Only a few years ago El Niño was a six-foot-one blond teenage lightning-bolt that made La Liga fullbacks play on their back foot over the entire 90 minutes of each match in fear of being burned by one of his amazingly well-timed, onside runs that more often than not would end in an embarrassing goal at their expense.
When Fernando Torres was at Atlético Madrid, the then peach-faced teenager terrorised every team in La Liga and everyone knew that it was only a matter of time before he would be snatched up by either Los Blancos, Barça, or some big club in the EPL.
And in 2007, it was Liverpool’s good fortune to do so. The transfer cost to Liverpool was estimated at about £20m, but that included Luis Garcia going to Los Colchoneros as a “make weight” to the transaction. It was a steal!
But that was then and this is now. Then, after three highly productive years at Liverpool—scoring 61 EPL goals—Torres was sold to Chelsea for £50m.
Unfortunately, Fernando’s goal scoring prowess hit a small bump in the road that seemed to coincide exactly with this transfer. The transition to the Chelsea system, with five different managers over four different seasons, has been met with what most would suggest is a resounding failure. Only nineteen goals in almost four seasons!
And to show you how bad things have become, how terribly things have deteriorated over time, a handful of some critical performance metrics essential to the success of any striker will be examined under the harsh light of statistical analysis. Here goes.
Opta performance measures for the three seasons Torres last played for Liverpool will be contrasted with the time he has been with Chelsea. The focus will be on what strikers do: scoring. Attacking skills will take center stage.
I am going to ignore some of the finer points such as dribbling and passing and creating chances—which are, of course, important. But at the end of the day, the focus for premier strikers is putting the ball in the net—and that’s what the entire furor surrounding Fernando Torres has been about. Or, I should say, about the lack of putting the ball in the net.
So we are going to focus what has happened over the past six plus seasons and if his production regarding his goal scoring performance, in terms of the obvious cumulative numbers, and those more subtle measures dealing with efficiency and effectiveness of goal scoring have experienced significant change or not.
Here is what we found.
This is the goal standard for all forwards. There is no substitute. There is no question that a dramatic drop-off in performance occurs at the point of the £50m transfer sale from Liverpool to Chelsea.
The productivity level post-Liverpool is disturbingly low. In fact, Torres has been able to produce only 19 league goals during his first four seasons at Chelsea compared to an average of over 20 EPL goals per season during each of his three seasons at Liverpool. Simply put, he’s operating at less than 1/3 of what he once was. This amounts to a cost of about £2.63m per goal not counting his salary that is reported to be £175,000 per week.
Goals attempts (or shot attempts) is clearly not the same as putting the ball in the net, however it reflects a type of work rate measure for strikers regarding their effort or their “attack ferocity” on goal. So, a metric that determines the frequency of a forward’s efforts of goal attempts is very important. For this article, I’m going to look at the metric of how frequently Torres attempts shots on goal (they may or may not be on target). More specifically, on average, how many minutes between shot attempts?
Again, no surprise here. As soon as Torres leaves Liverpool and goes to Chelsea, the time between each shot essentially doubles. It is clear that he has either become more hesitant in shooting or that his service has, comparatively, “dried up.”
Of course, the differences between the Liverpool 4-4-1-1 system that was perfectly tailored to his style, as compared to the myriad of very different Chelsea formations, could be expected to have somewhat of a negative influence on his productivity. But the question remains, what is a reasonable time lag for any world-class striker to adapt to a new system before his anticipated scoring skills return. Simply put, it never did.
The minutes between goals scored show a similar relationship to the previous metric. There is an immediate and dramatic decrease in performance at the transfer point between Liverpool and Chelsea. In fact, it looks more like a discontinuity as the minutes between goals increases over 800% at the beginning of the 2010-2011 season. Even for the remaining seasons, Torres requires an average of about 3-4 times the amount of minutes to score a goal as when he was with Liverpool.
It’s clear that Torres has developed either a reluctance to shoot on goal or that his fit within the various Chelsea systems have resulted in poorer service. As a result, it may be useful to look at metrics that assess the efficiency or effectiveness of when Torres does attempt goal efforts. The percent of shots on goal measures Torres shot accuracy, and is not a measure of quantity—perhaps this metric is still on par with his earlier career.
Shot accuracy = shots on target/total shots taken
When the data is grouped across multiple seasons by club, there is a slight drop-off in shot accuracy when Torres moves between Liverpool (52.6%) to Chelsea (48.2%).
When the data is analysed, no statistically significant difference is established between these two performance numbers. So, accuracy is not an issue. Keep in mind that shot accuracy is often overrated, strange as this may seem. The reason is that there is no direct “payoff” with shot accuracy: it is a booby prize unless it ultimately ends in a goal. “Close, but no prize” may be a better name for this measure. The great strikers put up volumes of numbers. Some are more accurate than others but at the end of the day, strange as it may seem, quantity plays a more important role than accuracy.
Of far greater importance than shot accuracy is shot conversion rate or shot effectiveness. This can only be measured by the portion of the shots taken that are ultimately goals.
Shot effectiveness = goals scored/total shots taken
In order to do well with this metric, a player has to not only create numbers of attempts but with enough accuracy and quality to end up in the goal. It is the all-inclusive metric for goal scorers.
You can be highly accurate, you can have high quality attempts, but if you do not create enough shot attempts the end result will be poor. This is, arguably next to total goals scored, the most important metric for a striker.
During his Liverpool seasons, Torres’ performance was simply superb: he scored with 26.8% of his attempts. When he transferred to Chelsea he’s averaged about a 13.5% conversion rate of his shots on goal. This latter figure is not a particularly low figure among strikers. Although the stark difference in metrics between Liverpool and Chelsea seem clear, we can formalise the difference with a statistical comparison between the two levels of performance.
A statistical comparison between Fernando’s performances at Liverpool with that at Chelsea shows a statistically significant higher performance (no surprise). The 13.3% advantage in shot conversion rate at Liverpool is reliably superior so that there is little chance it could have happened accidentally—only one in 380. In more classical statistical terminology, Fernando’s conversion rate was significantly better at Liverpool with a p-value = 0.0026.
A shot can be on target but it is no guarantee that it is going to end up as a goal. In addition to being “on frame” the shot must often be well struck and also placed intelligently. If not, an on-frame strike can be weakly hit or well hit but directly at the keeper. This metric is a particularly good indicator of the shot quality.
Shot quality = goals scored/shots on target
How different is the change in performance between the two clubs? The average for Fernando’s Liverpool performance was 50.8% while his Chelsea performance dropped to 27.9%. Think again what this metric measures. It says that the quality of the Torres’s shots that were on-frame were about half as effective with Chelsea as his earlier performance with Liverpool! That is simply shocking. The statistical findings support what you might suspect: the difference is statistically significant with a p-value = 0.0032. That is, the gamble that this difference is erroneous is about one chance in 438. We may, therefore, assume that the findings are highly reliable.
Strangely, it is not that Fernando Torres is suddenly unable to hit the target … that somehow his accuracy is gone and his shots are amiss. There really isn’t that much of a change between now and his heyday at Liverpool. So what has gone wrong?
Fernando is simply not attempting shots at the rate that he once was. And, make no mistake, it is not simply how accurate ones shot is on goal. Daintiness is not the world of the striker. To a large degree, “Give me the ball and let me take my damn shot!” Is the necessarily selfish world of the striker.
Having the opportunity to strike the ball as many times as is reasonably possible, to create chances to score is what the great strikers do. And, to face facts, whether it is simply Fernando’s fault or the combination of Fernando and the myriad of talented midfielders that he has been playing with over the multiple seasons with Chelsea who have trouble connecting with his style of play is not important. It just is not happening.
Nothing has worked for Fernando Torres playing in Chelsea colors. Not for five different managers and multiple combinations of world class teammates. The bus has left the station. For whatever the reasons, the magic chemistry between what was at one time the beautiful play of Fernando Torres has been replaced by the now hesitant, seemingly always in the wrong place, at the wrong time, Fernando Torres.
And it’s probably terribly unfair to continue to anticipate that things are going to change during his stay in Chelsea colors.
More likely than anything, Fernando is not going to be able to really exhale until he is back in La Liga or, possibly, applying his talents in Serie A if the latest rumors bare any relevance. Only time will tell.
Regardless, this tale is a sad one because stories that deal with what could have been always are.
Here’s hoping that El Niño finds his magic again … wherever that may be.
Joel Is an avid football and modern jazz fanatic. He sees the connection between the improvisational elements of each ... the connection between Andres Iniesta and Lionel Messi as well as Miles Davis and Bill Evans. He wrote a weekly football column for the Wall Street Journal Europe between 2010 and 2011 using sports analytics to explain player and team performance. Joel is a professor of Business Analytics at the University of San Francisco, School of Management.
Jan 26, 2015 0