Do the numbers match fans' perceptions?
When I first started this website back in February last year, I did so with the intention of putting together lots and lots of statistic-based articles which would hopefully shed some light on tactical issues, as well as providing people with something a bit different to read. Of late though, I've been slack - partly due to work commitments eating into time I would otherwise have put to use researching articles, and partly because Dan over at www.wearepremierleague.com does a much better (and more thorough) job of it.
That's not going to stop me though. I haven't been lost in numbers for some time now, and if I'm being honest I miss it, so it was with this partly in mind that I thought I'd put together a blog looking at how our central midfielders compare statistically. I actually did a couple of similar pieces last April looking at Leon, Ki & De Guzman (links here and here), so it'll be interesting to see how our changes in playing style have been reflected in terms of raw data when compared with last year's statistics.
Without further ado then, let's get into it. Last season, our central midfield consisted primarily of Jonathan De Guzman, Leon Britton & Ki Sung-Yueng, who were essentially vying for two spots given Michu's preference for playing behind the frontman (who for a good swathe of the season was Danny Graham). This season we've seen Ki loaned out, leaving De Guzman & Leon with new signings Jose Canas & Jonjo Shelvey for company in the centre of the park. Michu played a couple of games up top early this season and has since missed out a lot through injury, so I'm going to concentrate on the four we'd consider "central midfielders".
Let's start with passing. We do like a pass in SA1 don't we? Leon & Jono's pass completion is pretty much the same as last year, Leon's down a percent or so to 90.7% and De Guzman's marginally better than last season's at 89.6%. The two newcomers to the team this year figure pretty much as you'd expect - Canas's is almost identical to Leon's at 90.6% with Shelvey a little lower at 84.1%. Given he enjoys more of an attacking role that's to be expected, and he also has the least average passes per game with 52.2, with De Guzman averaging 60.3, Canas 61.4 and Leon 64.8. He was always going to have the most wasn't he? Interestingly when compared with last year's Leon and De Guzman are actually playing about ten more passes per game - a sign of how much we've been retaining possession this year even if on occasion we've struggled to do anything with it.
That takes us neatly to chance creation. While you'd expect Canas and Leon to boss the passing charts, you'd probably fancy Shelvey & De Guzman to come out on top when it comes to passes in the final third, and it works out fairly predictably. Leon plays 0.3 key passes per game, Canas 0.1 (ouch). but Shelvey manages 1.3 De Guzman 1.8. Goals and assists are also an area you'd expect there to be a disparity, and you'd be right - Canas and Leon have no goals and assists between them. Shelvey, however, now has five goals and three assists while his Dutch teammate has one goal and four assists. You don't score without shooting, and that's neatly reflected in the stats too - Jonjo has 2.3 shots per game, Jono 1.6, while Leon & Canas come in at 0.3 and 0.1.
In terms of ball retention, again what most fans would have suspected is reflected in the raw data. Leon & Jose Canas are both dispossessed 0.7 times per game, but De Guzman is over double that at 1.5. Jonjo comes in with a nice round figure of once per game, and that's impressive given how often he tries to beat his man. As for losing the ball through poor control it's fairly even between Jose (0.8 - and better than Leon!), Leon (0.9) and Jonjo (1.0) but again De Guzman comes across a little loose, gifting possession on average 1.5 times per game. Still an impressively low figure, mind.
Defensively, Jose & Leon boss the tackle stats and given for the majority of the season only one of them has been selected, that means the other central midfielder hasn't really been gaining possession back that much. Leon averages 1.9 tackles, Canas 1.8 while De Guzman comes in at 1.3, Shelvey 1.1. I guess they are up the field a lot more, but we're already seeing the benefits of playing with two defensive midfielders so I expect the midfield stats to tighten up a lot in the near future. Interceptions too is another area with a disparity - Leon averages 1.9, Canas 2.1, but Jonjo and De Guzman only make 0.8 & 0.9 respectively.
If I'm being hones there isn't as much to read into these statistics as I thought there would be. I've been critical of Jonathan De Guzman in the past for sloppy play but the statistics simply don't relay that - more often that not he retains possession when passing, even if he is mugged more occasionally than any of our other midfielders. I do feel though, that as I said in the Liverpool review if he is willing to run until it hurts, he could be some player. I struggle to recall too many times where I've seen him busting a gut to get forward in particular, whereas all the other players we've looked at here can never be accused of not putting in a shift.
What is interesting to me is we're obviously playing more passes on average. Given the increase in calibre of the squad if Monk can get the team playing the same football we were at the end of Rodgers' tenure/the beginning of Laudrup's then it's automatically going to be football of a higher standard given the players on the field are capable of that much more.
It's good to see Monk appreciates how important it is to control the centre of the midfield, and if the next year goes as his first few games in charge have I'd expect the passing stats to be even more impressive if we were to look at them again next February. Whether De Guzman is still here then remains to be seen as his second loan deal expires in the summer, but despite how often I get frustrated with him there's clearly a quality footballer in there somewhere and I'd love for him to kick on at the Swans.
Canas is someone else I'd like to give a mention as he's bedded into the side so well. Since signing from Betis he's looked every bit a Swansea City midfielder, but it's infuriating we've had to wait this long to see him paired with Leon Britton regularly when the possibility of such a great partnership was just so obvious. The Spaniard has taken to life in SA1 like a duck to water and I'm expecting a great second season from him, as he continues to settle in and his English improves.
Overall, this just shows that if you put statistics in context you can in fact infer a fair amount from them. There are useless statistics, sure, and there are statistics taken completely out of context, but if you compare these statistics with your average Swansea fan's opinion of the four midfielders we've looked at it's likely they'd come to the same conclusions. For me, the foreseeable future should include Leon & Canas at defensive midfield with one of De Guzman, Shelvey, Michu, Pablo or Pozuelo ahead of them. Get Ki back in the summer and we're covered at defensive midfield, and then it's simply a case of figuring out who's better ahead of them.
It should be fun finding out!
Below are the stats I've collated from www.whoscored.com for this article, as well as statistics from April last year (bottom) for ease of comparison.
