2014

The Spanish Inquisition - Roberto Soldado

By Will Humphrey | March 23, 2014
2014

The Spanish Inquisition - Roberto Soldado

Once the last ball of the season has been kicked and a few hundred Premier League footballers have flown off to exotic locations to rest their battered feet, I'd like to think club staff get themselves round the table for a season debrief. They'd discuss what went right, want went wrong and what went........Mmmkay. Come May, this process might take a little while longer at Spurs than it would at other clubs. The league position isn't a disaster by any stretch, but player trades on the whole haven't worked out. Sooner or later, the guys at the table (after a long time sidestepping it) would stumble upon what was maybe the crowning turd in the transfer water-pipe. Before the season started, Colin Trainor, voiced concern about the transfer of Roberto Soldado here. Taking penalties away, Soldado has scored 2 league goals this season compared to 19 last. I'm going to try and build on Colin's work by asking the kind of simple questions I'd be asking if I was sat down at that table, discussing the reasons why its all gone paella-shaped. Is Soldado taking less shots? This season he's taking 2.7 shots per 90 mins. Last season he was taking 3.1 shots per 90 mins. That's an extra 15 shots per season if all minutes were played out. He's taken 52 shots this season compared to 100 last. None of this explains the huge difference in goals we've seen. How many shots is he getting on target? Is he getting into the same areas to take them? It's nice to visualise this one. The deeper the red on the graphic below, the bigger the volume of shots on target from that zone:

SoldadoSOTSComparison

The graphic quite clearly demonstrates part of the problem. Same areas, but a drastic reduction in volume. Soldado is a penalty box striker, pure and simple.

So how did Valencia 2012/13 compare to Spurs 2012/13? Was it a complete style mismatch when they bought him?

The graphic below shows the volume and type of chance the two sides made in the penalty area last season.

Valencia Spurs Comparison 1

I didn't see much of Valencia last season. I'm not sure what I was expecting to find but it wasn't this. The two teams were almost identical in this regard.

 Ok, so what about Valencia 2012/13 versus Spurs 2013/14? Have Spurs done something different?

Valencia Spurs Comparison 2

Once again Spurs are well on track to at least equal Valencia in every department bar set piece deliveries.

So if he's getting in a similar amount of shots per 90 as he was before, he's just not putting them on target as much, right?

Right. Roberto's gone from getting 43% of his open play shots on target last season, to just 23% this season. It's a whopping drop but it's fairly common as you can see in this cracking piece by @willtgm. Will's work suggests that this drop is nearly as likely to be luck or other factors influencing play as much as it's down to the skill of the individual involved.

Those other factors. The teams as a whole might have been similar in style and volume, but maybe Soldado himself isn't getting on to the same type of chance as last season?

This is what that looks like:

SoldadoComparisonTypes This is pretty much bang on identical too if you take minutes played into account. There's a small drop of passes played into him from in the hole. However, as he didn't score from any of these last season, it's difficult to say it's a problem. The only thing is, this is how both his open play goals for Spurs have come about. Does it take less shots on average to score in La Liga? It does, but the differences aren't significant. Definitely not significant to explain the Soldado puzzle. Might that say something about defensive pressure on the ball? It might, but the data for this isn't publicly available. The video of Soldado's goals is here: http://www.youtube.com/watch?v=Xj-XhTAmvag Looks like a mix here. On some occasions there's time and space to finish and on some occasions it's a tight spot with good movement and finishing on show. The way Spurs played under AVB (high press, hemming opponents in their half) might look a little different to this. Spurs didn't build quickly and as a consequence opposition defences were fairly set. However, under Tim Sherwood things are different and it still hasn't happened for Soldado.

How would the average shooter in the Premier League do with the same volume and type of chances that Soldado's had this season?

My expected goals model suggests 4 goals should have been scored in open play with the shots on target he's had. That would be the benchmark. The Spaniard is therefore only 2 goals behind where he should be. Simulating these shot situations thousands of times over means we can put a measure on how good or bad that performance might actually be.

SoldadoSim

We can expect a Soldado-esque performance just over 10% of the time. Considering there's only a 20% chance you'll actually get the benchmark performance of 4 goals, Soldado's efforts might not sound so bad. However, the reality is we're looking at an 83% probability that your benchmark shooter would outperform him.

 None of this is really explaining it. Has he just been unlucky?

I'll be honest. it's not a term I like very much. When it comes to what happens in football, everyone has a reason for everything. However, when it comes to goalscoring, the longer I delve into the numbers the less reason I'm finding for stuff that happens. My studies into goalscoring so far suggest that even the 'best' strikers in the Premier League come back to mean over time. Is this simply what Soldado is doing? Here's his rolling performance starting from his last season with Valencia until now (this is far back as the publicly available data goes):

Soldado Rolling

Looks like a case of him regressing back to where he should be to me. Will's work on shots on target (linked above) is fairly important. Next season Soldado might just start getting shots on target again and probably getting goals. The staff at the club could no doubt start watching hours and hours of footage and maybe put their finger on the reasons why. I'm sure something could be found and they'd be happy with the explanation. Increasingly, my explanations around goal scoring sound a lot like: sometimes shit happens.

Some nice analytics could probably tip the odds slightly in your favour over time. You should stop looking for certainties - they don't exist. No one wants to hear that message. But if someone like me can get to grips with it, so can you.