Vote Yes on Prop B(et)
TLDR: Bet that top rebounders will outperform when playing the Rockets
Last year I theorized that top rebounders should outperform their rebounding averages when playing against three-point shooting teams. The general logic was that more three pointers lead to more misses, which translate into easy rebounds. I ran a bunch of analysis that supported my theory – top rebounders did indeed have higher rebounding averages when playing against three-point shooting teams.
Why Does this Even Matter? $$$ of Course
There’s a prop bet that allows you to bet on the over/under rebounds for individual players. The over/under line is typically set around the player’s averages. For example, Karl-Anthony Towns averaged about 12 rebounds per game last year. The over/under line would thus be set at 12, and a bettor could wager that KAT would either have less than or more than 12 rebounds for an individual game.
Not Being Lazy Anymore
I took another look at the numbers and found a major flaw. Since I used averages instead of specific games, there was a large amount of skew. For example, KAT played the Dallas Mavericks four times last year and recorded 20,12,11, and 12 rebounds. That averages out to 13.75 rebounds, which implies he outperformed his season average of 12.3. However, on an individual game basis, he underperformed the 12.3 average three out of four games. The one 20 rebound game significantly skewed the average.
I’ll be honest – I was too lazy in my first analysis to look game by game for each rebounding player. Without any sort of database this is a huge manual effort, but this time around I figured I should. I analyzed the top ten rebounders, game by game, when they played the top five three-point shooting teams. My original theory didn’t hold – some top rebounders outperformed their averages more times than not, some didn’t. Overall it was about a wash.
FYI – I defined the “top rebounders” as the top ten players who ranked in the highest defensive rebound percentages. That is to say, those players grabbed the most defensive boards relative to their team. I figure this subset would benefit from the missed threes the most. I also filtered for players who averaged at least twenty minutes per game.
As I went through the cumbersome process described above, I found a much stronger correlation against the Houston Rockets. Surprise, surprise. The Rockets of course attempted, made, and missed the most threes last year. They averaged 41.4 attempts, which is 16% higher than the second closest Brooklyn Nets.
The graph below shows how each top rebounder performed when playing the Rockets, relative to his overall season average. In total, there were 14 games of outperformance and 7 games of underperformance. We’ll call the outperformance games “Wins” and the underperformance games “Losses”.
If you were to bet $100 on each of these games, you would make a profit of $517 net of bookie fees (assuming the bookie fees are -115). On a total investment of $2,100, that translates into a solid 25% return.
Picking and Choosing
This may seem like I’m picking and choosing to juice the returns (which I am), but there are some qualitative factors at play here. The table below shows the average minutes per top rebounder when playing the Rockets. Notice that there’s a direct correlation – those that averaged at least 30 minutes were successful. Cutting out these bottom four players leads to 13 wins and 1 loss, for a whopping 74% return.
Why couldn’t these four players at the bottom stay on the court? I would suspect that they either are bad matchups against the Rockets (i.e. Harden or Paul will take advantage of them on switches), foul trouble, blow out games, or all the above.
What about Pace?
I also looked at how top rebounders performed against the highest pace team, the New Orleans Pelicans. Its logical that higher pace means more shots means more defensive rebounds, but the table below actually shows a net loss.
I’m going to test my theory throughout the season and will hopefully make some money. Check back in a year!
Note: All numbers were sourced from NBA.com statistics