From Patterns to Parity

The (selective) post derby meltdown reveals much about various insecurities, not least the changing Scottish refereeing landscape.

The fallout from the Glasgow derby League Cup Semi-Final continues unabated into a fifth day, and the object of most of the angst is refereeing.

As you know, I have applied an analytical framework to this topic to uncover any underlying patterns behind refereeing performance in the SPFL. Especially as it impacts what has traditionally been “the big two”. In what the media likes to portray as a “two-horse race”, where giants of the game collide in a league of otherwise much smaller clubs, one would expect a parity of outcomes in refereeing decisions.

Given the predictable and, frankly, tiresome debating of refereeing decisions post-derby matches, I decided in 2021 to “eat my own dog food”. That is, to apply some data-driven rigour to the question rather than bitch and moan from the sidelines.

My starting hypothesis was that there would be no significant differences in how refereeing performances impact Celtic and The Rangers.

How wrong I was.

Patterns of Assistance

Using public data, with the added insight of a Sheffield Wednesday supporting professional referee from Yorkshire, I pulled together a framework for analysing the big calls in games involving Scotland’s perennial top two. Penalties, red cards and their impacts were analysed.

I will link all the detailed articles at the end, but the TL;DR is this:

• Opponents of The Rangers needed to enter their penalty box 372 times before a penalty was awarded, whilst the league average was 147 (142 times per penalty against Celtic);

• Whilst The Rangers gained a penalty every 147 forays into the opposition box, Celtic needed 203;

• The expected points gained by The Rangers’ penalties were 2.95 times more beneficial than those awarded to Celtic based on game state (time of award and score at that time);

• The Rangers’ average time of penalty award was the 48^th minute, but 59^th for Celtic;

• The Rangers’ average time of penalty concession was the 67^th minute, but the 45^th for Celtic;

• In the period under analysis, The Rangers’ opponents were given 12 more red cards than the Ibrox club. Celtic’s opponents received four more;

• When combining the expected points impact of both penalties and red card based on game state, awards favouring The Rangers were over three times more beneficial than those favouring Celtic; and

• Based on the independent Yorkshire Whistler incident review, The Rangers benefited from an estimated 13 points over three seasons due to incorrect calls. In two of those seasons, the league-winning points gap was four and seven points, respectively.

All the above were deemed statistically significant using Z-score analysis. A statistically significant Z-score (greater than +2.0 or -2.0 from the mean) indicates that a data point is far enough away from the mean to be unlikely to be caused by random chance. In English: there was something fishy afoot.

The collective results strongly indicated patterns of assistance for one club, The Rangers, in refereeing calls in the SPFL over multiple seasons.

John Beaton, Brendan Rodgers, “Incompetence” and Crawford Allan

The patterns of assistance became statistically significant mainly during the oversight of Crawford Allan as Head of Refereeing from 2020 to 2024.

There had been several blunders from SFA referees, even allowing for the introduction of the Video Assistant Referee (VAR) framework partway through a season. The SFA had been spectacularly caught out misleading Celtic in particular, but Scottish football in general, in November 2022, as detailed here.

In the 2023-24 season, senses were heightened by the fact Scotland’s champions would get automatic Champions League entry, probably for the last time for many years, as Scotland’s Coefficient reminds us. The stakes were high.

In March of 2024, John Beaton, the VAR, effectively re-refereed a Celtic tie at Heart of Midlothian to the detriment of the Hoops. Post-match, manager Brendan Rodgers called the display of the officials “incompetent”.

An SFA disrepute charge followed.

In the hearing, Celtic arrived what may be described as “lawyered up”. Additionally, they were very well prepared on many matters pertaining to refereeing procedures and performances.

No minutes are available, but by July, Crawford Allan had resigned.

Despite high-profile non-Scottish candidates, Willie Collum was appointed as Head of Refereeing for the 2024-25 season onward.

Willie Collum, YouTube, the KMI

Whilst most clubs would agree Willie Collum was consistently not great as a referee, he has been refreshingly innovative as the head of service.

Two key initiatives have increased transparency over refereeing in Scotland.

Firstly, Collum holds a monthly VAR review show on YouTube and Sky TV with Gordon Duncan. This examines contentious calls from the last month in the league, referencing the audio and video files used by the officials in the matches.

Whether you like the outcomes or not, this is gold dust insight into how the referees arrive at decisions. Collum adds his commentary on what should have happened.

Additionally, the independent review panel, the Key Match Incident body, publishes a weekly report detailing its views on the key calls, the outcomes, and whether it believes they were correct or not.

Therefore, not only are we, the fans, better informed, but the referees are on notice that errors will be subject to public scrutiny.

It is likely that referees DO NOT want to be YouTube stars on the Willie and Gordon show.

The Impact

We are a quarter through Collum’s second year of operation. Has the introduction of light into refereeing performances disinfected some of the patterns of assistance stench?

Last season, the results from the Yorkshire Whistler analysis concluded that there was a very small 1.29 expected points estimated impact because of incorrect calls. That seems to fall within the parameters of natural variance.

Here is the profile of decisions reviewed and incorrect calls identified over the period.

01

Not only did far fewer incidents need to be referred, but the error rate amongst those decisions fell dramatically to 13% compared to an astonishing 32% in 2021-22.

Does this mean refereeing standards are better?

Probably not. Scottish referees are objectively poor. Look at the paucity of assignments for SFA officials to major UEFA European club matches and international ties. It will be no shock if, once again, no SFA official is at the FIFA World Cup next summer.

This is not a surprise. The referees are part-time in a full-time sport. The SFA continue to select their officials from an unnecessarily narrow pool, specifically the Lanarkshire FA. Seemingly, there is virtually no football in Inverness, Aberdeen, Dundee or the Borders. Or indeed, Edinburgh.

Additionally, there has not been any major reform of the SFA.

When asked, “why do patterns of assistance to The Rangers exist?”, I would have to defer to experts for the answer. I can provide the evidence. I suspect that terms such as unconscious bias, culture, internal controls and procedures, training, and reward structures would all feature. James and I always talk about attribution. How you attribute the patterns of assistance is one for the organisational consultants, sociologists and historians.

So, the culture of the SFA remains unchanged, as does the lack of a risk control framework over refereeing conflicts of interest, supporting preferences, and family ties.

Overall refereeing quality has not demonstrably improved.

Yet, behaviour has been changed. It is not obvious that the patterns of assistance persist.

ALL clubs now seem to be affected by poor refereeing performances, more or less equally, or at least within the boundaries of normal variance. This is all we Celtic fans ever asked for.

Back To Black

Which brings us back to Hampden and the black pall of bitterness and recrimination over Sunday’s semi-final.

The same sociologists and historians would no doubt have a field day analysing the Scottish press reaction to the match. The insistence on concentrating on decisions that may have negatively impacted The Rangers only, and the wheeling out of ex-referees, not to defend Nick Walsh and team, but to back up The Rangers’ position.

This all serves to avoid scrutiny of the footballing performance. Of a transfer window where the club declared a £20 million net spend, the most expensive transfer window in the history of Scottish football on a net spend basis. On the evidence of performances on the pitch, many of them look, well, not very good.

Of the role and intentions of the new ownership at The Rangers. Indisputably, these are not fans. Ownership details and structures remain hidden from public scrutiny. The terms of purchase regarding the degree of leverage and repayment arrangements, a mystery. Slum landlords in it for a buck any which way, or sports management professionals looking to build a world-class operation? Either way, the Ibrox clientele have lost control of their club, probably forever.

A club determined to modernise and improve their standing in football, so they apparently say. This sounds laudable, but their work is all ahead of them. The entire Ibrox section of Hampden, belting out the legally established as racist “Famine Song” (exhorting those of multi-generational Irish heritage to “go home”) and legally established as sectarian “Billy Boys” (venerating a British Fascist and member of the Klu Klux Klan) passes without comment in Scotland. In 2025.

Rangers (old and new) remain the only British club charged with racist and discriminatory chanting by UEFA and have been sanctioned multiple times.

So, there is much to deflect from, and some thorny topics that many will find uncomfortable to address.

Scottish referees are undoubtedly poor. Nick Walsh and team may well have made several big errors on Sunday (I await an expert and neutral opinion on this, as I am biased and not suitably trained).

However, the evidence suggests that the incompetence now infects all equally, or at least within statistically acceptable boundaries.

For those at Ibrox, however, parity is another country. They’d rather “go home” to the comforts of the past.

The Patterns of Assistance Files

The detailed analysis laying out the evidence for patterns of assistance aiding one club can be found in the following articles: