I’m (not quite) a believer – looking behind Arsenal’s shot numbers
Anyone who follows me on Twitter will know my stance on Arsenal’s title hopes – I’ve been unconvinced. A gut feeling from history points to them having a collapse at some point in the New Year, injuries could well trouble them again, but more importantly I was also skeptical of their good shot numbers, a way to gauge underlying performance.
As I explained in more detail here (http://everyteamneedsaron.com/2015/11/21/why-i-dont-think-arsenal-are-title-contenders/), after 11 games in the season they were converting their ‘statistical possessions’ to ‘statistical attacks’, and ‘statistical attacks’ to shots at very high rates, as well as preventing their opponents turning statistical attacks into shots. These are, I think, inner workings behind the curtain of shot numbers, and in these three categories, compared to a four season sample (Bundesliga 14/15, La Liga 14/15, Premier League 14/15 and 13/14), they came 5го, 2й, and 4го.
Basically, even the Barcelonas and Bayern Munich’s of the world don’t seem to do this well across the board, and the current Arsenal team is certainly not as dominant in their league as these other giants are in theirs. At the time, their shot numbers were leading to analysts rating them as serious title contenders alongside what was, at the time, a very dominant Manchester City. Moving the ‘chain of conversion’ backwards however, away from only shots and goals, I was unsure about them (though my gut was also playing a part in this).
They have just beaten a City side which looks a shadow of its former self, though the final stages suggested that the two teams were closer than the previous 70-80 minutes had made it appear. Prior to that, I had a look at their shot, statistical attack, and statistical possession numbers for the first 16 games of the season.
In their three stand-out categories highlighted above they had cooled slightly, fitting in 6го, 6го, and 7го against the 4 season sample. Coming high in both the attacking categories is not necessarily a red flag, particularly as they are now performing closer to what would appear to be normal (28.3% of their statistical attacks producing shots cooling off to 25.1%, against a general average of around 21%, as the most extreme example).
More worrying was the difference between the statistical-attack-to-shots conversion difference between their attack and defence. After 11 games the difference was nearly 12 percentage points between their conversion rate and their opponents’ whereas now it’s just over 7 – the highest in the 4 season sample is 7.95, so you can see why this seemed a red flag.
Unsurprisingly, given that their ‘underlying conversions’ are lower now, their shot numbers are too. They take fewer and concede more shots now (19.2 per game down to 16.4, 10.5 up to 11.3) and take and concede fewer shots on target (6.7 down to 5.6, 4.2 down to 3.9). This clearly means that their shot differences are also less impressive now, 3й in the league for raw shot differences per game and 4го for shots on target, behind City in both.
Shot numbers clearly aren’t the be-all and end-all of football, and most statistical predictions are now giving Arsenal the edge for the title. Their underlying conversion numbers are now no longer suspiciously good, suggesting they, and their shot numbers, may stay around their current levels. Could they win the league? Maybe. That, at least, I will concede. Will they? Who knows.
Присоединиться Брокер по ставкам на футбол в Истбридже сегодня!