinking though… in a parallel universe, or perhaps growing up in another country, would Beckham have been groomed to play as a central midfielder? His combination of vision, passing range, agility and excellent fitness levels is an interesting blend for a central midfielder.
BECKHAM AS A CENTRAL MIDFIELDER?
At this point, I have to confess that I’d be lying if I said there wasn’t a slight agenda behind this piece.
Months and months and months ago, I remember seeing a video on Youtube – which sadly, seems now to have disappeared. It was Beckham playing for England in a friendly against Italy, in 2000. He’s deployed as a central midfielder and boy, it’s pleasing to watch. There are shades of Pirlo, as he picks the ball up from the centre backs and sprays it around with ease, with some particularly eye-catching long, raking, diagonals on show. It’s not a perfect performance, and it’s only a friendly match, so you could question the intensity of the opposition defending, but there’s a huge amount of promise there.
Could he do it on a rainy night in Stoke… at a higher level?
Back then, through the archives I went again.
Thankfully, Transfermarkt have a great feature where you can filter by matches played in a certain position (thanks guys). There was one game highlighted as a central midfield appearance for Beckham that was misclassified, and he was actually playing on the right wing, on the video. I can’t for the life of me remember which game that was now to allow them to change it though (sorry guys).
Anyway, I decided to have a look at Beckham at Madrid, in the Madrid derby no less, in a league match at the Bernabéu. We’re in 2003 now, or in old money… blonde ponytail Beckham.
I think a lot of these clips speak for themselves. A selection of Beckham’s forward passes are below, including some of the rare unsuccessful attempts. This isn’t even all of his forward passes , there are loads… honestly.
This is a masterclass from Beckham. A passing clinic. Fantastic to watch. Left foot, right foot, long, short, driven, clipped, outside of the boot, disguised. Everything you’d want to see. And done with a flair and technical ease you rarely see from English players.
One of the things I wanted to get to with Beckham was how capable he was at dribbling. I knew he wasn’t a 1v1 specialist but how good was he in tight spaces? On the evidence of this game he was pretty capable, using body feints and quick feet to evade pressure. As mentioned before, he’s not lightening quick, but he was quick over the first couple of yards, and sometimes that is all you need.
There was is one instance where he’s unaware of the Atleti payer behind him who proceeds to nip in and pinch the ball. But this is not an indication of any kind of trend in the game, he’s tidy throughout in this regard. Some of his dribbles are evidenced below.
Again his fitness levels are apparent. He looked just about this fittest player on the pitch, in this particular game. There are sprints right up until the last seconds. Literally. The last clip in the below is him sprinting off again, trying to add to Real’s lead before the referee blows the final whistle. There is some real selfless running also. On a couple of occasions, he makes a bursting run forward when there is practically zero chance of actually receiving the ball. Of course, he doesn’t get on the end of passes here but what he does do is open up space for teammates to move into. Not all midfielders are willing to do that for the team.
So he’s great on the ball and can definitely run (that’s 2 of Roy Keane’s holy trinity for midfield players) but can he defend?
Defensively, I’m impressed with Beckham in the game. There aren’t a huge amount of defensive actions from him here, but that’s often because he’s run past someone to get in a good position early. There’s one occasion where he doesn’t have the pace to stop a transition that leads to a Torres chance, but this is a one off. In truth he’s rarely in trouble in the game and rarely out of position. In one of the clips in the below, he shows excellent defensive awareness, cutting off two passing lanes and then is unfortunate not to nick the ball before a foul is called. Some of his defensive effort and actions are highlighted below, including the aforementioned instances.
He’s not naturally a brilliant tackler but he’s there to get in a good position early, to put his body in the way, to make it difficult for opposition attackers.
It is just one game. But it’s one game where he is man of the match on the day. And it’s a beautiful performance to watch.
The two questions this blog set out to look at were:
How good was David Beckham?
Could he have been a central midfielder?
The short answers here are: very and yes, at least to a certain degree.
Is this a bit of a Beckham propaganda piece? Probably. This was the guy who had me curling rolled up socks into wash baskets as a kid, providing my own commentary and wheeling away in celebration when one crept off the far post and in.
His enormous celebrity status may have impeded his legacy, I think. And encouraged people to misremember what a good player he was. It would have been a kick in the teeth to go back through the footage and him not to be as good as I had thought. This is not the case. What’s evident is that Beckham was a very good player. Which shouldn’t come as a surprise really, but will be a counter to some who believe his legacy comes more from his haircuts and questionable fashion choices than his ability on the pitch.
The cold, hard, numbers show he offered more goals and assists per 90 minutes than anyone in the data set, bar Robert Pires, over his selected 3-year peak. That’s more than Kevin De Bruyne, Giggs, Figo, Zidane and far ahead of anyone who currently operates as a wide midfielder in the Premier League. The video shows a supreme athlete with fantastic technical ability and vision.
Could he have played as a Pirlo-esque holding midfielder? Maybe. Should he have played centrally given he delivered world class goal and assist numbers from the right? Probably not. Would it have been good to watch? Almost certainly.
Addition: As a few people have asked, the first chart is a single-axis scatter. It’s only looking at G+A p90. There is no variable on the x-axis. The data is just jittered to make it easier to see.
It is also important to interpret these type of charts carefully. It does not mean Walcott is a better player than Luis Figo. It means his G+A p90 output, over the selected timeframe, was higher.
Sancho Quinn, 11/05/20