One of the trends I've mentioned in recent DSS charting posts is the surprisingly low success that the Kentucky Wildcats have had when grabbing the numerous missed shots they create. In conference play, the Cats are grabbing 66.6% of opponents' missed shots (per KenPom). That mark ranks 6th in the SEC and would be tied for 216th in the nation. In all games, Kentucky is grabbing 69% of available defensive rebounds, good for 117th in the country.
The fact that UK has been better overall than they have been in conference play indicates that they were doing much better before conference play began. That's not surprising given the respective level of competition between the conference and non-conference schedules, but UK faced a number of excellent offensive rebounding teams in November and December including North Carolina (6th in the nation in OR%), Loyola MD (12th), Old Dominion (16th), and Lamar (21st), and Louisville (38th). It's not like they beat up on patsies in this particular area.
Indeed, when Kentucky entered conference play they were allowing teams to grab 31% of their own misses against a set of opponents who were otherwise grabbing 35.1% in their other games. The most common explanation is that as a shot blocking team, Kentucky often sees their best rebounders out of position as a result of attempts to block shots. I think that's likely part of it, but I'm not convinced it's the whole story.
For the Vanderbilt game, I decided to keep track of where rebounds came from, based on the location of the missed shot that created the opportunity. I marked down whether Kentucky or Vanderbilt grabbed the rebound and whether it was a player rebound or a team rebound (more on that in a second). The results are in picture form below.
Before we get to the rebound location, here's some background as to how rebounds are credited. They are divided into 3 categories, of which only 2 are relevant to the conversation.
Team rebounds might or might not show up in a box score depending on where you look. They do show up in the ones available on UKathletics.com and are included in determining Offensive and Defensive Rebound%. Deadball rebounds are rarely included in a box score, although they are listed at UKathletics.com. They are not included in the Rebound% calculations and so I am not concerned with them.
I initially tried to keep track of offensive rebounds that occurred because of help defense leaving the eventual rebounder, but that proved to be too much to do along with everything else I was trying to keep track of. So I'm afraid I don't have anything stronger to say about that theory right now. The combination of the results on blocks and the good job the Cats did rebounding after long jump shots are certainly in favor of the block theory.
In hindsight, it might have been more useful to keep track of where the rebounds were grabbed instead of where the shots were taken. In the end this may not matter much. In this game, Vanderbilt scored 15 points on 14 possessions where they grabbed an offensive rebound (they had some possessions with multiple offensive rebounds) for an efficiency of 107.1. That's not much different from their overall efficiency of 103.3 in that game. One game is hardly conclusive, but maybe good defensive rebounding is overrated in this instance? That would be a really counter-intuitive conclusion, but it might be interesting to explore further.
2 recs | 31 comments
The times when they don't get a block
You would think that we miss out on as many blocks as we get. If that’s the case, that would mean that there were around fourteen times when Vandy had the ball inside with either Davis or Jones out of position (having gone for the block). That seems like a lot of possessions where we’re at a defensive or rebounding disadvantage. I know this particular analysis couldn’t prove it, but I really do believe our aggressive nature on defense is part of the explanation for our less than stellar rebounding.
chicagoblues - February 16, 2012
It's something I want to look into further
When I have time, I’m going to re-watch some games so that I can focus attention to what happens when multiple players go for a block.
JLeverenz - February 16, 2012
You saw the exercise in yesterday's practice
where the defender had to hold off an opponent (a pad-holding manager) until the ball hit the rim before going for the rebound. That’s exactly what Terrence didn’t do to start the Vandy game. He left Goulbourne three times, once to try to block Davis’ man, and twice to drift toward the rim as the shot went up. After leaving his man, Jones had no idea where Goulbourne was or what Goulbourne was doing. Goulbourne was completely free to manuever for three rebounds, one of them a slam to score the first points of the game. Cal blew a gasket. Or two. He blew two gaskets.
Wheatgerm - February 16, 2012
yup
Miller did sort of the same thing later in the half. On a long 2 by Taylor he drifted away after the shot instead of boxing him out and the ball bounced right back to Taylor. The possession ultimately ended with the foul against Jenkins on the 3pt attempt.
JLeverenz - February 16, 2012
Improvement Nevertheless
If you have the wherewithal review some of the earlier (Nov-Dec) games and compare the Cats block out technique then to now and I think you’ll find remarkable improvement — perhaps an overlooked portion of this team’s significant maturation. That said, I doubt this team will ever be an outstanding rebounding team and due mainly to the fact that they won’t be together long enough. Ah, yes, another price to pay for one-and-dones but worth it I would aver.
Wild Weasel - February 16, 2012
That's something I would like to do, time permitting
I don’t have any kind of decent memory when it comes to things like block out technique, so going back to compare then and now would be illuminating.
JLeverenz - February 16, 2012
There's no doubt that's true.
They practice it and work on it — we saw that yesterday.
Glenn Logan - February 16, 2012
FWIW altho we give up rebounds on defense we do a good job grabbing rebounds on the offensive end. That more than makes up for it.
The difference between our eff FG% and the eff FG% of our opponents is so large. The fact that teams don’t gain anything at all as far as rebounds go when you account for our offensive rebounding is fine with me.
Plus if it’s due to the blocks, getting defensive rebounds might come at the expense of blocks.
Syracuse isn’t good at defensive rebounding either (prob cause of zone) and they are doing pretty good as well lol.
TheAnswer1313 - February 16, 2012
It may not be as big a problem as I thought
In the abstract, it would seem that defensive rebounding is something a team needs to be good at in order to do well defensively – it’s not much good to force a missed shot if you can’t get the ball back and allow the opposing team to just try again. But maybe there is something to how the Cats are defending that causes a lot of the offensive rebounds to be concentrated in only a few possessions that helps to limit the actual number of second chance points.
JLeverenz - February 16, 2012
There are a few factors that we haven't considered.
For one thing, a blocked shot recovery counts as an offensive rebound. Based on examining a few games, blocks go about 2-1 to the offensive team, meaning they get credited an offensive rebound for a shot that had zero chance of producing points.
If, for example, neither team had a block in the game, our DR percentage would likely have been significantly higher, although we may have also lost the game by Vandy getting off more shots, but that’s ancillary to my point. The point is, shot blocks produce a higher percentage OR per incident than regular shots do.
Intuitively, I know this matters, and has some impact on our DR numbers, I’m just not sure I know how to properly quantify it. But the reality is, shot blocking produces artificially inflated OR numbers because of their nature — most shot blocks tend to go either out of bounds or right back to the shooter.
Because Kentucky blocks a very high number of shots, it is my suspicion that the OR’s created by the blocks have a significant impact on the way our rebounding statistics look. I’m just not sure how I can prove that, exactly.
Another part of the value of a blocked shot versus a regular FGA is that it has effectively zero chance of success, which means every blocked shot that the defensive team recovers is exactly like a turnover than and not like a missed FGA, even though it only turns up as a DR in the stats.
Glenn Logan - February 16, 2012
On further review ...
… it may be closer to 6-5 than 2-1, which means the impact is smaller than I originally thought.
Glenn Logan - February 16, 2012
Millions would disagree, but this is interesting.
What effect do blocked shots have on our DR? Another way to put it would be, what effect do blocked shots have on the offense’s probability of scoring on that possession, as DRs are available only on missed shots.
I like the turnover analogy. A blocked shot reduces the probability of the FGA to zero, thus amounting to a turnover if defensively rebounded. But because a shot was attempted and not made, both teams have an opportunity to gain possession of the loose ball, something that can’t be said of all TOs (a travel call, e.g.). Thus a DR in the stats.
If offensively rebounded … the offense has much the same probability of making a basket that it had before the block. But less time on the shot clock might force the offense’s hand and lower the probability of scoring, thereby increasing the chance of a DR or TO. How quickly an opponent shoots the ball, it seems, should figure into the equation.
If the block comes early enough in the shot clock, the recovering offense has basically a second opportunity to succeed in making a basket, effectively doubling the chances of scoring on that possession. If those chances were 1 of 3 on the first attempt, they jump to 2 of 3 for the possession. Who takes the next shot, and from where, would seem to matter. To the extent that the offense claims more rebounds off blocked shots (Vandy 7, UK 5), the difference must improve the odds of scoring unless time is of the essence. And those increased odds must decrease our chances for DRs, though it seems we can’t be talking about anything significant.
Wheatgerm - February 16, 2012
Well, interesting points.
I boil it down to this, taking your theme of when the shot is blocked as relevant, because I think a case can be made that it is:
If a shot is blocked early in the possession (i.e. inside the first 10 seconds), then the recovery has substantially the same value as an offensive rebound.
If it occurs in, say, the last third of the shot clock, I would say it has much less value than an offensive rebound and trends toward the value of a forced shot.
Any block recovered by the offense within the last five seconds would be just like forced shot.
How to quantify that? I have no clue. :-)
Glenn Logan - February 16, 2012
Although I am not the stats guy, I have to ask the two of you (and anyone else for that matter)
Are we at the point where we can assess a value to the blocking of a shot depending on when it occurrs? And If we are, does that also mean that where the shot is blocked means as much as when?
I mean do we have the wherewithal to account for which is better, a 3pt shot blocked with 10sec remaining on the clock, or a layup that was blocked on an attempted fast break?
I mean, to me, if we are blocking 10-12 shots a game, that seems to be a fairly difficult number for the opponent to overcome. Assuming that a team shoots 60-70 shots per game, that is a significant %. But can we really get a true representation, and ultimately develop a different strategy based upon what kind of block, and when it occurs?
Greg Alan Edwards - February 16, 2012
That's a good question
I think trying to break down blocks by position and time remaining is not likely to lead to any kind of definite conclusions because of how finely the data is getting parsing.
Statistically speaking, in order to draw meaningful conclusions we need a lot of data – the more the better. If we start dividing up blocks by time/position we lose any ability to determine if time/position have an actual affect or if the result is random (ie due to other factors). To use your example, UK may only have 1 or 2 block all year that come within 10 seconds and are on 3 point attempts.
JLeverenz - February 16, 2012
This is something I've been thinking about
Because I wonder how true it is for UK? Davis in particular seems to do a really good job at directing his blocks to a place where another UK player can get the rebound, but maybe that’s just selective memory on my part.
It may be worth going back and finding out what happens in possessions in which UK blocks a shot and what happens when they don’t.
JLeverenz - February 16, 2012
It would be impossible to quantify
but the intimidation factor looms in the analysis. Davis blocks a shot, the rebound goes to the shooting team who then goes up with another shot. That followup shot will likely be hurried or significantly altered because Davis is still nearby and can come out of nowhere to block again.
kywineman - February 16, 2012
Could be.
It’s difficult to quantify because we really don’t keep statistics for “intimidations.” JLev does a good job of capturing some of that in his defensive style sheet as a defended shot, but that’s still a level of detail that is very hard to capture.
Glenn Logan - February 16, 2012
Yeah, it's hard to know if a player was intimidated or not
For example, I would not say that Vanderbilt was ever intimidated by all the shot blocking. They continued to try to go inside even after Davis and Jones had blocked several shots.
JLeverenz - February 16, 2012
Perhaps intimidation is not the best word to use.
Anticipation might be better. The players may continue to drive to the basket but their awareness of the threat might lead to thinking and shot alteration, both of which are most likely going to be working against the success of a shot attempt.
kywineman - February 16, 2012
I can see that
But I think that falls into the range of “things that can’t be quantified”. The only thing I could see is if Opponent’s 2pt% on unblocked shots is lower than it would be expected to be, but that would be really flaky evidence.
JLeverenz - February 16, 2012
Intimidation
There must be an app for that, if not now soon.
Wild Weasel - February 16, 2012
Yea...
..i can see it now. The Anthony Davis Call Blocker :-)
JohnPike - February 16, 2012 via Android app
I NEED THAT APP!!!!!!!
And can I get it for the land line, too? That’s the worst. Do Not Call List my ###.
GriffinRC - February 18, 2012
More Statistics: Cats Vs Orange
KSR provides a statistical comparison of Syracuse and UK — really, really close.
Wild Weasel - February 16, 2012
Interesting analysis
It’s actually surprising that Vandy managed to corral just one offensive rebound on 3 point attempts. I think there’s some statistical analysis (or maybe it;s anecdotal?) somewhere that says 3 point attempts are more likely to be rebounded by the offensive team, since the ball tends to bounce farther off long misses.
jc25 - February 16, 2012
I've often heard that
I think I recall reading somewhere (it might have been Pomeroy) that missed 3pt shots get rebounded at about the same rate as 2pt shots by the offense. I don’t remember where or when I read it though.
JLeverenz - February 16, 2012
Great Discussion
I have been pondering the same things. I have long thought blocks were generally overrated, but I don’t think that about this team. I also don’t think I would have thought that about Bill Russell’s teams if I had seen them. (Which might be good data to go look at if it is available.)
My eyes tell me much of what said above is correct, but think there are even more variables.
First, I think missed blocks are a bigger part of the OR problem than actual blocked shots. A block usually goes back the way the blocker is facing so he has a good chance recovering it. However, a missed block takes the player attempting the block away from the basket and essentially eliminates him as a rebounder. IE the ball is behind him.
Second, with all the switching this team does, the 4 and the 5 are often away from the basket which takes them away from most rebounds. Which would be ok, if the other team’s 4 and 5 were away from the basket as well but they are not. In that Vandy game I was incredulous at the effectiveness of Wiltjer in guarding Tinsley or Jenkins 15-18 feet away from the basket..
Third, there is such a thing as a “bad miss” and by that I mean the ball is shot so poorly that it doesn’t follow the normal arc and bounces off the rim at unusual angles. Which results in the ball skipping away from the basket or dropping straight down. I think a bad miss turns a 70/30 ball into something like a 50/50 ball. (I first started thinking about this last point years ago while watching a game at Denver Christian HS. Twice a kid shot a FT that clanked hard to a corner and she recovered the ball. DC kept the ball made the basket each time and won the game by 3 or 4.)
If you have the data and the video, consider trying to tease out missed blocks and do a with/without to see if there is a big difference in the split of rebounds. .
All that being said, I think the biggest problem is guys not holding the block out and “leaning” to the other end of the floor. Get that corrected and the splits will be better.
darkandbloody@gmail.com - February 17, 2012
I think you have a good point about missed blocks.
As I said before, shot blocking teams historically don’t put up great DR numbers. Remember 2010? as good as they were on the offensive glass, they were only a pedestrian 108th in defensive rebounding. 2009 UConn was another great shot blocking team, yet they only managed to be 55th nationally in DRs.
That evidence is hardly dispositive, and I don’t really have the time to effectively analyze shot block% against DR%, but my suspicion is that once you get above about 15% blocks in a game, it begins to affect your defensive rebounding.
Glenn Logan - February 17, 2012
Missed blocks
Good comments dark. The missed blocks are what interest me the most and I originally tried to pay attention to them while watching the Vandy game. Unfortunately it was just too much to keep up with along with DSS charting, 3pt tracking, and rebound placement. I do have plans to go back and watch some games and look for it specifically.
Your comment on bad misses is apt as well, although I’m not sure that UK is affected by those any more than any other team.
JLeverenz - February 17, 2012
Bad misses
I said too much without saying much. I think the Cats play such good defense they create more bad misses than most teams and I do think they tend to be 50/50
darkandbloody@gmail.com - February 17, 2012 via mobile
You must Login with your SB Nation account and be a member of A Sea Of Blue to post a comment.