Jam Session: Handicapping the Dunk Contest
By Matt D'Anna
flickr |
Here’s a bad idea: let’s analyze the NBA Slam Dunk Contest based on the seemingly-unrelated-yet-publicly-available statistics on previous winners and current contestants, identifying patterns, anomalies, and commonalties among the past and the present.
Challenge. Accepted.
In order to tackle this bold endeavor, I’ve devised a somewhat logical three-step process to predict this year’s Slam Dunk Champion (assuming there isn’t confusion on who actually wins, like last season; I digress).
1. Average Champions
What does the average dunk champion look like? By examining the basic box score stats — at the per 100 possession level — we can develop basic ranges of “normal.” Specifically, finding the mean and standard deviation for each stat creates a window that incorporates 95% of the prior champs (which is roughly 26 of the 28 winners since 1984). Then, a comparison of this years contestants across these 21 metrics looks…like…this:
A few caveats about this chart. First and foremost, these stats are based on only the season a player won the title; the entire season. This is an inherently flawed comparison, but we can at least observe some trends. Second, Hal Brown has developed a much stronger metric for measuring consistency that I am not utilizing here. That metric is designed to capture the degree of variance from the norm; depending on how well the results of this article pan out, we may have to dip into that method for the next dunk contest. For now, these ranges examine which of these four players fall outside (either above or below) these established ranges. Third, the only ‘standard’ box score stats that were excluded were defensive rebounds and total rebounds; both were significantly skewed. Having said that, Mason Plumlee clearly falls outside of the norm. Interestingly, this may be more of an indictment against Big Man dunk champs; regardless, he would appear to be going into this contest at a pretty big disadvantage. But remember, once upon a time a young Mason did this:
Further, a note about NetRTG. The chart describes an even split in offensive and defensive
on average
. Well no, not really; this is a cautionary tale to the allure of averages. As you can see, there’s nothing average, ordinary, or normal about this set of NetRTGs:
The point here: players have been a part of really good (see: Jordan, Michael) and really bad (see: Jordan, Baby) teams, and still won it all. However, no one has been on as bad of a team as Zach LaVine currently is. A -23 NetRTG would be historic for a champ.
So far, Giannis Antetokounmpo and Victor Oladipo have the slight stats-based edge.
2. Textual Clues
A significant portion of previous slam dunk champions — and contestants overall, really — are first and second-year players. Coming fresh out of high school and college, often the best data available are scouting reports. A manual scrape of relevant scouting reports from Draft Express of the last ten champs (starting in 2000 with Desmond Mason), reveals some of the most commonly used terms:
Names aside, there’s some interesting patterns. Let’s compare this to each of this year’s contestants:
In a visual scan that admittedly lacks scientific scrutiny, it appears LaVine and Oladipo have scouting report closest resembling the previous champs.
That’s two categories for Oladipo so far, with the Greek Freak in a close second.
3. Dunk Counts
BREAKING: Historic data on dunks is not robust. Basketball Reference has reliable count information since 2001-02. Thus, since Jason Richardson we have 12 years worth of dunks per game and dunks per minute to compare to this year’s crop.
The graphs only differ slightly. Despite the skewed appearance, consider this: there’s three big men (Griffin, Howard, and Smith) and three Nate Robinsons. They essentially cancel each other out, and make the average (a shade under one dunk per game and a shade over one dunk per 36 minutes) a decent measure. Further, when comparing this year’s quartet to past, Victor Oladipo just barely edges out Zach LaVine as the closest to average. Victor’s counts resemble the Dunk Firm of Wall, Ross, and Green. Plumlee enters the Griffin zone, and Giannis falls in an uncomfortable No Man’s Land.
The Result
So when it’s all said and done, how do they finish?
That’s Oladipo, LaVine, Giannis, Mason if you couldn’t tell. If this “prediction” is based on some sense of rationality, it’s worth noting that this is the Slam Dunk Contest, an event judged on irrationality and uniqueness. I’d say my confidence in this prediction is best summarized as:
Data and photo support provided courtesy of Basketball-Reference, Draft Express, Jason Davies and his wonderful World Clouds, and Wikipedia – yes, Wikipedia.