A major match-preparation step is reached when referees are ready with knowledge. As a well-known figurehead, Collina, is reported to have stated - if you know before what can happen, you are ready. Here's a data visual from ~200k matches and >36K cards issued in a 10-year period - http://bit.ly/1UlWewL It is youth soccer, and there are distinct trends. For instance, there are potential two time-periods, 2/3rds into the first half and between 2/3rds and 3/4ths into the second when most misconduct gets penalised. And that referees are more strict when match time is multiple of 5 (strange, but fact). When scrolling over the years, these trends remain, so referees are ostensibly consistent. And while this is club and school data, it is very likely a similar trend in adult soccer. If sports analytics is useful for practical refereeing, then this could be a model that can be used to become acutely aware of critical time-periods in a match. Do others have similar observations from their personal logs or leagues they serve? --
The multiple of five is probably and artifact of referee estimating the time of infraction to the nearest 5 minute mark.
At this time of year with the heat in adult league, sometimes it's hard to get someone to take a throw in 10 minutes in to the second half. Over the weekend at both of my games around noon, I don't think there were any fouls in the second half.
Some states HS are 90 now. This will be the 3rd year in WI of running 90 minute matches with no OT until playoffs.
Thanks for compiling this data and presenting it to us. From the raw data given, the trend you see is quite strong. What one can infer from this is unfortunately very little. You imply that ref's might need to be more vigilant around certain times but the data reflect when misconduct is most likely to be sanctioned, but that may or may not be when most sanctionable events occur. So, this might be more useful for referees to be aware of our own patterns and whether those patterns fit the needs of a particular game. But I am tantalized by the possibilities of working your database. For example, you could see the effect of an early caution on the total number of cards on a game on the likelihood of a send off, on the odds of mass confrontation (which could be inferred by games with multiple cards recorded at the same time). Does a caution within the first 20 minutes lower the chances of subsequent misconduct? That would be an interesting answer. Another angle on that could be looking at games with five or more cautions and looking at the distribution of those. Conversely, in games with a send off, what was the number/time distribution of antecedent cautions? I certainly agree with mckersive that your data is in desperate need of some smoothing. Sure, minutes should be a continuous variable but looks like a lot of your refs measure time in increments of five minutes.
I don't find this surprising. Efforts at managing with cards (either a particular player or the game in genarl) fail, leading to a card -- that takes time. And (for better or worse) many referees are reluctant to give a card late as it is too late to change behavior.
While the data is super interesting I'm not real sure how referees might utilize it. Too many qualitative variables (i.e., level of play, game length, refereeing style, etc.) to measure it against. Sent from my iPhone using Tapatalk
This is historical information. When time after time, the 'trend' remains, all the qualitative variables average out. Check out the rest of the pages in the visual for details about age group, gender, year, day of week, etc. The drilling-down for more information is not provided in the dataViz. However, if I was a referee on a B14 match for Fall season, I would prepare for those critical time periods when I need to wrest control of the game back. Alternately, as chwmy hinted - - I would prepare for alternate course of action, as in. Hey, I know this group becomes ornery in second half, so let me lay down my business stall with prices clearly marked in the first half. Will my second half be smoother going? We all know referees came make a positive impact on the game flow, letting the game breathe, so let this be a template. Together we can change history, right? The referee coordinator can take the lead too. --
So, you're hypothesizing - based on the data you're presenting - what exactly? And just to clarify; the data you're sharing with us represents one - and only one - single referee, correct? Sent from my iPhone using Tapatalk
I'd wager the data is not all from 90 minute games, and the drop off after minute 70 is for the same reason as the drop after minute 80.
Okay, when I said the data was interesting I was just being polite... What I was really saying was the following: "The data is interesting but it doesn't mean jack crap. Period." Any conclusions drawn from the statistics being presented are merely serendipitous. For all we know the the referee(s) for which this data represents could be a complete moron or somebody who simply doesn't realize he or she consistently struggles at certain points in a match for any number of different reasons such as lack of fitness, skill, knowledge, etc. Sent from my iPhone using Tapatalk
Bootsy baby, it's 200,000 matches! 36,000 cards!! I think we can safely assume there is more than one referee!
I would suggest that the graphic tells me that games tend to not have a lot of misconduct in the first half and that the first 10 minutes after half tend to be about the same. But things start getting hot around the 55th minute and get steadily hotter for the next 25 minutes. I know this data has a lot of high school games that are only 80 minutes long and USSF games up through U-16 are also only 80 minutes long, but the drop off after 80 is still fairly significant. In my experience, there frequently is a realization by that point in the game that the score is the way it's going to be and getting more physical isn't going to help us change the outcome. I'd suggest that this sort of information and interpretation may be most helpful for experienced grade 8's and those seeking grade 7, as they have to transition their thinking and approach from "see a foul, whistle and point" to player management. I know that there is a tendency in, say, adult games, to feel like "oh, they're playing nice today. I don't have to be the hard guy today." and then all hell breaks loose after a while because you didn't whistle up some contact that irritated a player, so he goes out to get justice on his own. I think it also reinforces what most of us know, that tired players and losing players tend to try to do more than they are physically capable of doing, they get clumsy and they don't perform as well as they know they are capable of doing, so they get frustrated and/or mistime their tackles, etc. This could be used to start a discussion about game management in a classroom setting. A bunch of clips of MLS games don't, IMHO, lend themselves very well to talking about managing the game over 90 minutes.
I (in my old age) have no problem with giving a card late in the game or even after the final whistle has blown. You're right that at that point there is no effective management, but at that late point in a game I am no longer trying to manage a player. If I card one, I am telling him in no uncertain language that if you don't STOP whatever it is that you are doing , then you will be sent off from the field. Players know that we, as refs, (I was in that boat too), are frequently reluctant to discipline them towards the beginning or end of a match, and use it as an excuse to commit misconduct.