Ken Pomeroy had an interesting article up the other day about offensive rebounding. One of the points that caught my eye is this one:
Every single season since 2007, the chances of a missed shot being rebounded by the offense has decreased as stopping transition is prioritized at the expense of getting second chances. If you listen to enough coaches, it doesn’t take long to understand that this is the prevailing thinking, though it may be a bit misguided. You still have Jamie Dixon, Rick Barnes, Tom Moore, and Pat Skerry out there holding the line on attacking the glass, but their ranks decrease with each passing season.
Pomeroy’s reasoning is that fear of transition is what is keeping teams from hitting the offensive glass more. He isn’t stating this axiomatically, but more in the sense that this is how coaches see it, and so the emphasis has shifted more to transition defense than offensive rebounding.
This has the perfect tie-in to Kentucky. Last season, transition defense was a huge problem with the Wildcats, so the natural assumption might be that the reason this was so was due more to an emphasis on offensive rebounding than on transition defense. Supporting this theory was the fact that Kentucky was the #2 team in the nation in offensive rebound percentage, (Kenpom.com) at almost 42%, with only Quinnipiac having a higher OR%.
So let’s look at this in a bit of detail. When a shot goes up in basketball, here are the possible outcomes:
- A defensive rebound
- An offensive rebound
- A foul
- A made basket
Only one of these outcomes, in a well-played game, will produce a transition opportunity for the offense. So it seems on it’s face that more offensive rebounds act to limit transition opportunities, at least in theory. Transition opportunities also come on inbound plays and turnovers, but on inbound plays, they can only happen if the defense is unprepared or out of position.
A quick comparison of hoop-math.com’s transition numbers is unhelpful. Just for example, Quinnipiac, the best offensive rebounding team in the country, wound up in defensive transition 44.2% of the time, 7th most in the nation. Grambling, who wound up in transition the most, was one of the least efficient offensive rebounding teams. Kentucky, who was the second best rebounding team, wound up with shots taken against them in transition 19.7% of the time; 258th.
Of course, this simply doesn’t account in any way for differences in personnel and competition. Kentucky competed at the highest level of college basketball, while Quinnipiac and Grambling did not. Personnel does make a difference.
Fortunately, a study by some MIT folks on NBA teams provides us general guidance about this issue, although even their analysis cannot factor in personnel, but rather examines it from the general perspective of a player. You can read this if you are so inclined, but I tend to skip the tricky parts and head straight to the conclusion, which is:
In conclusion, our results suggest that focusing on the offensive rebound immediately after the shot goes up seems to trump the gain a team gets with ahead start on getting back. In the case of a defensive rebound by the other team early threat neutralization (as opposed to merely getting back early) can help limit the negative impact of transition baskets.The generalizability of these conclusions is limited by the data. For some teams we lacked data. Moreover, there are many factors we have yet to consider, e.g. the positioning of the defensive players, the game situation and especially the personnel on the floor.
What they have come up with is this general premise: the value of crashing the glass seems to exceed the value of making sure you get back on defense. The reason this intrigues me is obvious: Kentucky is going to once again be one of the tallest and biggest teams in the country next season. Generally (and we all know that this doesn’t universally apply), the taller team has an advantage when it comes to getting rebounds off of missed shots.
Grantland also has an excellent article on this subject, citing the MIT study as well as pointing out that a couple of NBA teams seem to have taken the study to heart:
The authors have since looked at data from the 2012-13 season and found the same effect, says Jenna Wiens, the lead author and a PhD candidate at MIT. (She also once played water polo, meaning she is automatically cool.) And as a fun aside, Wiens and her team found two teams changed their offensive rebounding philosophies dramatically between 2011-12 and last season: Minnesota and Washington. The Wolves appear to have sent an extra player, on average, to chase offensive boards, while the Wiz focused more on transition defense, Wiens says.
The Wolves nugget speaks to the importance of personnel. In 2011-12, they had Kevin Love, one of the world’s great offensive rebounders. Love missed most of last season, and the team shot a pathetic 30.5 percent on 3-pointers. When you have Kevin Love, you can sustain on the offensive glass without over-pursuing. When you don’t have Kevin Love, you might need to send an extra player, especially when you know you are a horrible shooting team that needs extra chances to survive.
It’s important to note that Kentucky did not crash the glass last year as a rule. They kept people back to maintain defensive balance, but it absolutely did not work well. In fact, Kentucky often got beat down the floor off the inbounds play after a made basket, and it wasn’t because they pressed.
But the statistical evidence supports the idea that if you send an extra person to the offensive glass, you will gain a net benefit, especially if you have personnel suited to that. Like Minnesota’s Kevin Love in the example above, Kentucky had one of the NCAA’s great offensive rebounders last season in Julius Randle, so sending an extra rebounder was unnecessary. In theory, it looks like Kentucky will once again have the kind of size and energy that it takes to get offensive rebounds in large numbers, but will they have a Randle-eque rebounder who can take care of business without help from the wings?
If not, it may actually benefit Kentucky to send an extra player inside for offensive rebounding purposes. The net benefit seems to be clear if you have the personnel to support it, and I think Kentucky will, especially with Poythress playing the three. That will put more pressure on the guards in transition, something that they haven’t executed very well to date, but I have suspicions that was due more to a lack of conditioning than anything else, related to the Harrison’s late arrival to Lexington.
Having said all that, there is this final observation by Zach Lowe, author of the Grantland article:
There is a growing sense that teams have gone too far in offensive rebounding paranoia. Jeff Van Gundy wonders if this is especially true in the playoffs, when the pace slows, teams face the very best defenses, and points become more scarce. "You have to ask yourself: In the playoffs, is offensive rebounding more important?" Van Gundy says. "Because scoring is harder. And so, should we construct an offensive rebounding identity early, so that we have another weapon in the playoffs?"
And yet, the correlation between offensive rebounding and winning is still very low, according to several stats experts around the league. Is there a chicken-and-egg thing going on there? Or does offensive rebounding really not matter, as Van Gundy and other coaches have found in the historical data? After all, we’re talking about only a handful of possessions each game. Teams snag only 11 offensive rebounds per game on average, and not all of those lead to baskets or free throws. Would jacking that number up to 14 or 15 really make a difference, especially considering that the very existence of an offensive rebounding chance flows from a negative event — a miss? Maybe teams who are "good" at offensive rebounding are good because they miss a lot — because they are bad at offense.
You may wonder how the NBA stuff translates to college, but my feeling is that NBA trends translate pretty well. There is probably an ideal balance of times when extra players should hit the glass, and when they should stay back. A lot of that no doubt depends on the scouting report, and the tendencies of the opponent.
But as a general proposition, I would posit that the fear of transition is not a good reason to avoid crashing the glass. In the general case, it’s better statistically to crash the offensive glass than not.
Nominally, of course, it’s better if you never have to get an offensive rebound. Alas, that is never the case in reality.