Every time I think of the Roberto Soldado transfer to Spurs I can’t help but think the North London club could have done better.

Perhaps in the modern game money doesn’t really matter and clubs are determined to get their target regardless of the asking price.  However, call me old fashioned, but I like to get value for my money and thus I’ll appraise Soldado in terms of how Tottenham did in terms of Value For Money.

Yes, he scored 24 goals last season in Spain’s top flight and on the basis that Tottenham expect someone other than Gareth Bale to score their goals they needed someone who could put the ball in the net with a little more regularity than Jermaine Defoe and Emmanuel Adebayor have been able to in recent times.

But my over-riding belief is that they could have put the transfer fee of £26.5m to better use in my opinion.

Why do I say that?

He is 28 years old, and he still has a decent few years playing time left but given his age he’s unlikely to improve much and he will certainly have a minimal resale value at the end of his contract.

But that’s not the main reason for my dissatisfaction; my main gripe is the fact that his conversion rates really aren’t overly special.  And special conversion rates are what I would expect from someone who I’m paying close to £30m for, even more so when he’s not being bought for his creativity (with less than 1 key pass per90 last season).

It might seem strange to say that his conversion rate is nothing special given that last season he converted his 100 shots at a clip of 24%.

On the face of it, this favourably compares to Van Persie (18%), Bale (13%) and Lukaku (17%).  In fact in the Big 5 Leagues last season, of the players to score more than 15 goals only Cvitanich, Messi and Higuain had a higher conversion rate than the former Valencia striker.

The Quality of his Attempts

However, regular readers of my pieces will know that we need to provide a little more context than just the player’s raw conversion rates.
We need to compare the conversion rate against what their conversion rates should have been when accounting for the shots that the players attempted.

On this “shot quality” adjusted measure Soldado no longer sits atop of the rankings.  His ExpG Eff rate (as discussed in this previous piece) was 1.01.
Don’t get me wrong, 1.01 means that he finished his shots a tiny bit better than the average player did but it’s pretty damn close to being just an average conversion rate of the chances he attempted.

The reason for the barely average rate when adjusted for shot quality given his astonishingly high apparent raw conversion percentage is immediately apparent when we see the positions of Soldado’s shots:


When I was at school I would have been called a “moocher” if I had those shot locations.
59% of Soldado’s shots were struck from centrally within the penalty area, and he only had 10 shots from outside the penalty area.

In his defence, it is great that he was there to attempt those shots, and he puts them away with decent, nay average, aplomb.

But I feel that the raw conversion rates that he has posted drove the transfer fee to a higher range than it otherwise should have been.  He didn’t score any more goals than the average player would have done if they had been lucky enough (or perhaps good enough) to be presented with those opportunities.
Credit undoubtedly needs to be given to him for being in the right place at the right time, but all I’m asking for is that the location of his shots are taken into account when attempting to analyse his shooting performance from last season.

The Extent of Soldado’s chances

The measure that Constantinos Chappas and I have created that objectively rates the quality of chances that a team or player is presented with is the Average ExpG.  This quantifies the probability of the average shot taken by the team or player in question.

According to my records, 102 players in the Big 5 leagues scored more than 10 goals last season.  Rather interestingly, Soldado’s Average ExpG value is 2nd highest out of those 102 players.  The quality of the chances he attempted was second only to Pazzini of Milan.
Just let that settle in for a second……. The chances he attempted were from the 2nd most attractive locations in the Big 5 leagues.

That fact is ultra-important but it has generally been overlooked in most of the analysis of Soldado’s transfer that I have seen.  And that is the primary reason why if I ran a football club I wouldn’t have spent £26.5m on taking him from Valencia.

I read with interest the suggestion made by Ted Knutson in his preview of Tottenham on this site that, in a worst case, the acquisition of Soldado may ensure that Spurs get a much more interest Adebayor.
I guess that may just be true, but if so, you would have hoped that Spurs could have come up with a cheaper performance management system!!!

As a general point, I want to make the readers aware that I would have loved to be able to go back and include data from before the start of last season and see what his shots were like then.  But unfortunately I am limited to the data that Squawka provides us with, and that is only one season at this stage.

Tottenham’s Style of Play

One other thing that I’m not totally at ease with is how Soldado’s style appears to be polar opposite to how Spurs has played in the recent past.
You can see that Soldado thrives on putting away very good chances.  I wonder if he is aware how Spurs played football last season?

He is moving to the only team in the Premier League who took more than 50% of their shots from outside the penalty area.  Working the ball into areas that most people would feel are good shooting opportunities certainly wasn’t top of Tottenham’s “How to Play Football” manual last season.

Perhaps his signing suggests that Gareth Bale is indeed leaving White Hart Lane this summer and thus Soldado is more likely to receive the sort of service that he has been used to.
However, in the event that Bale does stay with Spurs for another season I would fear that Roberto Soldado could cut a rather frustrated figure as he ducks to get out of the way of Bale’s long range  efforts.

Wrap Up

OK, so it’s easier to criticise an action than it is to do something positive yourself.  So after claiming that Spurs could have done better in their search for a striker I think that it is only fair if I suggest a few alternatives as to how Levy’s £26.5m could have been better spent.

In my next piece I will outline a pool of strikers that I would procure from if I was tasked with doing so.

  • David Nielsen

    Hi Colin

    Again a very good piece. I truly enjoy reading these “number-crunching articles”, and I find them to be very informative and horizon-expanding. As a coach myself I have to make a point clear though:

    Given the fact that Soldado (or any other player for that matter) takes most of his shots from “good” areas – would I be wrong to suggest that precisely that makes the asking price increase? If for example I was to play for Valencia last season, I seriously doubt that I would have been able to fire a 100 shots on goal let alone from such good positions. If I had been able to I might just have had a better conversion rate than Soldado, but – and this is my point – getting into those scoring positions (holding off defenders, making well-timed runs, cutting ahead etc.) is just as much what you pay £26.5m for. “Any fool” could score if presented with only penalty shots. But in the real world – all statistics and numbers aside – you have to work your socks off to even get the chance of shooting at goal. Soldado might just prompt Spurs into taking more shots from inside the area, and maybe that is part of the plan AVB has for bringing him into the side.

    If you fully wanted to investigate the value-for-money on this deal, I think you also need to look at how often does he actually work himself into positions that are good scoring positions. ‘Cause in the end it all comes down to the actual figure for how many goals scored. Soldado might just prove the missing piece in Tottenhams puzzle. He can get the shots from the positions they were missing last season. And that might just be what propels them into the top4 even if his conversion rate isn’t that good. That he has the 2nd best overall position for taking his shots in all of the top5 leagues is also worth paying big money for.

    Looking forward to your next article. I am very intrigued to find out, which strikers you would rather have spend that sort of money on.

    Best of wishes and once again thank you for expanding my horizon


    • Colin Trainor

      Thanks for your constructive comments.

      What you have said seems reasonable and you have made some very fair points. However, I just don’t know how to test what you are saying, or indeed quantify how much of a skill getting in a good position is, as the data to do this is way out of my reach.
      Hence why I haven’t tried to incorporate those soft factors into my analysis.

      All I would say is that on our ExpG Eff rankings the cream does tend to rise above a mark of 1.00. amd Mr Soldado didn’t manage that last season.

      • David Nielsen

        Hi Colin

        Thanks for your answer. A fair point if you don’t have the numbers. I don’t know quite enough about the numbers, but it seems to me that number of shots taken in prime positions would make quite a good indication as to whether a player is good at what you call ‘soft factors’…? And as I mentioned before – he DID manage to score 24 times. And even though his finishing might not be extremely good, he does seem to get into positions that would almost guarantee, he could replicate that number.

        Well being a Man Utd man myself, I obviously hope, I’m wrong… 😉

        Keep up the good work. Loving it!

        • Colin Trainor


          As you’re a Man United fan; Chicarito had an ExpG Eff of 1.08. Pazzini the player with the best shot locations also earned an ExpG Eff of 1.08. Then you have class acts like Higuain who posted a 1.33 last season. Those are people who are typically getting shots in the same areas as Soldado but posting superior finishing figures.

          I have 9 players who posted better ExpG Eff figures than Soldado in the Big 5 leagues last season despite having similar shot profiles, ie a minimum of 3 Shotsper90 and less than 20% of their shots coming from outside the area.

          Not all of them are for sale but Soldado trails them in terms of scoring the opportunities presented despite the similar shot profiles. That’s kinda the crux of the article.

          • David Nielsen

            OK. Thank you for clearing that up for me.

  • Brett

    This post seems entirely contradictory to this post:


    • Colin Trainor

      That’s probably because the two articles are written by 2 different writers Brett.

  • http://None Zebra

    Good analysis, but not sure I agree that Soldado’s positions aren’t almost entirely skill based. If you watch him play, then you’ll see his movement in and around the box is really spectacular. I agree that there was better value, particularly in the form of someone like Luis Muriel

  • Novice

    Hi Colin,

    Thanks for posting this, some very interesting thoughts and bringing a new angle for looking at strikers. I’m really fascinated by the work you have done on ExpG have a few thoughts/questions that I wanted discuss with you.

    1) Have you had look at how well the Expg eff rate of players tallies with clear cut chance conversion rate? In my mind this is a more objective measure of how clinical a particular striker is when presented a chance, because of the subjectivity involved in what constitutes a ccc.

    2) Reading David’s comments got me thinking: whilst expg eff rate is potentially a very good indicator of one facet of a player’s game it does not tell the full story. Given how little time a articular player actually spends on the ball, 95% of a player’s work is done off the ball and his movement, in my view, is equally important as finishing ability if not so:. being a good finisher is not enough if you can’t work yourself into goal scoring opportunities (the opposite of this is perhaps Giroud or Gervinho). Have you looked at ExpG per 90 min played? i.e. how often does a player get into himself into the good goal scoring positions. Whilst this is not perfect because it also depends on his teammates getting the ball to him in that position, it maybe a good proxy to judge the quality of a player’s movement. (and perhaps there is a way to adjust by taking into account the passing accuracy of the team).

    Thanks again and look forward to hearing your thoughts

    • Colin Trainor

      I’m glad that the articles are proving thought provoking; and it’s not just one way. The reason why I am sharing many of my findings is because I am hoping to learn a little bit more and fill in some of the gaps that I haven’t been able to cover myself.

      I have never looked at CCC as I prefer not to deal in subjective measures, and I agree with you that I prefer our methof od deciding on the quality of a chance.

      I’ve had some thoughts on your second point, and whilst I certainly welcome all such new ideas I just think that a player’s ExpG per90 is determined so much by the team that he plays for. In my mind, the weighting would be much more in favour of the team than the skills of the individual player.

      Again, I’m not sure how we could go about measuring this. You would need to see how the average ExpG per90 would fluctuate for 2 players who play the same position against the same teams. Such analysis wouldn’t be very practical.

      However, you’ve got me thinking. Perhaps what could be done is measure the ExpG per 90 for Demba Ba and Fernando Torres during their spell at Chelsea last season. Both players played the same position and for the majority of the time it was one or the other. So they strike me as the perfect test scenario to see if ExpG per90 is drivem by team or player.

      Ideally I’d like more than one example.

      Is anyone aware of any other player combinations where it was either one striker or the other at a club last season?

      • Mark

        Adebayor and Defoe at Spurs

  • Pingback: StatsBomb | Premier League Shooting – The First 3 Weeks()

  • Pingback: AVB – człowiek, którego przerosła Anglia | kolejnynudnyblogopilce()

  • david

    Really interesting article, I now want more!

    “…In my next piece I will outline a pool of strikers that I would procure from if I was tasked with doing so.”

    Could you tell me where this next piece is?

  • david

    Already read that article and many others. Follow you on Twitter already too. Great stuff!

  • Pingback: StatsBomb | Has Britain Got Talent? Is A Lack Of Data Holding Back British Football Clubs?()

Improve Performance and Productivity in Your Club:
State-of-the-art Football Analytics