A couple people asked in other threads if I had looked at referee bias statistics. I took a quick pass at this at the end of last season but never posted the results -- so here is that post. The season referred to is the 2017-2018 season.
Team Bias
Most of us believe it exists for a few particular teams, but referees don't work the same team often enough to produce meaningful data. How a team plays offense or defense, its personnel, and the particular opponent cause teams to foul more or less, and there's too little data to ferret out whether an individual referee subconsciously favors a particular team.
For example, last season 48 different referees worked NC State games. Our 18 ACC games used 33 different referees. In our conference games, four refs worked three games each (Dorsey, Henderson, Simmons, Valentine), 13 worked two games, and 17 worked just one of our games. We never saw the same crew twice all season and only once did we have the same pair of refs (Dorsey, Gaffney) a second time. More on that below.
Home Team Bias
Oh yes, it exists. An extensive study several years ago concluded that there is a home team bias and there's a separate metrics argument that this bias actually accounts for the majority of home court advantage. In our games that was true; we averaged fewer fouls than the opponent at home and more fouls on the road. We were called 22% more often on the road and our 17 ACC opponents collectively were called 19% more often at PNC than at home.
Score Bias
The same study found that teams are more likely to get called for a foul when ahead in the score, and less likely if behind. That works to keep games closer and seems to be an entirely subconscious bias.
Another aspect of score bias is that fouls tend to trend towards evening out over the course of a game. That is, when refs have called several consecutive fouls on one team there is a tendency to call fewer fouls on that team and more on the other team. Specifically, the bigger the difference in fouls between the two teams, the more likely it was that the next call would come against the team with fewer fouls. For instance, when the home team had five or more fouls than the visiting team, there was a 69 percent chance the visiting team would be whistled for the next foul.
Whistle Bias
Speaking of whistles, some are quicker than others. I hope coaches pay attention to this one. Ken Pomeroy did a big study last year that found Ted Valentine and Roger Ayers called significantly fewer fouls than any other NCAA referee -- about 3.5 fewer per game. (My wife says Ted spends too much time staring at his reflection on the floor.) John Hampton is another slow whistle guy -- he calls 2.7 fewer fouls. So if you get two of those guys in the same crew you can presumably play ultra-aggressive defense. Of the guys we saw more than once Dorsey, Desau, L. Jones, Gaffney and Sirmons all call noticeably more fouls than average -- all in the +1 to +1.5 range.
Incidentally, we saw Dorsey and Gaffney together twice (the only pair who repeated all season) and that pair calls 3.01 more fouls than average. Those two games produced 42 and 40 fouls, well above the average of 34 fouls in our other games.
Is there a fatigue effect?
Referees often work as many as 90 games. Doug Sirmons, who did 4 of our games, worked 91 games in 21 states for 20 different conferences. Roger Ayers, who did 5 of our games, did 88 games in 24 states for 21 conferences. That's like 5-6 games a week, with significant travel time between games. The travel time alone has got to be fatiguing. (The multi-conference officiating alliance formed a few years ago was designed to alleviate that with better scheduling, but I haven't seen whether it really made a difference.) There is no longer a website that tracks referee workloads, so the effect of travel fatigue can't be quantified. But I have a hard time believing that it doesn't affect foul calling.
That's it. Got to prep for today's game....
Team Bias
Most of us believe it exists for a few particular teams, but referees don't work the same team often enough to produce meaningful data. How a team plays offense or defense, its personnel, and the particular opponent cause teams to foul more or less, and there's too little data to ferret out whether an individual referee subconsciously favors a particular team.
For example, last season 48 different referees worked NC State games. Our 18 ACC games used 33 different referees. In our conference games, four refs worked three games each (Dorsey, Henderson, Simmons, Valentine), 13 worked two games, and 17 worked just one of our games. We never saw the same crew twice all season and only once did we have the same pair of refs (Dorsey, Gaffney) a second time. More on that below.
Home Team Bias
Oh yes, it exists. An extensive study several years ago concluded that there is a home team bias and there's a separate metrics argument that this bias actually accounts for the majority of home court advantage. In our games that was true; we averaged fewer fouls than the opponent at home and more fouls on the road. We were called 22% more often on the road and our 17 ACC opponents collectively were called 19% more often at PNC than at home.
Score Bias
The same study found that teams are more likely to get called for a foul when ahead in the score, and less likely if behind. That works to keep games closer and seems to be an entirely subconscious bias.
Another aspect of score bias is that fouls tend to trend towards evening out over the course of a game. That is, when refs have called several consecutive fouls on one team there is a tendency to call fewer fouls on that team and more on the other team. Specifically, the bigger the difference in fouls between the two teams, the more likely it was that the next call would come against the team with fewer fouls. For instance, when the home team had five or more fouls than the visiting team, there was a 69 percent chance the visiting team would be whistled for the next foul.
Whistle Bias
Speaking of whistles, some are quicker than others. I hope coaches pay attention to this one. Ken Pomeroy did a big study last year that found Ted Valentine and Roger Ayers called significantly fewer fouls than any other NCAA referee -- about 3.5 fewer per game. (My wife says Ted spends too much time staring at his reflection on the floor.) John Hampton is another slow whistle guy -- he calls 2.7 fewer fouls. So if you get two of those guys in the same crew you can presumably play ultra-aggressive defense. Of the guys we saw more than once Dorsey, Desau, L. Jones, Gaffney and Sirmons all call noticeably more fouls than average -- all in the +1 to +1.5 range.
Incidentally, we saw Dorsey and Gaffney together twice (the only pair who repeated all season) and that pair calls 3.01 more fouls than average. Those two games produced 42 and 40 fouls, well above the average of 34 fouls in our other games.
Is there a fatigue effect?
Referees often work as many as 90 games. Doug Sirmons, who did 4 of our games, worked 91 games in 21 states for 20 different conferences. Roger Ayers, who did 5 of our games, did 88 games in 24 states for 21 conferences. That's like 5-6 games a week, with significant travel time between games. The travel time alone has got to be fatiguing. (The multi-conference officiating alliance formed a few years ago was designed to alleviate that with better scheduling, but I haven't seen whether it really made a difference.) There is no longer a website that tracks referee workloads, so the effect of travel fatigue can't be quantified. But I have a hard time believing that it doesn't affect foul calling.
That's it. Got to prep for today's game....
