Football fans the world over complain about refereeing decisions and we all understand that a referee has the thankless task of using his judgement to interpret the rules. We know that referees are subject to review and their performances are scrutinised by an observer subjectively using their experience and knowledge but also, objectively, using statistics to analyse decisions. If the review is done on a match per match basis and the overall, cumulative, statistics are not taken into consideration then particular decisions, which may, on review, have been seen to be wrong, might be dismissed as judgement call errors on the day or excused away but statistical evidence over a longer period should not be discounted. Presumably, if a referee were to consistently be seen to make mistakes in his judgement then something would be done to correct this situation. If it is the case that, when the statistics are taken into account, a bias is highlighted, then at what point does this trigger some kind of investigation?
The following text and statistics includes comments taken from a number of Hibernian supporters on their supporters website and from other websites* providing the source data. I have taken the liberty of editing it and presenting it in such a way as to hopefully give it clarity.
After yet another contentious refereeing display in the recent Hibernian v Celtic match, one fan (1) presented some statistics on the club’s fans forum. This was followed by other fans presenting more statistical evidence. Please note the comments below, taken from the initial post, which I believe demonstrate that these statistics and the conclusions drawn from them are presented as objectively as possible...
“First of all, I do not want to appear as a conspiracy theorist, but I thought that I would have a look at every referee’s performance at our games since the start of season 2011-12. I broke it down to the numbers of games each were in charge, number of yellow and red cards given to our players and number of penalties awarded to us. I did start to look at the cards, pens etc given to opposition players in these games to get a contrast but I was only able to find that information going back to Jan’ 2013. In this period we have had 117 competitive games refereed by Scottish refs. We have had 19 different refs, four have only been in charge for one game each. I have also worked out the average of yellow cards per game for each ref. I will let you draw your own conclusions. I list them in alphabetical order so as not to sway your thinking.”
Referee / Games / Yellows awarded against Hibernian / Reds awarded against Hibernian / Penalties awarded for Hibernian / Average yellows per game
Crawford Allan,------- 6 games, 7 yellow, 1 red, 1 pen, average yellow cards per game 1.17
John Beaton,--------- 5 games, 12 yellow, 1 red, 1 pen, average 2.40
Iain Brines,----------- 7 games, 4 yellow, 0 red, 0 pen, average 0.57
Craig Charleston,------3 games, 6 yellow, 0 red, 0 pen, average 2.00
Kevin Clancy,---------4 games, 5 yellow, 0 red, 0 pen, average 1.50
Willie Collum,-------- 14 games, 21 yellow, 1 red, 2 pen, average 1.5
Brian Colvin,--------- 3 games, 3 yellow, 0 red, 0 pen, average 1.00
Mike Conroy,---------2 games, 2 yellow, 0 red, 1 pen, average 1.00
Stephen Finnie,------3 games, 6 yellow, 1 red, 0 pen, average 2.00
Bobby Madden,----- 13 games, 16 yellow, 0 red, 1 pen, average 1.23
John McKendrick,---- 1 game, 1 yellow, 0 red, 0 pen, average 1.00
Steve McLean,----- 13 games, 25 yellow, 1 red, 3 pens, average 1.92
Alan Muir,----------- 5 games, 12 yellow, 0 red, 0 pen, average 2.40
Calum Murray,------- 7 games, 10 yellows, 0 red, 0 pen, average 1.43
Euan Norris,--------- 5 games, 8 yellows, 0 red, 2 pens, average 1.60
Steve O’Reilly,------- 6 games, 16 yellow, 0 red, 2 pens, average 2.67
Charlie Richmond,---- 1 games, 1 yellow, 0 red, 1 pen, average 1.00
George Salmond,-----1 games, 1 yellow, 0 red, 0 pen, average 1.00
Craig Thomson,----- 17 games, 46 yellow, 2 red, 0 pen, average 2.71
B. Winter,----------- 1 games, 2 yellow, 0 red, 0 pen, average 2.00
Total 117 games, 181 yellow, 7 red, 14 pens for. Average yellow per game 1.54
Referee / Games / Yellows awarded against opposition / Reds awarded against opposition / Penalties awarded for opposition / Average yellows per game
Crawford Allan,------- 4 games, 6 yellow, 2 red, 1 pen, average yellow cards per game 1.50
John Beaton,--------- 3 games, 5 yellow,0 red, 1 pen, average 1.67
Iain Brines,----------- 3 games, 5 yellow, 0 red, 0 pen, average 1.67
Craig Charleston,----- 1 games, 0 yellow, 0 red, 0 pen, average 0.00
Kevin Clancy,-------- 3 games, 5 yellow, 0 red, 1 pen, average 1.67
Willie Collum,--------- 7 games, 11 yellow, 1 red, 0 pen, average 1.57
Brian Colvin,----------2 games, 4 yellow, 0 red, 0 pen, average 2.00
Bobby Madden,------- 6 games, 15 yellow, 0 red, 1 pen, average 2.50
John McKendrick,----- 1 game, 1 yellow, 0 red, 0 pen, average 1.00
Steve McLean,------- 6 games, 5 yellow, 1 red, 0 pens, average 0.83
Alan Muir,------------ 2 games, 3 yellow, 1 red, 1 pen, average 1.50
Calum Murray,-------- 1 games, 0 yellows, 0 red, 0 pen, average 0.00
Euan Norris,---------- 1 games, 3 yellows, 0 red, 0 pens, average 3.00
Craig Thomson,------- 6 games, 8 yellow, 0 red, 2 pen, average 1.33
Total 46 games, 71 yellow, 5 red, 7 pens for, Average yellow per game 1.54
(2) Put in a slightly different way, and including roughly equivalent statistics on Heart of Midlothian matches where Mr Thomson has refereed for comparison, here, below, are Mr Thomson’s games from start of 2011/2012 season to present. (The opposition statistics are in brackets).
Hibernian
Games = 17
Yellows = 46 (21) 68%/32%
Reds = 2 (2) 50%/50%
Penalty Awards = 0 (5) 0%/100%
Hearts
Games = 15
Yellows = 27 (36) 43%/57%
Reds = 0 (3) 0%/100%
Penalty Awards = 4 (1) 80%/20%
To further explore Mr Thomson’s statistics (3) an analysis of the figures for clubs (**) in which he has refereed over 25 games would suggest the team with the highest likelihood of getting a player booked is Hibernian (2.28 per game or 73 in 32 games) whilst the club involved in games where the opposition are least likely to get booked is also Hibernian (1.38pg). The side least likely to get awarded a penalty from Mr Thomson is Hibernian (0.06pg or 2 in 32) and most likely to get a penalty awarded against them is also Hibernian (0.28pg or 9 in 32).
(*2) Statistically, based on the data above; Mr Thomson is more than twice as likely to book a Hibernian player as an opposition player. He is 30% more likely to award the opposition a penalty during a match against Hibernian and 0% likely to award a penalty to Hibernian.
He is also more than 25% likely to award Hearts a penalty when officiating one of their games and 20% more likely to send an opposition player off. He is 0% likely to send a Hearts player off.
These statistics are based on the last 17 games that he has taken charge of involving Hibernian, which covers the current season and the 2 seasons prior to it, and the 15 games that he has been in charge involving Hearts during the same time period.
(*3) At the time of writing this document; this season Mr Thomson has refereed 14 SPFL games and handed out 48 yellow cards. He has awarded two penalties this season, one in the Hibernian v Celtic match against Hibernian and the other was awarded to Hearts against Dundee United when Hearts lost 4-1.
(*3) Mr Thomson’s overall statistics are that he has refereed 32 Hibernian matches, Hibernian have won 11, drawn 3 and lost 18. He has booked 73 Hibernian players and sent off 4 whilst booking 44 opponents and sending off 2. He has awarded Hibernian 2 penalties whilst giving the opposition 9.
Having booked Hibernian players 73 times and opposition players 44 times gives Mr Thomson a ratio of 73/44 or 1.66. Simply put, if a referee had booked the same number of players from each team then that ratio would be 1. One would anticipate that over a long career an impartial referee would look to have a ratio of 1 or as near as possible to it and would perhaps be concerned if there was evidence that he effectively booked on overage 1.66 players from one team as opposed to 1 from every other team they played.
(4) Here are Mr Thomson's full statistics on the website listed by another Hibernian poster (5), which takes all his games involving all the current Premiership /SPL into consideration.
On average, Mr Thomson awards 1.70 yellow cards per game (excludes Hibernian, includes double-yellows).
On average, Mr Thomson gives Hibernian 2.41 yellow cards per game.
He gives Hibernian 42% more yellow cards than his average.
On average, Mr Thomson awards 1.38 yellow cards against Hibernian opposition per game.
He gives Hibernian opposition 18% fewer yellow cards than his average.
In Hibernian games, he gives Hibernian 75% more yellow cards than the opposition.
After Hibernian, the next worst “sufferers” are Kilmarnock. In their games Mr Thomson gives Kilmarnock 27% more cards than the opposition.
In Hearts games, Mr Thomson gives the opposition 24% more yellow cards than Hearts.
On average, Mr Thomson awards 0.13 pens per game to all other teams, and 0.06 per game to Hibernian; in effect, he is half as likely to award Hibernian a penalty as any other team.
On average, Mr Thomson awards 0.10 penalties against other teams and 0.28 against Hibernian i.e. he is three times more likely to give a penalty against Hibernian than any other team.
In the 6 derby games that Mr Thomson has refereed he has given 20 yellow cards to Hibernian and 11 to Hearts. That is an average of 3.33 per game to Hibernian. He has given 1 double yellow to Hibernian and none to Hearts, which brings the average to 3.66 yellows to Hibernian. He has given 1 straight red to Hibernian and none to Hearts. He has given no penalties to Hibernian and 1 to Hearts. He has never given more cards to Hearts than Hibernian in any of these games he has refereed.
(3) Hibernian are only second most likely to have a player sent off (0.13 or 4 in 32) which is only slightly behind Rangers (0.15 or 5 in 33). We are third least likely to have an opposition player red carded (0.06 or 2 in 32) a short distance behind 2nd least likely Aberdeen (0.06 or 2 in 31) and surprisingly Celtic (0.04 or 2 in 51). There is a higher probability of Mr Thomson sending off an opposition player against Hearts than any other team (0.26 or 10 in 38) which is double that of the next best. For balance Hearts are second most likely (behind Hibernian) to have a player booked by Mr Thomson but second least likely to have a player sent off (behind Dundee United). For context it is also worth noting that Hearts have been fined for poor discipline on seven successive occasions, which does make one consider if that indicates that every other referee was getting it wrong in Hearts games during that period.
If we look at other referees who seem to get the highest profile encounters such as Mr Brines, Mr Collum and Mr Murray then we find that Mr Thomson’s booking rate for Hibernian players is still the highest overall at 2.28 per game. Mr Collum is most likely to book a Hearts player (2.27 pg).
Bookings for Hibernian players are Mr Thomson at 2.28 pg; Mr Murray 2.07pg; Mr Collum 1.67pg and Mr Brines 1.67pg. Mr Murray has only refereed 14 Hibernian games.
Bookings for Hearts players are Mr Collum 2.27pg; Mr Brines 2.16pg; Mr Thomson 1.87pg and Mr Murray 1.57pg.
If one were to plot many of the above statistics on a graph, there will be a fairly steady linear progression, with Hibernian as a complete outrider on every single one.
We believe that there is a clear pattern here, and if it was purely coincidence we'd maybe come up as the most "hard done by" in one, maybe two, of the categories listed above, but we can't see an argument to explain these statistics stronger than that which indicates that Mr Thomson is biased against us. That's before you put in the observational arguments, anecdotal evidence of major, game changing, decisions almost relentlessly going against us.
As far as I can determine, these statistics haven't been skewed in any way, however, it is important to note that, if you go further back than the two seasons prior, the figures appear to be somewhat more level. What that indicates is something that would require more statistical information and investigation as to why that might be but it doesn’t alter the conclusions that can be drawn from this sample of statistical data, which is certainly large enough to merit serious concern.
Whilst all of the above statistics provide interesting reading, we do appreciate that a fuller analysis of Craig Thomson's record with every Premier league team would be required. If it is then clear that he punishes Hibernian more than is statistically reasonable, regardless of opposition, then we would expect action to be taken.
It is also important to understand the context of the games in which decisions are given. A cursory scan suggests that Mr Thomson has refereed Hibernian games a disproportionate number of times against Celtic, Hearts and Aberdeen. This beggars the question; are these games where you'd expect more disciplinary activity?
One of the other thing's these statistics don't show is what time in a game bookings are given, some might suggest that booking a defensive player after 10 mins has a bigger sway on the outcome than, for instance, booking a midfielder on 80 minutes and, it goes without saying, that picking up a booking at any point in a game can obviously have an impact on up-coming games, particularly if the player also picks up a suspension. No Hibernian fan needs reminded of the consequences of the two fouls committed in the early stages of the Hibernian v Hearts final and the timing of the penalty decision and subsequent sending off in that same match. It could be put that, once the damage is done, the referee can be as fair as he likes. We may not dispute that Hibernian would have lost that final match; nor the recent game against Celtic but the eventual score-lines in both games may well reflect the referee more than the opposition.
It may be said that the figures do appear to suggest a bias against Hibernian and a bias in favour of Hearts. The prima facie case does seem difficult to explain in any other way. The red card and penalty data is perhaps too small a data set to draw any firm conclusions from and maybe there is a perfectly innocent reason that sways Mr Thomson to flash a card more readily at Hibernian players than our opponents. It would be interesting to hear it. It would also be interesting to see fouls awarded for/against if that information were available to see if there is any correlation with these statistics.
We understand that Mr Thomson has built up the reputation as "Scotland’s best referee". Our view would be to ask if we really expect this kind of statistical imbalance in terms of the decisions from our “best” referee. Hibernian are one of Scotland’s best teams but their games involving Mr Thomson would appear to indicate otherwise.
In summary, we would ask that these statistics be brought to the attention of the football and refereeing authorities in Scotland and, at the very least, investigated. I should point out that not all the posters on the thread were in agreement in their opinion of Mr Thomson. Many argued that he was merely incompetent. It was also noted that statistics can be used to prove anything. The challenge was then laid down to demonstrate, from the statistical evidence, how Mr Thomson was being fair to Hibernian. We imagine the authorities have all of the relevant statistics at their fingertips and we eagerly anticipate his vindication.
(1) Statistics posted by Big-mo
(2) Statistics posted by Hibercelona
(3) Statistics posted by Kaiser1962
(4) Statistics posted by Nailrod
(*5) Statistics posted by Golden Fleece
* www.transfermarkt.co.uk
(**) Motherwell (26); Aberdeen (31); Hibernian (32); Dundee United (32); Rangers (33); Hearts (38) and Celtic (51)
This document has been created using contributions from Big-mo, Hibercelona, Kaiser1962, Nailrod, Twa Cairpets, matty_f, tomf , Sunny1875, Golden Fleece, Hexham Hibee
-
Hibernian v Craig Thomson... (by tomf)
- '); loadVbCss(yuipath + '/treeview/assets/skins/sam/treeview.css'); //-->
Hibercelona- 19-02-2014, 06:20 AMNailrod- 21-02-2014, 01:36 AMvBulletin Message