Commentary

Food for thought: refereeing and data

As an engineer and a computer scientist, I am comfortable with numbers. I believe that the numbers do not lie, but they do not necessarily tell all the truth.

Data, statistics and data analysis had been used extensively in sports, especially in the non-soccer U.S. sports. Some teams like the Oakland A’s of the MLB used analytical data to select players which led them to a success disproportionate to their financial status. Lots of teams tried to mimic the success story of A’s even in soccer. The use of data in soccer has limited usage; it depends on how you interpret the data. For example, having a high possession rate in games does not always lead to success. At the end of the day, data does not score goals, players do.

Data is not used extensively in assessing soccer refereeing with the exception of using -- maybe -- distance covered by a referee in a professional game. Different countries have different styles of refereeing. We have the “letter of the LOTG” referees and the more common sense -- man manager -- referees. Even at the highest level, you see specimens of the whole spectrum. Some referees are known to draw the cards easily and some with a second thought. In all levels of soccer officiating, “game control” is the No. 1 priority. The way you achieve it as long as you stay in the basic LOTG is a matter of style and you as the referee bear the consequences of your style.

CIES very recently has released a report on the top 31 leagues and their referees. Well, to be exact, the report has the average number of cards card per game in those leagues. As far as I remember, this is a first time for CIES to do an analysis on refereeing. The report is for the current season. Nevertheless, the data in the report is interesting to analyze.

The highest average of cautions per game is in the Greek Super league with 5.23. The list is headed by three Eastern European leagues (Romanian, Turkish and Ukrainian) and Spain's La Liga. For red cards, the Romanian league leads with 0.3 red cards per game. It's followed by Ligue 1, Greek League, Serie A and the Turkish league.  Clearly, these results reflect the nature of these leagues and the player/referee culture of these geographies. Not surprisingly, the lowest average number of cautions is in the Northern leagues: Norwegian (2.83), Swedish, Finnish, Danish and Dutch leagues. Nearly on the average there are twice as many cards shown in the Greek league versus the Norwegian League. In red cards, the Danish league (0.08) has the lowest average followed by the Norwegian league, EPL, Swedish and Belorussian leagues.  The Greek referees on the average show 4.5 times more red cards than their Norwegian colleagues.

Yes, the EPL has a very low average red card per game. Is this natural? Do the players coming from all around the globe when they play for the EPL commit very rarely red card offenses? Or are the EPL referees more lenient than other leagues?

We have to compare the officiating in EPL not just with cards but with other metrics to other prominent leagues and MLS. I developed a table using 2015-2016 season results from www.whoscored.com data. The table includes the group stage of the Champions League and the Europa League:

League
Fouls per game
YCs per game
RCs per game
Foul/YC
ratio
Foul/RC
ratio
YC/RC
ratio
EPL 21.57 3.12 0.16 6.91 134.81 19.50
La Liga 27.28 5.29 0.29 5.16 94.07 18.24
Serie A 30.41 4.86 0.34 6.26 89.44 14.29
Bundesliga 29.16 3.81 0.13 7.65 224.31 29.31
Ligue 1 26.76 3.79 0.31 7.06 86.32 12.23
*Champions League 24.75 3.96 0.17 6.25 145.59 23.29
*Europa League 26.88 3.93 0.23 6.84 116.87 17.09
MLS 24.75 3.29 0.22 7.52 112.50 14.95
AVERAGE
26.445
4.01
0.23
6.71
125.49
18.61
* Group stage.


The data includes average number of fouls per game for that league. Although not all cards relate to fouls, still there is a correlation between the fouls and the cards.

The most number of fouls are called in Serie A and the least in the EPL. Naturally, it is the number of fouls called and not the number of fouls committed, since a foul is only a foul when the referee calls it. So the EPL referees call less fouls than their European counterparts.  Few fouls also mean less stoppage and hence more playing time for everyone. The EPL has also the lowest average of yellow cards in a game which is not surprising. The trophy for the highest average number of cards in a game goes to Liga officials. Serie A also leads the average number of red cards per game. Surprisingly, the German league showed the least number of red cards in the Bundesliga last year. The referee committee of DFB must have given some firmer instructions to the referees since this year the average rose from 0.13 to 0.20.
The average number of fouls per yellow card is an interesting metric; the higher values show that referees in that league call a tight game that is not complemented with cautions. The high value can be interpreted such that the referees are calling trifling offenses with the intent of controlling the game, but they wait too long to issue cards.

The average number of fouls per yellow card ranges from 5.16 (Liga) to 7.65 (Bundesliga). It is apparent that the German referees like to call a tight game, but they are a bit stingy when it comes to cards. If you take the average of 6.71 fouls per yellow card, La Liga and MLS referees are off the norm a little bit, too.

When the same sort of ratio is applied to the red cards and fouls, we see that the German referees on the average have to call 224.31 fouls before they issue a red card. On the other hand, the French referees are fast with their red cards. It takes on the average 86.62 fouls for some player to see a red card.

The same is true for the YC/RC ratio. For each 12.23 yellow cards, French referees show a red card, whereas the German referees wait 29.31 yellow cards.

What is interesting is that this year’s average of yellow cards by EPL referees is higher than last year’s (3.12 vs. 3.88), but the average red cards have dropped down from 0.16 to 0.11. Either PGMOL is happy with the use of red cards -- hence game control -- or the players are better behaved players this year. It is clear that the EPL has a different approach to officiating, which this table indicates, but then some of the best referees of the world are from the EPL.

Refereeing in the UEFA competitions seem to reflect the European league norms. The refereeing in MLS seems to be close to the European norms in terms of fouls called, yellow cards and red cards. It is a bit closer to EPL norm, which is not surprising. Unlike some commentators who want to think that in MLS too few cards are shown, the averages show that PRO officials are doing a job which is close to global norms.

You can use this data in any way you want to reach conclusions and interpretations, but these or similar data is all you will find to analyze referee performances across leagues and years. Unfortunately, there are no uncontested statistics about correct and incorrect calls. Most of the foul calls in a game fall into the category of gray decisions.

Soccer refereeing is a very difficult task; one should realize that there is no such thing as perfection in refereeing, although excellence can be sought.

Ahmet Guvener (ahmet@ahmetguvener.com) is the former Secretary General and the Technical Director of the Turkish FA. He was also the Head of Refereeing for the Turkish FA. He served as Panel member for the FIFA Panel of Referee Instructors and UEFA Referee Convention. He now lives and works as a soccer consultant in Austin, Texas.

advertisement

advertisement

>
8 comments about "Food for thought: refereeing and data".
  1. R2 Dad, April 22, 2017 at 1:37 a.m.

    Good stuff, good metrics, Ahmet. It doesn't appear to be available, but would be very curious when those fouls/cards are issued (first half vs 2nd) since I think there is a correlation between the number and timing of cards e.g. referees call fouls in the first half but no cards, then need to issue a flurry in the 2nd half because they thought they had a handle on it but in reality the game was spinning up and then got away from the center. Another metric that might be of use are the number of post-match punishments issued by the league that the officials missed during the match.

  2. Kent James, April 22, 2017 at 10:43 a.m.

    So the evidence is in; in England, soccer is still clearly a gentleman's game, since the EPL had the fewest fouls! Something about "lies, damn lies, and statistics"... :-)

    Interesting data...

  3. Mark Landefeld, April 22, 2017 at 2:08 p.m.

    The problem I have with the ratio analysis is that it doesn't seem to account for the application of advantage.

    That could skew the numbers a bit.

  4. Allan Lindh, April 22, 2017 at 2:25 p.m.

    What this leaves out is the very cynical strategy of some teams to intimidate or injure the essential skill players of their opponents, with vicious tackles, usually early in the match when cards are less likely. Until the Masters of the Universe have the guts to come down hard on this destructive practice, the beautiful game will continue with its ugly patina. Statistics could be compiled on this, and maybe they would be enough to convince the owners to protect their most valuable assets. Someone should tell them that if we want blood and guts, we'll watch hockey or cage fighting.

  5. R2 Dad replied, April 23, 2017 at 2:34 p.m.

    i.e. how often are teams carded for Persistent Infringement?

  6. Bob Ashpole, April 23, 2017 at 6:17 a.m.

    What is the value of this information other than to predict how many cards and fouls will occur? What would interest me is looking at a correlation between fouls and misconduct and injuries, but that would require additional information and some care in defining "injury" in an objective manner.

  7. beautiful game, April 23, 2017 at 2:19 p.m.

    Good piece of analysis. Add to this metric how many fouls are committed and not whistled in order to gain a much clearer picture. And that means rewind the taped games and with neutral review of each and every tackle to be fair minded.

  8. Fire Paul Gardner Now, April 24, 2017 at 11:23 a.m.

    Very interesting article and stats. Thanks!

Next story loading loading..

Discover Our Publications