Last I left off on discussing the referees, I discussed some general issues with the Last Two Minute (L2M) Reports and their reporting patterns. It’s evident that, over time, the L2Ms have increased their rates of grading “no-call” plays which may caused by an effort to increase the reported accuracy of the officials. It may also be that the NBA is simply having an adjustment period to how they should grade games based on the L2M criteria. I urge you to read what I wrote before for the finer details, but the L2M reporting procedures have certainly changed since they were first released.
But this is an Atlanta Hawks-eccentric website, so it might be helpful for me to dive into how the L2M reports have graded the Atlanta Hawks. So let’s try and put some perspective on how the L2M have deemed the officiating at the end of Atlanta games since March 2015.
Total Games and Calls
The NBA has pre-defined the games which qualify for a L2M, which is nice for us because it shows that the NBA does not pick and choose which games to release their evaluations to the public. Only teams which are involved in a game which is within 5 points at the 2:00 mark of the 4th quarter (and apparently if the game goes to overtime or a team trailing by more than 5 points at this mark wins the game…) is evaluated. There cannot be any favorites from the NBA on this issue, unless you think the NBA systematically rigs games through their referees so that certain teams are going to play closer games than others.
It’s a bit pointless to look at which team has been involved in the most L2M reports or has the most CC, CNC, IC, or INC…that will basically reflect teams which have the most games which are close. Instead, let’s use player information to determine which teams are “most favorable” to the referees. What I mean by this, is let’s rank the NBA teams by the frequency at which the L2M reports deem the committing player’s team was deemed an INC out of all the graded plays from the reports:
We can see that Atlanta is currently 6th in terms of number of reports and tied for 2nd for total number of quarters reported (thanks to the 4 OT game against the New York Knicks). This has nothing to do with the referees and is mainly just the luck of the draw, a team like Toronto is more likely to be within 5 points at the 2 minute mark than a team like the Los Angeles Lakers. About the only thing the number of reports can help with is our “small sample size” issue – ie that certainty of a result improves with the number of observations. We can be more certain about results from the Raptors than we are for the Lakers in terms of the following sections.
Referees Biased Against Road Teams?
It’s often suggested that the officials are favorable to the home team, but is this really the case? Or in our context, are the referees affected by the home crowd as graded by the NBA? Let’s go to the tape:
Each row sums to 100% and it looks like there is a bit of a difference in the way that the referees call a game for a home team versus an away team. If you’re a home team, then expect more no-calls (correct or incorrect). This does not necessarily mean the home team more aggressive, this relationship might just reflect that home teams are more aggressive already. Without more data, it’s hard to disentangle what it is here that causes the difference in no-call versus call percentages.
But what the above information is helpful for is a nice baseline. If Atlanta has a home-court advantage, then we should see a divergence from the the home averages for CC, CNC, IC, or INC percentages. That would tell us, relative to the rest of the league, that Atlanta has a better/worse outcome based upon some sort of a home crowd influence.
So, does it exist? Let’s take a look for Atlanta’s splits across home and away:
It appears that as we’re looking across the different L2M reports for Atlanta…there’s not too much of a difference. I mean, if you squint hard enough you can find a difference but if anyone believes that Atlanta gets jobbed on the road, well that’s just not founded from what we have here. Keep in mind that I’m only referencing the last 2 minutes of a game here, so it still might be the case that Atlanta is jobbed outside that time period. Or not. We don’t know.
Across The League
OK now let’s take a look across the rest of the NBA. Is there some sort of a difference across the rest of the NBA involved with the teams in particular? Well, let’s take a look:
It doesn’t seem like there’s anything aggregious here. Yes, there are some teams which have higher rates of INC and IC over this time (as judged by the committing player), but nothing is so out of whack that one should be upset. However, we’ve already established that the home team tends to have higher rates of INC and CNC. Might we see a greater difference if we break things down by home versus away? Let’s inspect:
Well there’s certainly some variation, but nothing seems egregious. It’s to be expected that there’d be some variation in these calls anyway due to the sheer number of teams. But at the same time, if the NBA Referees Union wants to claim that they are consistent across teams then that is a claim that doesn’t really hold water. They do vary across teams in terms of how they make their calls. But this is not nececssarily a fault of theirs, it might be that certain teams vary in how they play.
Again, as I mentioned in my previous post on the L2M reports, it’s hard to draw conclusions in the data. This should be seen as a call for the NBA to release the full 48 minute reports so that we can further evaluate the differences across teams in terms of how the referees evaluate them.