A goal by Theo Walcott is always a memorable event.
So many of his tallies live forever in our memories because of their breathtaking, straight-to-Youtube form, but they are even more to be cherished for their rarity.
Like a bizarre avian mating ritual to an ornithologist or a rare bloom to an orchidologist, Arsenal supporters (Arsenalogists?) treasure Walcott goals not least for their anti-quotidian quality.
In 133 appearances for Arsenal (72 as a starter and 61 as a substitute) Theo Walcott has netted only 18 goals. Rounding generously, that's a strike rate of .14. He has also tallied 20 assists for the Gunners, giving him a combined 'points per appearance' of .29.
Though he has provided highlight after highlight, Walcott's goal and assist tallies, by any statistical measure, are underwhelming. His potential to collapse defenses, create impossible matchups on the wing, and move behind opponents' flanks to deliver incisive crosses are there for all to see, but his actual rate of production does not match his hype.
He is 21 years old. He has been a phenom for five years (four of them with the senior Arsenal squad).
Among players roughly his age, Cesc Fabregas and Alex Song are less than 22 months his senior, but are important starters. Nicklas Bendtner and Denilson are less than 14 months older, and both have been important contributors for the past two seasons.
Denilson—ostensibly a defensive midfielder—has 11 goals and 11 assists in fewer appearances, yet he is the constant subject of fans' ire and impatience.
Bendtner, despite his obvious shortcomings and often uneven form, has blossomed into a remarkably efficient target man. In his last 80 Arsenal appearances, he has tallied 27 goals and 10 assists, while proving that he can contribute either as a center forward or a winger.
Walcott certainly takes more than his share of criticism, but we still watch, wait, and buy his kit—whatever the number on the back. Everybody expects that the next season is going to be the breakout season and we all want comfortable seats on the bandwagon when the lad finally bursts forth with 15 or 20 goals.
But will that ever actually happen?
Substitute or Starter?
Yesterday, while basking in the glow of another goal from the Newbury youngster (not yet aware of the horrors that awaited), I decided to look a little deeper into the numbers Walcott has compiled in his tenure at Arsenal.
I was hoping to determine whether Walcott is best suited to the bench role he has assumed for much of this season, or if there is perhaps an argument in favor of his seeing more of the pitch in order to improve his strike rate and overall contribution.
I examined Walcott's Arsenal goals, assists, and average minutes played per appearance since the 2006-07 season and sorted them according to whether they were recorded as a substitute or as a starter.
I opted not to use strike rate, preferring to add goals and assists together to get 'points,' because in his role as a winger, Walcott's ability to set up others' attempts is just as important as his ability to convert his own. Also, it enabled me to work with a larger, more significant statistical sample.
In total, Walcott has made 61 appearances as a substitute and 72 appearances as starter for Arsenal. In those 133 appearances, he has tallied 38 points (18 goals and 20 assists), which means he averages about .29 points per appearance.
By way of comparison, Nicklas Bendtner's average is .38 over his Arsenal career, Eduardo's is .42, and with 71 goals and 36 assists in 194 appearances, Robin van Persie sets the current gold standard at .55 points per appearance.
Those comparisons should seem unfair.
Walcott's ideal position is different from that of the tall strikers, Bendtner and van Persie. Also, almost half (45 percent) of Walcott's appearances have been off of the bench.
Additionally, as soon as Bendtner and van Persie became regular starters, their strike rates and points per game improved dramatically. That would seem to suggest that Walcott might follow a similar trajectory if given more starts.
Walcott's Statistical Anomaly
After comparing Walcott's numbers as a starter to his numbers as a substitute, I do not think more starts and more minutes are the spark Walcott needs to evolve into a more productive and more consistent player. If anything, he actually needs fewer starts and fewer minutes, however counter-intuitive that may sound.
In the three seasons from 2006-07 to 2008-09, Walcott started progressively more matches and logged more minutes per appearance.
However, his point per appearance flat-lined, even in 2008-09, when Walcott started 27 matches for the Gunners and came on as a substitute only eight times. This failure to immediately make giant strides is not news in itself.
What is truly remarkable is the relationship between Walcott's points per appearance as a starter and his points per appearance as a substitute.
It is logical to assume that a player is more likely to score a goal or tally an assist in a match in which he starts and plays over 75 minutes than in a match in which he comes on as a substitute and only sees the field for 10-30 minutes.
In fact, over a large enough number of appearances, a striker should be two to three times as likely to score in a match he started than in a match in which he appeared as a late-game substitution. Three times as long on the field means three times as long to score.
As young players in the Arsenal system, both Bendtner and van Persie exemplified this intuitive trend.
In his first three seasons at Arsenal, van Persie tallied 34 points in 64 appearances as a starter (.53 points per appearance) and 12 points in 46 appearances as a sub (.26 PPA). Thus, he was a little more than twice as likely to contribute as a starter than as a substitute.
Similarly, in three seasons since '07-'08, Bendtner has totaled 41 points in 65 starts (.63 PPA) and 12 points in 55 appearances off the bench (.22 PPA). He, therefore, has been almost three times as likely to score or assist a goal as a starter than as a sub.
Walcott does not follow this logical trend.
In fact, he shatters these expectations.
In his 61 appearances off the bench, he has tallied 16 points for an average of .26 points per appearance.
In his 72 matches as a starter he has tallied 22 points for an average of .31 points per appearance.
These figures are statistically similar, meaning that, over the course of his Arsenal career, Walcott has been just about as likely to tally a goal or an assist in a game in which he played 20 minutes as he is to contribute in a game in which he played 80 minutes!
Lest anyone suggest that these numbers are skewed by his first couple seasons of lower level competition and shorter league appearances: Over the last two seasons, Walcott has averaged .26 points per appearance as a substitute and .23 points per appearance as a starter.
His numbers as a sub have remained about the same, therefore, while his numbers as a starter have actually declined despite having more starts and increased time on the field in those starts.
Accounting for the Unexpected
Walcott does not follow the expected pattern. For whatever reason, he is just as likely to contribute as a substitute as he is to contribute as a starter. This shouldn't be.
Possible explanations for this curious statistic abound.
Substitutes are given specific directions by their managers and enter the field fresher and fitter than the men who have been playing for an hour or more. For Walcott, who is already faster than just about everyone else on the pitch, these two ingredients could combine to give him even more of an edge than the average substitute.
Another explanation could be that, as a young, still-maturing player, Walcott is somehow better able to motivate himself in short concentrated bursts than over the course of 90 minutes.
Or perhaps, like many talented professional footballers, Walcott's ego can't tolerate being relegated to a bench role so he feels the need to prove himself worthy of a regular starting role.
Then of course, it could be that the tail is wagging the dog in this case. Arsenal as a team have a greater-than-average tendency to score late goals. This means that anyone who is on the field in the final 20 minutes has an increased chance of participating in a goal-scoring play and that a player who leaves the pitch after 70 minutes is losing a prime opportunity to pad his stats.
Whatever the internal mechanism within Walcott that drives him to do better off the bench, or whatever the environmental factors that make him more efficient in such a role, Walcott is more effective as a substitute than as a starter.
He averages about 25 minutes per appearance as a sub, and in each such appearance he has about a one in four chance of scoring or assisting a goal. This in itself is a respectable number.
For a regular starter to only tally a goal or an assist every four matches would be mediocre, but for a substitute, it's as good or better than van Persie and Bendtner.
Walcott is the archetypal super sub. Why force him to be an uneven performer in an overambitious strategic role when he is demonstrably a clutch performer in a more limited tactical role?
If England fans, Gunners, and members of the press would simply scale back their expectations and realize that he is never going to be a 25 goal-scorer, Theo Walcott could perform a key, but limited role for club and country.