Proof that the media is inventing statistics to show there’s no ref bias.

By Tony Attwood

The Daily Mirror has decided to return to the old, old theme of dirty Arsenal – suggesting that somehow Arsenal are the nastiest club in the League when it comes to ill-discipline.

In doing this they are playing their old game of selecting the statistics that suit their view, while adopting the notion that whatever the referee says, is right.  However as we have shown before, there is no reason at all to make this assumption but every reason to suspect profound and continuing bias.

The Mirror’s statement opens with, “Dani Ceballos’ sending off against Villarreal on Thursday brought up an uncomfortable milestone for Arsenal boss Mikel Arteta.

“Since the Spaniard replaced Unai Emery in December 2019, 10 Gunners players have received their marching orders in the 80 games played in that time in all competitions.

“Shockingly that is twice as many as any other Premier League club in that period, suggesting Arsenal are suffering from a major discipline problem….”

Of course that may also suggest that the anti-Arsenal bias that we have found in refereeing statistics in the past has returned full time, but let’s follow the Mirror’s argument.

“… it is an issue that has been harming the Gunners ever since Arteta took charge and is part of the reason for their poor Premier League season.”

Now that is interesting because I have several times expressed concerns that Mr Arteta is going overboard on discipline, and handing out his own penalties for behaviour he has not liked, which has led to the demise of Ozil and Guendouzi and punishments for Aubameyang, and we hear other players as well.

The Mirror’s point is that Arteta’s strong disciplinarian approach contrasts with the approach of Emery whose “attitude was much more relaxed, with the Spaniard preferring to issue warnings rather than hand out fines….” and that “he would let ill-discipline slide if it had a positive effect on his squad’s morale.

“That approach proved a popular one with his players, with only five players getting sent off in the 58 games played in his first season at the Emirates.”

Now that invokes quite an extensive set of cause and effect links, without any evidence at all.   So let’s have a look at where Arsenal are in the discipline table this season…  

Yellow card  2nd yellow Red          
1 38 0 0
2 37 0 1
3 44 0 0
4 Wolverhampton Wanderers 44 1 0
5 40 1 1
6 41 1 1
7 44 2 0
8 51 0 0
9 Brighton and Hove Albion 36 2 2
10 39 1 2
11 48 3 0
12 53 0 1
13 54 2 0
14 46 0 3
15 57 0 1
16 West Bromwich A 44 0 4
17 41 1 4
18 54 0 2
19 59 0 3
20 63 0 3
Arsenal estimated 2020/21 47 1 5
20 Arsenal 2019/20 85 1 4
16 Arsenal 2018/19 71 1 1
12 Arsenal 2017/18 0 2
4 Arsenal 2009/10 0 0

The final four rows show Arsenal’s figures in previous years compared with now, which does show a decline in disciplinary issues for Arsenal, not an ever escalating increase!   And you’ll see from the “Arsenal estimated” row that assuming that in the remaining few games Arsenal continue to pick up cards as they have before, their number of cards will be way down on recent seasons.  Back in fact to the 2009/10 level.

So this story is nonsense. Even if it is true that the analysis of the “last 80 games” shows figures the paper suggests, that is irrelevant compared to the normal model of looking season by season. 

Just like the story about a lack of crowds having no effect on home/away performance, which New Scientist ran recently.  That also is statistical nonsense because the current season’s figures are utterly different.

But why are they doing this?

If you are a regular reader here you will know that from the start of this season we have been warning Arsenal that referees will punish them more severely than other clubs – backing this up with our regular analysis of tackles, fouls and yellow cards.   We have repeatedly showed that referees were punishing Arsenal unduly harshly (as indeed is reflected in the Yellow Card column above).  Thus we argued that the only way for Arsenal to survive was to cut down on the tackles – which then gives referees fewer chances to give out yellow cards.

(We also pointed out one or two other tricks going on – most notably Leicester’s tackling and their penalty haul.  I’ll return to that in another piece).

The point we made was that if Arsenal cut its tackling back dramatically, referees would not be able to harm the club so regularly by dishing out yellow cards for the sort of actions other teams dish out with impunity.  

I don’t know if Arsenal read our commentary, but they have certainly adopted the tactic.  The number of tackles has dropped so that Arsenal now tackles far less than other team.  And the tactic worked as our yellow card haul has dropped.

Yet here we have the Mirror and the Athletic seemingly trying to say that Arsenal get more yellows than anyone else.  If the last 80 game data is true (and they produce no evidence) it is irrelevant because these things are measured season by season.  It is wholly artificial.

Most worrying of all is the way that this story has followed on from the denial of the fact that home/away statistics have changed this season.  We seem to be entering a new era in which the media simply creates statistics to suggest that all is ok with refereeing.