Officiating bias in the ACC
- Thread starter TheBrickhouse
- Start date
This article did nothing for me. I don't see any real evidence about any bias.
They said they looked at 3500 games, and the ACC and the Big 12 favored the home team regarding penalties. Well, what about the other conferences? If they either did not favor either team, or favored the visiting team, are they also going to talk about a bias there too? Out of a 3500 game sample, there is no way to come up with data that will support no bias one way or the other. Do they really think all conferences, 3500 games later, should end up dead even in penalty yards for the home team and visiting team?
They also mentioned that favored teams were penalized more in the ACC and Big East. Once again, what about the other conferences? Out of 5 power conferences, and 3500 games, why is it surprising that a study might find that 2 of the conferences had favored teams penalized more? Wouldn't this also mean that favored teams in the other conferences were penalized the same as the non favored team or less? Duh! Out of 3500 games, no way would a study find that all conferences ended up equal in this data.
These guys proved nothing.
They said they looked at 3500 games, and the ACC and the Big 12 favored the home team regarding penalties. Well, what about the other conferences? If they either did not favor either team, or favored the visiting team, are they also going to talk about a bias there too? Out of a 3500 game sample, there is no way to come up with data that will support no bias one way or the other. Do they really think all conferences, 3500 games later, should end up dead even in penalty yards for the home team and visiting team?
They also mentioned that favored teams were penalized more in the ACC and Big East. Once again, what about the other conferences? Out of 5 power conferences, and 3500 games, why is it surprising that a study might find that 2 of the conferences had favored teams penalized more? Wouldn't this also mean that favored teams in the other conferences were penalized the same as the non favored team or less? Duh! Out of 3500 games, no way would a study find that all conferences ended up equal in this data.
These guys proved nothing.
Read the paper at the end of the article, which will answer a lot of your questions.
If you have more confusion after that, I am in close contact with one of the authors and can pass on your critiques, questions, and comments.
Fascinating report. So, during the years VT dominated the ACC they did it despite being an upstart, new to the league, and generally favored in almost all games. Pretty impressive to overcome all that and still be the dominant team in the league for those years.
So maybe I was right in complaining about all the questionable late hit penalties against VT during that period.
It's a shame this came out as VT is no longer an upstart, new to the league or favored in many games. We could use the extra help.
You are right. Can you remember any game that an underdog upset VT that may have cost VT a shot at the national title? Or BCS bowl game?
Sorry, but I still see nothing worthy of a bias claim. For example, they talked about the Big Ten having equal penalty treatment when they play in conference games, and they are penalized less when they play out of conference. That can be explained by the fact that most of their OOC games are against lesser opponents, including 1-AAs, so when you are a heavy favorite against a weak team, you usually have fewer penalties. And, if you are in a 30 or 40 point rout, officials will not call everything like they would in a 3 point game. Kind of like in basketball when a minor off the ball foul is rarely called in a 30 point game. But, overall, the difference is not significant enough anyway to show a bias.
Then, it talks about the ACC and the Big 12 home teams having 6 and 4 fewer penalty yards per game when playing in conference games than when they play OOC games at home. 6 and 4 yards? Are we really worried about that? So, out of hundreds a game, a Big 12 team might average 60 penalty yards a game in conference and 64 out of conference? That can be compared to the Big Ten being the other way by explaining that the ACC plays the hardest OOC schedules, and the Big 12, with 9 conference games, plays fewer OOC games. But, I think the best way to explain them is to question why anyone
would expect the numbers to always be even here.
Why assume the in conference and OOC opponents are equal? Are these examples where the teams were always playing similar type teams? Even if so, I still don't understand why this is telling. Did they really think the study of hundreds of games would result in an equal amount of penalty yards for in conferences and out of conference games? When you do a sample
like this with this 6 main units, some will be above, some in the middle, and some below.
These numbers are not weighted one way or the other nearly dramatic enough to scream bias, and are all easily explained, but I respect the time and effort the guys put into this.
You are right. Predicting penalty yards with variables is definitely not an indication of bias.
The argument is more nuanced than that. It compares two treatments (in and OOC), and where there is significant differences in the predictors for both treatments is a sign of bias. For instance, why would home ACC teams have a home field advantage when they play in-conference games, but don't have that same advantage when playing at home versus OOC?
Are you familiar with multiple regression? These predictors are all statistically isolated and team-effects are modeled as second-level fixed effects to remove any variance between teams/coaches/players.
brickhouse said,
"""""You are right. Predicting penalty yards with variables is definitely not an indication of bias.
The
argument is more nuanced than that. It compares two treatments (in and
OOC), and where there is significant differences in the predictors for
both treatments is a sign of bias. For instance, why would home ACC
teams have a home field advantage when they play in-conference games,
but don't have that same advantage when playing at home versus OOC?
Are
you familiar with multiple regression? These predictors are all
statistically isolated and team-effects are modeled as second-level
fixed effects to remove any variance between teams/coaches/players.""""""
You ask why home ACC teams have a home field advantage IC and not out? Who says they do not have a home field advantage OOC? From 2005-2012 (the years they used for the study), ACC home teams went 221-163 (.575) in conference and 185-64 (.742) OOC. Leave the 1-AAs out of the OOC number, and the ACC still went 104-59 OOC (.638).
For example, VT went 26-6 at home IC and 14-0 at home against OOC teams. So, good luck getting on here and telling VT fans that the refs were biased against us in our OOC games. Maybe we had more penalties in our OOC games because we routed a lot of these teams and our back ups committed more penalties than our regulars? Or, maybe the law of averages says some teams will have more penalties OOC and some more IC, and, as a result, not all conferences will end up exactly even with their IC and OOC penalties. Because this sample is done as a whole conference and not as a team by team basis, we do not even know if we had more penalties OOC or not, right? Do we even know which teams this favored more? Were some really dramatic, and if so, wouldn't it make more sense to focus on the individual teams than a whole conference?
I am familiar with multiple regression, but it does not matter how many variables you try to use to predict this "bias" if the so called "bias" does not exist here.
"""""You are right. Predicting penalty yards with variables is definitely not an indication of bias.
The
argument is more nuanced than that. It compares two treatments (in and
OOC), and where there is significant differences in the predictors for
both treatments is a sign of bias. For instance, why would home ACC
teams have a home field advantage when they play in-conference games,
but don't have that same advantage when playing at home versus OOC?
Are
you familiar with multiple regression? These predictors are all
statistically isolated and team-effects are modeled as second-level
fixed effects to remove any variance between teams/coaches/players.""""""
You ask why home ACC teams have a home field advantage IC and not out? Who says they do not have a home field advantage OOC? From 2005-2012 (the years they used for the study), ACC home teams went 221-163 (.575) in conference and 185-64 (.742) OOC. Leave the 1-AAs out of the OOC number, and the ACC still went 104-59 OOC (.638).
For example, VT went 26-6 at home IC and 14-0 at home against OOC teams. So, good luck getting on here and telling VT fans that the refs were biased against us in our OOC games. Maybe we had more penalties in our OOC games because we routed a lot of these teams and our back ups committed more penalties than our regulars? Or, maybe the law of averages says some teams will have more penalties OOC and some more IC, and, as a result, not all conferences will end up exactly even with their IC and OOC penalties. Because this sample is done as a whole conference and not as a team by team basis, we do not even know if we had more penalties OOC or not, right? Do we even know which teams this favored more? Were some really dramatic, and if so, wouldn't it make more sense to focus on the individual teams than a whole conference?
I am familiar with multiple regression, but it does not matter how many variables you try to use to predict this "bias" if the so called "bias" does not exist here.
The data says that there is not an ACC home advantage OOC.
This was a study of penalty yards, not wins and losses.
There were 6000+ observations. The number of variables used was around 11 - year dummies, betting line, home field dummy, opponent penalty yards, total plays. It was a clean, parsimonious model with more than ample power.
Why do you believe that bias doesn't exist in the face of this evidence? Controlling for game effects, it is a strained argument to think that teams get significantly more penalties when the play OOC games in the ACC, whereas that doesn't happen in the other 5 conferences. Have you read the study yet?
I do not think the bias exists because the evidence is weak, and does not support anything.
When using a 3500 game study, why would anyone expect penalty yards to end even among all of the conferences? When a study is done like this, one conference will have to have the most penalties, and one will have to heave the least. It does not mean anything is biased.
For example, if you did this same research on the NFL, you would find that one division had the most penalties when they played out of their division, and another division had the least. Nothing biased about it. Just common sense.
Same with college basketball. There is no doubt that one conference will look like it is getting a raw deal with the most fouls called in their OOC games, but, if we looked at 3500 games, we certainly should not expect all of the conferences to end up with the same amount of fouls called. Therefore, one of the conferences will HAVE to have the most in this category. Does not mean bias exists.
As I said earlier, I respect their time and effort they put into this, but there is nothing here.
When using a 3500 game study, why would anyone expect penalty yards to end even among all of the conferences? When a study is done like this, one conference will have to have the most penalties, and one will have to heave the least. It does not mean anything is biased.
For example, if you did this same research on the NFL, you would find that one division had the most penalties when they played out of their division, and another division had the least. Nothing biased about it. Just common sense.
Same with college basketball. There is no doubt that one conference will look like it is getting a raw deal with the most fouls called in their OOC games, but, if we looked at 3500 games, we certainly should not expect all of the conferences to end up with the same amount of fouls called. Therefore, one of the conferences will HAVE to have the most in this category. Does not mean bias exists.
As I said earlier, I respect their time and effort they put into this, but there is nothing here.
For those who think college football and basketball is bad, go watch a few college softball games.
