We have suggested for some time that the constant decision by PGMO (the organisation that employs and selects referees for Premier League matches) to employ a very small number of referees is one that should be changed, for the simple reason that if there is a rotten apple among the refs there is something seriously wrong with Premier League refereeing. The current system allows one referee to take control if a single team many teams across a season – which means just one “influenced” ref could have a profound influence. We think no ref should ever oversee one team more than twice a season.
Untold has also highlighted the perfidious Type III match fixing, in which the owner, or the minions of the owner of a certain club, could go to a ref and persuade him to have a negative influence on a rival, ensuring that they were more likely to get a draw than a win for example. This approach is hard to spot since it is spread over many games, and a ref so minded but faced with the marked team scoring a couple of early goals, will simply referee the match fairly, and thus deflect criticism.
Third, our analysis of games across the last four seasons including analyses by refs who support clubs other than Arsenal has shown a wide variation in the ability of Premier referees to spot fouls, and a very high level of errors. Which leads to…
Fourth, the PGMO constantly tell us that over 99% of referee decisions are accurate. Indeed in 2013 PGMO claimed this for the third year running, while our figures constantly showed far lower levels of accuracy. Yet PGMO utterly refuse to give any details of how they get their figures, and thus we are unable to compare their claims with our numbers that are much lower.
All of this suggests that something is very seriously wrong. But we have been missing one point: evidence of how easy it is for a referee who is so inclined to fix a match.
Now thanks to Louis Huguenin we have it. It represents, to my mind, the final nail in the coffin, for even if one argues that there is no match fixing, what we have is a system that is so easy to fix, that precautions should be taken at once to stop the possibility. The failure to take such steps is the greatest indictment of all. Tony Attwood
A new metric: Card-Minutes
by Louis Huguenin
I have been a frequent reader of the Untold Arsenal blog but only very rarely comment, since I only want to write something if it can add value. A few weeks ago an article about the number of fouls and cards was published. Naturally it plays a role in a game at what time a card is given (at min 5 it is more important than at min 90).
For this purpose I collected and analysed data from both Arsenal and Chelsea games to create a new metric called “Card-Minutes”.
I wrote down for which team and at what time a card was given. I then calculated how long in minutes this card would be active in a given game and influence the player in his playing style (e.g. not make any risky last ditch tackles anymore). By doing this for both teams in a game I created a new value called “Card Minutes” that basically weighs the cards (a red card is counted twice if due to a second yellow card) given by the time they are issued during a game.
A card given early in a game at minute 5 thus values much more than one given at minute 90. By doing the sum for each team this results then in a new metric called “Card-Minutes” for and against a team in a game.
I collected the data for all the games from last season for both Arsenal and Chelsea. As source I used the tables from Swiss television . To get the number of fouls I used the BBC sports website as source.
Even though sometimes the sample value is rather low, for example for draws and losses, there is still some value in the data. I also have to say that I didn’t double or triple-check the data, which would have been very time consuming. But even if there should be one or two mistakes in the data, the data itself and the difference between Chelsea and Arsenal over last season in card-minutes is enough to be quite noticeable and a variation in a few low-digits doesn’t change this fact.
|CM received by Arsenal||CM received by Opponents||Difference|
On average Arsenal concedes 54 card-minutes per game while the other team receives 62 card-minutes. This makes a difference of only around 8 minutes over the season, so maybe that is why people say “it evens out” over a season.
However it is much more interesting to look closer to so-called contested matches, e.g. losses and draws. Since in draws and maybe also losses the teams are more closely matched, small differences have a much bigger influence. There it is possible to see that the numbers are quite different. If Arsenal had lost a game (this happened 6 times last season) then on average the opposing team was given 19 fewer card-minutes. At a draw Arsenal was given 4 card-minutes fewer and at a winning game the number of card-minutes given to the opposing team is 18 minutes more than for Arsenal.
The reason for the difference of card-minutes is that if Arsenal is losing, then the players have to be more aggressive to catch up and thus are more likely receive a yellow card. The same reasoning goes for a draw, slightly less competitive and thus fewer yellow cards and card-minutes.
Therefore for a win it can be assumed and confirmed that the number of card-minutes Arsenal receive is lowest with only 45. What is interesting though is that amount of card-minutes the opponents receive doesn’t vary much (Arsenal loss: 59, draw: 62; win: 63). As some of fans already suspect though, for Arsenal this is not true at all. If Arsenal is losing, then they suffer on average 19 more card-minutes, for a draw this changes to four and for a loss to 18 for the opponents.
|CM received by Chelsea||CM received by opponents per game||Difference|
The picture is quite different for Chelsea. They had only three losses last season for which they received around 62 worth of card-minutes, which is 18 less than Arsenal have for losses.
However the opponents receive 34 more card-minutes, which is quite a pronounced difference. If Arsenal were losing, the team has a handicap of 19 card-minutes. If Chelsea were losing they had an advantage of 34 card-minutes. One could now ask a rhetorical question: under which match conditions, from purely a data point of view, is it easier to regain control of the game?
If Chelsea had won the game, the opposing team received 28 more card-minutes compared to Arsenal’s 18. If Chelsea drew, the card-minutes was a staggering 93 (for the opponent 102). In comparison at a draw Arsenal received only 58 card-minutes. This number is likely due the different styles of Arsenal and Chelsea.
It becomes even more interesting if one compares the numbers from Arsenal and Chelsea directly. If Chelsea lost rather than won they only received 8 more card-minutes. If Arsenal lost rather than won, they received 33 more card-minutes.
How can this difference be explained? Can this be attributed to a potentially a more aggressive style of Arsenal in comparison? Not really. Chelsea had 74 yellow and 4 red cards last season while Arsenal had 67 and 2 in comparison.
On the contrary if we look back at the first game from last season between the two teams it was clear that Chelsea had applied a more robust style rather than Arsenal. If Arsenal also had a similar positive swing like Chelsea in their losses, they would have had an advantage of (34-(-19)=+53 card-minutes compared to what really happened.
Imagine hypothetically if in addition to all the other cards handed out in Arsenal’s losses from last season that in each game an opposing player had at minute 90-53=37 received a yellow card. Would that have made that player more careful in his tackles and fouls? Would this have made an Arsenal comeback more likely?
Can the difference of card-minutes received be due to only the difference in style between the two teams and the context of the game? Or is it more related to who issues the cards and at what time in the game?
27 July 2006: The Highbury clock was erected onto the wall of the Emirates Stadium. The new clock inside the ground, based on the design of the original clock of Highbury was unveiled in 2010.
27 July 2012: Arsenal lost a pre-season game against Man City as some supporters became anxious about the club’s direction following the departure of Robin van Persie and Alex Song.
- 117 players tipped as coming to Arsenal before the window closes
- Why Arsenal recruited a new defence first, then a new attack
- After two games we can start to make one or two judgements
- Winning the opening two league games is not that common for Arsenal
- Arsenal? Nothing to see here. It’s only news when they’re losing.