The Refs Are Biased (Part 2)

To conclude my two-part series on referee bias, I have taken my inspiration from the great American philosopher and writer Mark Twain.

In my initial piece I was categorical. Based on penalties awarded:

“The unbiased data for the past 20 years is very clear; of all the top teams in the Premier League Arsenal is the least favoured team by Premier League officials.”

Yet there remains great skepticism and reluctance to accept the stone, cold logic of the data. Apparently there is an enduring belief that the referees are acting without bias, that they are honest arbiters doing a thankless job under difficult circumstances. Apparently we accept the gospel of the mainstream media (BBC) that the formation of the Professional Game Match Officials Board (PGMOB) in 2001 meant:

“….they will also be more accountable regarding their performances in matches.”

And lead to:

“….improving the overall standard of refereeing.”

In today’s final installment, I will demonstrate based on data of Penalties-Against (PA) that Arsenal is the most penalized among the top clubs in the Premier League and by a very wide margin. In making these decisions the PGMOB has been far from accountable and the promised improvements in the standard of refereeing has been a singular failure.

The data for the past 20 years confirm that the traditional top-four clubs in the premier league (AFC, CFC, LFC and MUFC) have the lowest PA compared to the rest. This is exemplified by the first 10 years of the Wenger era. It is logical that during this era when Arsenal and United dominated the League they would have the lowest PA, with MUFC marginally ahead within a relatively narrow range of 0.2 goals. LFC had the highest PA which makes sense since they have been less successful in winning titles although averaging 3^rd in league position. Also note that during those first 10 years the top-four clubs all had less than the league average PA of 3.4. Below I will  demonstrate the significance of the latter information.

In the second 10-years was a 38% increase in penalties awarded by the referees compared to the first 10. Amazingly, not only did Arsenal experience 206% increase in penalties awarded against the club but the average number of penalties was higher than the average for all clubs in the premier league, 4.9 vs 4.7.

Meanwhile the percentage increase in PAs for all other clubs was substantially less.

The second greatest increase was for United at 143%. But at 3-against they are well within the league average. In the context of refereeing, it is apparent in the post-Ferguson era the officials now have a greater willingness to make penalty calls against them, when before they seem to be in abject fear of the United boss. For those who think this is merely an opinion, note in Ferguson’s last year (12/13) on his way to the title, the refs awarded zero penalties against United. In a team with two notoriously physical central defenders, Vidic and Ferdinand, who were arguably past their best, this was an amazing statistical achievement.

Back to the main point. According to the unbiased data, PL referees over the past 10 years found AFC to be not only the worst of all the traditional top-four clubs in penalty-area defending but that it had been surpassed by LFC, a club whose average league position declined from 3rd in the first 10 years to 5th in the second 10. In total, Liverpool averaged only 1 more PA vs 2.3 more for the Gunners, a 130% difference.  Hello!

Yet over the past 10 seasons, based on mean averages, in every statistical category, except for Draws (D) and Goals Against (GA), Liverpool was inferior to Arsenal.

Given AFC’s general league superiority over LFC, it defies any statistical justification for the referees to award 130% more PAs against the Gunners compared to the Merseysiders.

Furthermore, if the referees are calling 38% more penalties then over the long term it should be distributed according to the league ranking of the club, as it was in the first 10 years. That is an iron-law of statistics. If not there is evidence of clear bias.

As someone with a little knowledge of how the federal government measures fair lending violations against protected classes in the USA, a prima facie case could be made of “disparate treatment” of Arsenal by the PGMOB. Many banks and lending institutions have been heavily fined and sanctioned for such apparent bias. Who holds the PGMOB accountable?

Let me take this analysis a little further. If we expand the group of top-clubs to five, to accommodate the rise of Manchester City who were basically a yo-yo club until the advent of external money in 2006 allowing them make massive spending on new players. Virtually overnight they became a top club in the premier league and the 3^rd best in penalty-area defending, overhauling Arsenal. Is it justifiable?

Skeptics and critics of my findings in Part I, that they were getting more favorable treatment than AFC in penalties-for, argued that City simply had better attackers. Yet for much of those years a number of their offensive-players were ex-Arsenal such as Adebayor and Nasri and they had managers like Hughes and Mancini who were arguably more conservative than Arsene Wenger. Yet they averaged 1 penalty more, 6.4 vs 5.4.

In the PA department the difference was even greater. Within 10 years they averaged 3.3 vs Arsenal’s 4.9, a 1.6 goal difference. Were their defenders vastly superior to Arsenal’s, sufficient to gain a significant edge in penalty area defending?

Au contraire. Many of those years among their defenders were ex-Arsenal such as Kolo Toure, Gael Clichy and Bacary Sagna. So in retrospect the personnel was relatively similar. Furthermore, in the last three years under review they were managed by Pellegrini who was arguably as attacking in team setup as Arsene Wenger. Yet since 2006, there is only one season when they had more PAs than Arsenal.

In comparing the PAs conceded by both clubs in the past 10 years, I was struck by the data for 15/16 season.  The referees only awarded one penalty against them despite their having disasters-in-waiting such as Mangala, Otamendi and Demichelis as central defenders. Collectively they were poor, conceding 41 goals compared to Arsenal’s 36. My recollection was of them mis-tackling and clumping opponents in their own box and the PGMOB letting them get away with it. Yet both teams ended up with equal number of penalties against. Can’t stop Shaking My Head (SMH).

In analyzing the data it is significant to note for most of the last 10 years, premier league officials have been under the leadership of Mike Riley (the main actor in ensuring Arsenal’s loss in Game 50), in his capacity as manager of the PGMOB. While Riley’s bias in Game 50 was manifest, he is no longer a referee and we have no available evidence to prove how his past bias currently affects his subordinates. However what is evident from the data is institutionally they are increasingly biased towards the big-spending clubs. As I demonstrated in Part I the big money clubs, except Arsenal, are increasingly and constantly getting the Penalties-For at the expense of the smaller clubs.

“Gone are the days when a mid-table team, between 1997 and 2006, would be ranked first in penalties awarded; such as Southampton (97/98), Middlesbrough (99/00), Newcastle (00/01, 01/02 and 03/04) and Crystal Palace (04/05).”

Seemingly the refs have consciously or sub-consciously internalized the message that, to succeed in the PGMOB, one must favor the big clubs which, except for Leicester in 15/16, now permanently monopolize the top-positions in the league. What is rarely discussed in the mainstream media  is the prime reason for the PGMOB’s establishment was to make refereeing a lucrative full-time occupation. According to the BBC report in 2001, previously referenced, former FIFA referee Roger Milford who advocated the PGMOB stated:

“There are footballers out there earning £20,000 to £30,000 a week – and that is at the bottom end of the scale,”

“So referees deserve to earn the sums being talked about.”

If a lucrative career is the prime driver of a full-time referee then it is consistent with human nature to have little desire to make seemingly controversial decisions against a big-moneyed club which could potentially be a career ender.

That would explain why, in last week’s game between Chelsea and Hull,  when the underdogs were already one goal up, there was arguably a clear case for a penalty against the Blues which would have given them a crucial two-goal advantage. This appeal was vigorously waived off by the referee. It may also explain why the referee studiously ignored the foul by Chelsea in the build-up to their first goal.

Is it any wonder we rarely see any big upsets of the big boys by the lower level teams especially in the second-half of the seasons? Does this explain why there are reports of a significant fall off in attendance and viewership of premier league matches especially those involving mid and lower-level teams? Why attend or give these matches any attention when the results are already pre-ordained?

On the basis of the evident disparate treatment of Arsenal by the refs when it comes to penalties and moreso an increasingly systemic bias against the smaller clubs, I am forced to revise my assessment of the chances of Arsenal overhauling Chelsea in the run-in for the title. When I did my first An Open Letter to the Arsenal First Team blog, I estimated a 50:50 chance which meant it was uncertain and unpredictable statistically. Now I am convinced it ranges between 5%-10% which, should it happen, would make it miracle. My eyes have now turn to the cups as our most realistic chance of any silverware.

PS: Thanks to usamazaka at Positively Arsenal who provided the source information on the founding of the PGMOB.

Postscipt

A legitimate question was asked in the comments section of this blog why I excluded Spurs from the analysis. Because they were never a consistently top-four team I eliminated them so as to compare like-with-like between the 1^st and 2^nd-10 years. To eliminate any suspicion that the data is being fudged to prove a point, below is the PA numbers for THFC:

The striking thing is, being a markedly inferior team compared to Arsenal, never ever exceeding AFC in league position for the past 20 years, we have a remarkable statistic that for the past 10 years they had less PA vs Arsenal; 44 compared to 49.

I rest my case your honor.