“Unlike other sides at the top, Arsenal rarely rotate” and other fairy tales

 

 

• What would happen if Arteta were suddenly not available?
• The fight is on to try and stop Manchester City

By Tony Attwood

A little while ago a reader wrote in and said that Arteta should be sacked because he would not rotate his players.   We ran several articles to show that this notion of “not rotating” was not true.    Yet the story is back, as Football.London  has a piece saying “Unlike other sides at the top, Arsenal rarely rotate,”

The point here is not simply that outlets such as Football.London present information that is wrong, it is information that can be checked in a trice, and be seen to be wrong.  One only has to look at the data from Transfermarkt for example to see that Arsenal used 25 players in the Premier League last season.

Yes indeed some clubs used a lot more – Sheffield United used 35, Nottingham Forest 33 for example.   But they also had twice as many managers as Arsenal (two to one) and didn’t actually do very well.  

One of the reasons for clubs using a lot of players can be a regular change of management either before or during the season (Chesela used 32 players) or a lot of injuries (Newcastle also used 32).    But poor performance is generally the key (Burnely and Manchester United both used 31.  Although to be fair Aston Villa also used 31.  Which suggests that maybe there is not much link at all.

But let us consider these outrageous clubs that used 25 players last season.  There were three of them. One was Fulhham, and the other two were Manchester City and Arsenal, and it seems to me those last two had something in common although, obviously Football London would disagree.

Indeed FoLo have another bash at Arsenal with “Arsenal have damning £78m transfer reality as brand new FFP issue laid bare”.

That sounds like something frightening – and  we surely must read on, but if we do we can find the rather disappointing news that misses out the concept of “new” in the word “news” that,

“Arsenal could see a big chunk of their normal bench be sold this summer. With just 19 starts between them, it is looking like the end for Emile Smith Rowe, Reiss Nelson, Aaron Ramsdale, and Eddie Nketiah…”

Oh dear.

But let’s be fair, if we measure the use of players another way we can see Arsenal do use certain players a lot.   For a further examination of the players’ lists for the top three clubs shows that… Manchester C have five players who started thirty games or more and nine who started 20 games or fewer.

Liverpool on the other hand had three players who started 30 games or more.  Arsenal had eight players who started 30 games or more.

So Arsenal had more players who played 30 games or more.   But there is still one more question:  is that a good thing for a bad thing?  Since Arsenal ended up with 91 goals last season, more than any club other than Manchester City, does this mean something?   Or perhaps the fact that Arsenal conceded fewer goals than any other club means something…

The point is that choosing 30 games in the Premier League is an arbitrary number.  There is nothing to show that clubs who use fewer or more players do better.   If there were, we would have those statistics thrown at us regularly.

In fact, come to think of it, we did use to get figures concerning this thrown at us regularly.   And the Telegraph made a lot of it with “Revealed: The truth behind Arsenal’s terrible injury record – and how your club shapes up”.

That table from 2014 was used over and over again during the Wenger bashing era.   In the era of Klopp-worshipping a little less has been made of Liverpool’s figures.   Although it is possible to find the news that “Liverpool have lost 1,252 days from injured players during the campaign – considerably more than Arsenal (882) and Manchester City (654),”    That turned up in a report from Sky Sports.   Maybe now Klopp has gone, Klopp-bashing can be admitted as a journalistic endeavour.

There are two points here.   One is the data itself which involves finding a cut off point (more days lost through injury, number of players injured, number of players used, number of substitutes brought on, amount spent on players etc etc.)   Go through enough of these statistics and you will find one somewhere where Arsenal come out on top. Or bottom.

The second is that those statistics don’t actually mean anything, unless you can show that using more or fewer players, receiving more injuries etc actually has a direct link to performance on the pitch.   It’s called “cause and effect”.  Not something FoLo understands.

To give an example, I have spent hours trying to find a link between the number of fouls each team receives against them and, well, injuries, performance, goals against…. anything.  And I can’t.

The point is, just because numbers exist it doesn’t mean that they mean anything.   Having some players play more games might just mean that Arsenal have a much better injury treatment department.   Or it might mean that consistency works.

That’s the game the media plays.  Finding (or indeed quite often simply making up) numbers, and then claiming they mean something.  In short, when the media quote numbers, there may be no link either to reality, or to cause and effect and that is worth remembering.