2013

Chelsea's Striker Options

By Colin Trainor | August 1, 2013
2013

Chelsea's Striker Options

Given the huge amount of attacking talent currently residing at Stamford Bridge I wonder how Jose Mourinho is going to decide on which four attacking players he will field.

Formation

I assume that he will use a back four and will play two holding / central midfielders which will then allow him four out and out attacking players.

The widely held belief is that he will play three attacking midfielders, link men or “just off the shoulder” forwards.  Those 3 positions will probably be filled by some combination of Mata, Hazard, Oscar, Moses and the new signings of Schurrle and De Bruyne.  And the attacking talent has been assembled even before we consider the possibility of Wayne Rooney signing for Chelsea.

Such a formation would then leave room for just one traditional striker, and at the moment it would seem that this position will be contested by Lukaku, Torres or Ba.  It would appear that this position is Romelu Lukaku’s to lose but I wanted to take a look at the three strikers’ stats as well as visuals of their shot locations and placements from last season to see if this is indeed the correct decision.

Striker Options

Romelu Lukaku seems to be holding pole position in this battle right now, but at just 20 years old is he ready for such weight and responsibility to be placed on his shoulders?  Yes, he had a terrific season last year but despite the fairly large transfer fee Chelsea paid for him (£19m) perhaps he was something of a surprise package to the defences he came up against last season, might they be better prepared this season?

Demba Ba didn’t have a great first 6 months at Chelsea, in fact it went pretty awful for him with just 2 goals since his move in January from 46 shots.  That’s the sort of conversation rate that makes the current Fernando Torres look, well, like the Fernando Torres of old.

Torres doesn’t need me to write much more about him, suffice to say it appears that El Nino’s best days are well behind him at this stage.  Although the fact that he played in approximately 75% of all Chelsea’s available minutes last year suggests that Roman Abramovich may not feel the same way. At this stage it does look like his time at Chelsea is running out as there has been a lot of chatter concerning a return to Spain.

To help put some context on how the 3 Chelsea strikers performed last year, I thought I would take a look at their performance from a statistical point of view.

Player Statistics

 

The above stats are for the entire 2012/13 Premier League season, so Demba Ba’s figures include both his time at Newcastle and Chelsea.

All the figures, with the exception of ExpG and ExpG Eff, should be both obvious and well known to readers of this post so no explanation will be necessary.

Lukaku’s Per90Shots on Target value of 2.11 is pretty special and at more than 4.3 ShotsPer90 he certainly kept defences busy.  Demba Ba was even more impressive with the amount of shots he took but unfortunately for him he lacked a little accuracy which then reduced his SoT value. Torres’ numbers are really subdued.  Despite playing more minutes than Lukaku and Ba he had substantially less activity in all outputs (shots, shots on target and goals) and he rounds it off with just 1 shot on target per90, which is a very poor return for a top level striker.

ExpG

The new metric introduced in the summary box, ExpG , is the number of Expected Goals that we** expected a league average player to score based on the type of chances that the players attempted.  The inputs to this measure won’t be disclosed, but we find that it is fairly accurate and allows us to compare the quality of chances created and then the efficiency with which they were finished.

The ExpG Eff metric is  = Actual Goals / ExpG where an ExpG Eff of 1 represents an average player, a value greater than 1 represents above average finishing and less than 1 below average

**We refers to Constantinos Chappas and I. Constantinos can be followed on Twitter @cchappas

From a Chelsea viewpoint it is perhaps worrying that Lukaku is the only one of the trio whose actual goals tally exceeded their ExpG value.  So whilst the finishing skills of Torres and Ba were very poor, with an ExpG Eff of 0.73 and 0.88 respectively, even Lukaku’s 1.05 (as the best of the trio) was not exceptional by Premier League standards.

As a means of comparison; Van Persie finished the season with an ExpG Eff of 1.15, Walcott 1.40, Berbatov 1.19 and even Suarez earned 1.08.

In fact, of the top 12 Premier League scorers last season only Dzeko (at 0.84) had a worse ExpG Eff ratio than Ba and Lukaku. Interestingly, Wayne Rooney who has been strongly linked with Chelsea this summer doesn’t look like he’ll be the answer to their lack of a clinical finisher either as he posted 1.06 last season.

Shot Visualisations

In order to provide the bare statistics with some context I had a look at the shooting locations that the players were faced with and the placements of their non-blocked shots.

Torres

TorresShotsFFS

The shot location images I use in this piece have been taken from the subscriptions section of Fantasy Football Scout website.

I certainly wouldn’t encourage players to take speculative, often wasteful long range shots, but the almost total absence of long range shots for Torres appears indicative of a player that is very low on confidence.  He also struggled to hit the target (green dots) from many shots that were outside the width of the 6 yard box.

 

TorresPlacements

The above image shows the shot placement from the striker’s Point of View with the red balls signifying goals.

Looking at the shot placements it would appear that Torres strongly favours shooting toward the right side of the target.  Aside from that there was an unhealthy attraction towards the centre of the goal.  His lack of accuracy and the amount of easy saves that opposition keepers were allowed to make would have contributed to his awful ExpG ration of 0.73.

Demba Ba

BaShotsFFS

 

We can see a lot more activity on Ba’s image than the Torres one, with a particular penchant displayed  for attempting long range efforts

BaPlacements

On the whole, Ba seemed to have two types of shots.

Most of his on target shots tended to be very low ground shots, which at least is preferred to shots that arrive at the goalkeeper a few feet off the ground.  However, he seemed to lack appropriate accuracy control when he attempted to put some elevation into his shots.

Lukaku

LukakuShotsFFS

Lukaku’s shooting appears to be the happy medium between Torres’ lack of activity and Ba’s overzealous shooting.

He has a decent smattering of long range shooting, but the highlight of that image for me is that he displayed great skill in ensuring that shots from the right side of the pitch generally hit the target.  Undoubtedly this is due to the fact that he favours his left foot and thus the right sided shots give him the best angle, but the amount of green dots on that image is admirable.

 

LukakuPlacements

If I was being critical of Lukaku’s shooting its that he fired too many shots toward the centre of the goal at heights that were favourable to the goalkeepers.

A rough count gives me 19 shots in the central region that didn’t stay along the ground, and only 2 of them were scored.  That shooting pattern will certainly reduce a player's conversion percentage rate.

Perhaps that might explain why although good, the Belgian youngster’s actual goal tally compared to his ExpG was not exceptional by Premier League standards last season.

Conclusion

Based on the statistics from last season and the three strikers I have considered I don’t see any reason why Lukaku shouldn’t be the starting centre forward for Chelsea this season. Torres can be discounted entirely.  His finishing of the chances he had was very much below par, but this is compounded by the fact that he didn’t get himself in the position to be taking shots anywhere near often enough.

Ba just didn’t do enough last season to suggest that he is ahead of Lukaku.  Yes, he had more shots but his average ExpG per shot was 25% less than Lukaku.  The lower average shot ExpG is caused by attempting more difficult shots which suggests that Ba was less prudent in his shot selection. This also comes across clearly in their shot location maps.

As a result of Ba’s more speculative shooting, Lukaka posts better Shots on Target and Goals per 90 than Ba.  But the clincher for me is that Ba didn’t even convert his chances at the average player rate of 1.00 wheras Lukaku slightly exceeded that threshold (1.05 vs 0.88).

It will be interesting to see how Lukaku progresses this season.  There is no doubting that he is a handful and he should improve considerably with maturity, but he will need to. In my opinion, a club with the expectations of Chelsea should have a main striker who is capable of putting away their chances at a rate that vastly exceeds that of a league average player.  Perhaps Lukaku will develop into that player, but if not, it’s important for Chelsea that they have someone playing at the top of the pitch who can.