Testing Bill Self's theory: How was KU's late-clock defense against Texas A&M?


Kansas coach Bill Self had some interesting comments about his team's defense following the Jayhawks' 64-54 victory over Texas A&M on Monday.

The coach was most upset with his team's late-clock defense against the Aggies. Here's his quote:

"I think if you were to go back and look at us statistically — which I don't have any way to do this, but we probably need to start charting this — I would say the percentage of people scoring against us is far higher in the last seven to eight seconds of a possession than it is in the first 27. We don't finish possessions. How many times tonight did they score under five on the shot clock, or we foul? That's the thing that's so frustrating, because we're not finishing possessions, and that's something that we really emphasize."

So how poor was KU's late-clock defense against Texas A&M?

Let's take a look.

After going back to the tape, here is a graph showing each Texas A&M field goal, along with how much time was left on the shot clock. Please note I did not include Dash Harris' shot right before the halftime buzzer, as the shot clock was turned off.

As you can see, KU's late-clock defense doesn't appear to be as bad as Self remembered.

A&M scored only twice with less than 8 seconds left on the shot clock. That's only 10 percent of its field goals (two out of 20), which is much lower than Self's original prediction.

If you calculate it, the Aggies' average field goal came with 22.7 seconds left on the shot clock (among those 19 shots when the shot clock was on). That appears to be relatively early, especially considering A&M plays at a slow pace.

Here are the three times when KU was scored upon with less than 10 seconds on the shot clock.

With 12:20 left in the first half, KU center Jeff Withey switches on a ball screen, then doesn't get out far enough to contest Elston Turner's three-pointer.

On A&M's next possession (and perhaps this is why it stuck in Self's mind), A&M's Dash Harris gets by Elijah Johnson off the dribble, which forces help from Withey. Harris sees the help coming and dishes to Keith Davis, who puts in an easy layup.

Then, with 16:52 left in the second half, Withey helps with a ball screen, then trips and falls down in traffic. Elston Turner realizes this and is able to find David Loubeau for a wide-open layup.

To be fair, Self also said that KU fouling at the end of the shot clock should be considered as poor late-clock defense as well.

"I think it's breakdowns," Self said, explaining why his team was struggling late in possessions. "Travis (Releford) fouled twice because he reached. Or guys forget we're switching or trapping or whatever it could be. Then the other thing is, offensive teams are always most aggressive in the last five or six seconds of the shot clock because they've got to get a shot. So I think it's just a combination of things."

Going back to the tape, there were two instances of KU fouling with less than 10 seconds on the shot clock.

With 15:15 left in the first half, Tyshawn Taylor is whistled for an apparent blocking call ...

and with 17:40 left in the second half, Releford is called for a reach when he had help from Thomas Robinson behind him.

Taking the whole game into account, it doesn't appear that KU's late-clock defense is as much of an issue as perhaps the eye test told Self it was.

That doesn't mean the Jayhawks can't (and won't try to) improve in that area before their next game against Iowa State.

Postscript: A few commenters said they would like to see a graph that included KU's fouls and also the times when A&M scored points off an offensive rebound.

After going back to the tape, here's a graph that includes those two elements.

This seems to only support the statement above: KU's late-clock defense — at least against Texas A&M — did not hurt the Jayhawks as much as it might have originally appeared.


Rian Ankerholz 10 years, 8 months ago

Jesse: The numbers don't lie, but I remember having the same gut-level reaction as Coach Self experienced after the late-possession fouls. It is agonizing to let the other team off the hook after playing thirty seconds of tough defense. I am encouraged that, after a second view, the late-possession shots and fouls are not glaringly reflective of poor effort. I really like your attempts to apply some quantitative analysis to a rather subjective idea. As always, if you write it, I'll read it. Keep up the good work.

Rian Ankerholz 10 years, 8 months ago

p.s After looking at your graph again, it seems to me that it might be interpreted a bit more. What happens in the first five seconds of our defensive stand is reflective of our transition defense. The next ten seconds or so are reflective of the effectiveness of the other team's primary offensive set. As we might expect, that was the most effective period for A&M's offense, when they scored at least eight field goals (two more were on the fence between transition and the offensive set). After that, when the primary offensive set has not produced a score, the other team begins it's secondary motion. We were great defensively during that latter time. I think this gives some insight as to how our defense functioned, but I suggest it is more important as a lesson to our offense. Coach Self said the players were "selfish" (no pun intended) and we jacked up early shots outside of the flow of the primary offensive sets. That wasn't effective. Lesson: Execute the set plays first; over the long run, they give the offense the best chance to succeed if no transition bucket is made. The only exception that comes to mind was when Kentucky's length and interior defense really shut down our primary sets. Then we had to depend on improvisation, primarily Tyshawn's ability to create his own shot. In the final analysis, no Eureka moment here; just run the plays!

Carter Patterson 10 years, 8 months ago

Jesse, can you add KU defensive fouls to the chart?

Jesse Newell 10 years, 8 months ago

carter — I didn't chart all the fouls, but I might have time to do it tomorrow.

dylans 10 years, 8 months ago

How many of the short shot clock baskets were a result of a put-back? That should also be considered a end of shot clock error, I would think.

Jesse Newell 10 years, 8 months ago

Texas A&M only had two second-chance points in the game. That one field goal came at the 5:50 mark of the first half after a shot was blocked by Thomas Robinson. The original blocked shot was taken with 20 seconds left on the shot clock, so it wasn't a late-clock situation.

In other words, offensive rebounding didn't really have an impact on any late-clock defensive possessions against Texas A&M.

dylans 10 years, 8 months ago

Thanks for looking into it. What were the 3 shots made in 3 seconds or less? Seems fast for a fast break, were they inbounds plays?

Jesse Newell 10 years, 8 months ago

Dash Harris stealing an outlet pass from Withey was one. Harris picking Tharpe in the backcourt was another. The third was off of an offensive rebound following a block from Thomas Robinson.

dylans 10 years, 8 months ago

Thanks Jesse, it's great that you care enough to research and answer posters burning questions!

dylans 10 years, 8 months ago

Another point that I didn't think of last night- how many times did aTm miss after a defensive breakdown late in the shot clock. Bad defense is often times bailed out by a missed shot.

Stephen Simmons 10 years, 8 months ago

Thanks Jesse. I think adding fouls might tell a different story. Does your current graph chart free throws? In order to be fully complete, you could add fouls AND free throws made. That way if KU fouls with 5 seconds left on the clock we could also track how often it led directly to points. Slippery slope here, however, as fouls prior to the bonus that lead to another posession where A&M scored would also be fairly directly the result of a late shot-clock lapse, even if those points came early in the next shot clock.

Better buy some bigger graph paper.

Jesse Newell 10 years, 8 months ago

My best attempt at creating another graph, which includes fouls and free throws, is above.

ancient_hawk 10 years, 8 months ago

Scores early in the shot clock are also indicative of an offensive board and a put back.

Eric J. Baker 10 years, 8 months ago

I was wondering about this both during the game and especially after Bill commented on it, too. It's interesting that the fact of the matter is almost completely opposite what it seemed like. I think it being such a close game that when they make a shot late in the shot clock it's incredibly frustrating. Thanks for the analysis, Jesse!

Sparko 10 years, 8 months ago

I think Self's comments are actually illustrative of a team frustrated by a clog-it up zone. Kansas shot horribly from 3--probably the worst I have seen this year. But those shots were actually open looks. The team hits 40-percent of those, this is a 30-point cakewalk. A&M's strategy is one we'll see mid-majors like Davidson employee again in the tourney. Clog (sometimes with four players!!) the lane, hope Kansas forces the ball inside anyway or misses open looks. It can be successful if there is the perfect storm of bad-shooting, poor officiating, and a hot streak from three by the other guys.

Withey's Blocks and KU's perimeter defense in the Second Half won the game. Self may be frustrated by the defense, but the truth is completely opposite--the offense missed open looks and rebounds and allowed some junk run-outs. TRob had an awful FG night--but man did he come through at the line. Yesterday was one of the most impressive victories of the year. Kansas gutted it out and simply did everything it could to bring it home--short of draining long jumpers.

Eric J. Baker 10 years, 8 months ago

I don't think it can be mentioned enough that TRob was amazing from the line. If you had told me 2 years ago, or even last year, that he would go 10-10 from the free throw line with a big game on the line, I wouldn't have believed you. Just another facet of his game that has improved dramatically.

Scott Smetana 10 years, 8 months ago

Not much basis to Bill's comments... just love that he's preaching D again. That will keep us in any game.

Statistically Jesse, I'm interested how many minutes these starters are averaging vs. all other Self teams. I've never seen a bench play less for any team.

jayhawktalk 10 years, 8 months ago

I'd be curious to see the per minute bench productivity, including the negative stats - fouls and turnovers. I imagine it's abysmal, but would be curious to know whether that's just me being biased and pessimistic about our bench (so frustrating!), or if it's accurate.

5DecadeHawk 10 years, 8 months ago

Nice Job Jesse.

I was rather confused when Bill said he didn't have a way to do it.

All the data is already collected in each game's play by play log. It's just a matter of parsing the data and running the numbers.


I would suggest that end of half possessions should be included. In those instances the game clock effectively is the shot clock.

Also... it is extremely important to think about what is the proper way to factor in offensive rebounds.

For example: An opponent misses a shot clock buzzer beater, but our defense was not properly positioned when he shot it to get the rebound, and the opponent gets an offensive rebound and stickback.

In that situation, Bill Self might consider the poor rebounding positioning to be "not finishing a possession." Even though the stickback bucket after the rebound is officially recorded as being scored on a newly refreshed shot clock, it clearly is an incident of less than ideal defense at the end of the previous shot clock.

I would suggest that made shots are not the only important result of an opponent possession that are a sign of poor defense. Offensive Rebounds should be considered poor defense as well, or at least charted as a second category to be put in perspective.

While the above graphs are informative, they don't tell the whole story. It's important to plot all instances that end every possession, positively and negatively.

Plotting made/missed shots, offensive rebounds is helpful.

Plotting opponent turnovers and defensive fouls is also important.

Each category could easily have its own chart. Then there could be a comprehensive chart that combines them all in a logical way.

In any case. You've still done nice work Jesse.

Putting the game log into a spreadsheet, parsing the data, should make it easy to generate all the graphs and average out the numbers. Someone good at spreadsheets could probably plot every game for the entire season with an afternoon's work.

VancouverHawk 10 years, 8 months ago

I like the analysis as well. It would be nice to see some comparative information with other games - games where Self thought we played better late shot clock defense. Although it doesn't appear bad in this game, maybe the late shot clock defense has been even better in other recent games (e.g., the Baylor game).

Michael Luby 10 years, 8 months ago

Hey Manginwhore, Im sure you are such a genius and scholar at the game of basketball. You could probably teach Coach Knight or Coach K a few things right? Loser. I bet your just a janitor at a ballpark or something. Your not even a KU fan! Quit posting here and go be a D-bag somewhere else.

John Randall 10 years, 8 months ago

I'm sooo looking forward to your attempt at a "complete or useful" analysis of which ever situational stat coach comments on after each game.

Then I could be the one to type, "Come on."

Keith Conklin 10 years, 8 months ago

What really makes it seems like we have poor late-clock deffense is that when A&M made shots of that or we fouled it was at critical times and it made it seem more frustrating. If we wouldve Shot the ball better prob wouldnt have been as big of an issue. That game felt really sloppy and Coach Self said it best before half-time we played Dumb and We werent playing team ball. We did play better in the second half and Im happy with the outcome but if the play like that again against another team with better offense than we are in for a world of hurt. I know we have great D and I believe in this team. There are always things to improve on. Lets learn from this game and play Smart ball the rest of the yr. ROCK CHALK!

jhox 10 years, 8 months ago

What shocks me is that Self's perception was wrong. I've always been amazed at his ability to recall specifics after games. Honestly, I would have guessed just the opposite of what Coach Self was stating, because I thought our late clock defense was pretty good Monday night, but would have still taken Self's perception as the gospel if not proven otherwise.

John Randall 10 years, 8 months ago

It sounded more like "not 100% accurate" from where I was sitting.

StewardHawk 10 years, 8 months ago

Insightful stuff, Jesse.

I wonder about the apparent discrepancy with what HCBS believes about our late clock defense. How often did the clock run down under 10 seconds? I can recall just one clock violation, but it's possible I'm forgetting one or two. Perhaps what HCBS is upset about is that, when we did force their offense to work with less than 10 seconds, we either fouled or allowed them to score too often.

Do you have those numbers?

ccarp 10 years, 8 months ago

I'd like to see the shot clock violations on the chart too, or possessions over 35 seconds. Like Steward, I recall at least one. I'd like to know how many we get in a game and the season. I would love it if we could get a few more lock downs in a game. Go Hawks! Keep the intensity on the defense!

Ron Franklin 10 years, 8 months ago

He's always coaching. Always motivating. Don't believe his quote about not being able to track it. He's a genious, and he's been coaching and playing for over 25 years. He knows exactly what he is doing when he makes a statement like that. This is a group of players that needs to be called out regularly to keep them motivated.

Bill knows he's facing his first real BIG 12 road challenge this Saturday at the Hilton. He sent the boys out flat against Texas. He sent them out flat against A&M, and now he's pushing buttons and amping these guys up for Iowa St.

swavity 10 years, 8 months ago

This comment was removed by the site staff for violation of the usage agreement.

John Randall 10 years, 8 months ago

Certainly is the right time for 'amping up', with the toughest stretch of the conference schedule looming in front of the team.

Don Perry 10 years, 8 months ago

First point - Self's statement in the press conference is "people" - not necessarily focusing on TAMU game. There have been a number of instances in various games where the shot clock has moved under 15 seconds and a defensive breakdown has occurred. Playing defense for over half the shot clock and giving up a basket, offensive rebound and basket, or fouling ARE an issue for this team.

Second point - in a game where, as pointed out possessions are limited, and the final score ends within a spread of 3-4 possessions how big do the three scores with the clock under 15 seconds and a fourth just on the line, not to mention the fouls and offensive rebounds.

Good analysis and I think has it's merits. But, compare to the total possesions of the game and the impact those failed defensive possessions have on the final score. Close games dictate a more detailed focus on maintaining the defensive pressure throughout the possession.

Phoggie_Thinking 10 years, 8 months ago

Jesse, I am impressed that you would go back through the tape and give us video examples online. Good work!

It is interesting that Self's perception was not validated. I am amazed how often he states something and then you look back and he had it exactly right. I do think he is thinking season or at least the last few games when he says what he did and not merely A&M game.

Looking forward to seeing ISU game to see the adjustments in style. Hopefully it will lead to a victory, but it is sure to lead to a great contest.

Dyrk Dugan 10 years, 8 months ago

this is good analysis no doubt, but it all comes down to points a minute. if you avg. two points a minute, you'll score 80 points a game. if you average 1 1/2 points a minute, you 'll score 60.

anything less than 1 1/2 points a minute, is excellent defense......and that comes down to great half court D, ball handling on your end and good shot selection so you don't give up easy transition buckets.

of course, the game ebbs and flows, but for the most part Monday night, our D was very very good. we only gave up 24 points in the 2nd half. the problem, we couldn't hit the broadside of the barn from three, and we didn't take those medium shots that are wide open to us. Elijah, TRele and Tyshawn can all hit MUST start becoming more of a staple of our offense...because teams will start to pack it in more and more.

HCBS will always talk D....and lament how it could be better. he's a defensive minded coach for sure. but i think this team's D has been great this year...we just need to smooth the O a bit...and we'll be even more dangerous.

AsadZ 10 years, 8 months ago

Jesse, Thanks for the analysis and hard work. This is great.

On a different subject, KUSports used to post 3-4 minute highlights of KU games. I really enjoyed watching them. Would it be possible to start posting them again.

Jim Roth 10 years, 8 months ago

Looks like a big difference between the first half and second half. In the first half, 9 of 11 goals were after the 25 second mark, which generally supports Self's perception (using 10 sec instead of 7-8). In the second half, the field goals are more normally distributed.

It seems the trend was there first half, but adjustments were made for the second half. Jesse's analysis was based on the whole game, and thus didn't support Self's suspicion on the whole.

Sorry I'm late getting to this discussion. I really enjoy looking at the data.

panalytic 10 years, 8 months ago

I think this was a great idea to chart what Self was discussing as a problem for the team. I am not sure I'm reading the chart correctly though. I believe the box score has A&M taking 48 shots and I can't find 48 shots on the chart. Is the chart showing all their shots for the game? Is each mark on the chart one shot? Just wondering if I'm missing on how all their shots are accounted for??

Jesse Newell 10 years, 8 months ago

The chart is just made field goals. Texas A&M made 20. Nineteen show up on the chart, as I didn't include Harris' shot at the end of the first half with no shot clock.

kmatlage 10 years, 8 months ago

The transition baskets are skewing your findings. Self is talking about the possessions where we are getting down and actually playing defense. If you just exclude the shots with over 30 seconds left on the shot clock (though I would argue some of the ~28 second range shots should likely be excluded too, would have to look at the tape), the new numbers (include the 2 foul bail-outs you mentioned) are 5/16 for 31% instead of the 10% mentioned in the article. If you only count the shots at 25 seconds remaining or less (most of the ones above that we aren't really playing defense yet, though would need to look at the tape to tell), then it's 5/11 for 45%.

And that's at a point in the shot clock when normally people are putting up bad shots, but they were getting many good looks. It doesn't take into account missed shots where they got a good look. It also doesn't take into account our late clock percentage, which I think is likely more what's frustrating Self, even though it's not what he directly said.

When the shot clock is under 8-10 seconds, I believe we fouled, let them score, or let them get a good look more often than we were able to keep playing good defense. That, I would say, is more important to look at than what percentage of their total shots were made in that time frame. I believe this is more what gave Self the feeling he had during the postgame interview. I think it's unfair to look at the specific stat he happened to mention rather the the point he was trying to make, which was that our late-clock defense is bad. If we're letting them score almost every time we get to a late-clock situation, that is bad, regardless of if that ends up being the majority of their baskets or not.

That said, I don't believe we always have this problem, but I think we did on Monday.

Carter Patterson 10 years, 8 months ago

I love postscripts! Jesse, this along with the work you do in breaking out plays in football is just great stuff for the Jayhawk Nation. Appreciate all the work you guys do.

jaybate 10 years, 8 months ago

Huge props to Jesse for plotting the data.

Then huge props to kmatlage for parsing it just a bit further and revealing its truth.

Jesse had me convinced until I read kmatlage's super take.

I would like to add one more wrinkle to the work of both.

Notice how the under 10 second FGs and FTs are clustered in the first 5-8 minutes of each half; this is significant on two levels.

First, because they clump early in each half, they form Self's early impression of how the game went each half. First, or early, negative impressions often bias our outlook of what follows. They make us focus on the negative more than we might otherwise, during the rest of the game. With early negatives, we begin to view the game through the lens of what we are doing wrong, rather than through a positive lens. Early clumping of errors that particularly gall Self, make Self view the players' performances through a negative lens. This effect is probably greatly magnified, when the early first half mistakes, that stick in a coach's craw, then get heavily lectured on at half time, and then, instead of improving, the players make a clump of the same mistakes at the beginning of the second half. This two bad first impressions each half with the second bad first impression infuriating the coach, because he just lectured on it at half time.

Recalling Self's demeanor in the aTm game, he was agitated and angry for about as long as I have seen him in a game--for practically all but the last few minutes of the second half. At the end of the first half, the normally composed Self, could barely contain his anger at his team, while having to answer a few brief questions from Holly Rowe, something he usually has no trouble doing.

From all of the above, I infer late scoring is a major irritant to Self and experiencing it early in a half becomes something very hard for him to let go of the rest of the game.

Second, the second half clump appears almost consecutive failed possessions, or very close. They came much more suddenly than the first half and much closer together than in the first half. This at leasts suggest that Self's half time lecture may have actually given them such a negative outlook, something he often talks about trying to avoid, that they unconsciously got tight near the end of these early possessions and thought to much about making mistakes. This is tremendously frustrating to a coach, because on some level, that will probably not surface for a few days, he knows his coaching made the problem worse, not better.

Again, thanks to Jesse and kmatlage for amping up the QA of KU basketball here.

Rock Chalk!

John Randall 10 years, 8 months ago

Jesse, weren't all ten of the baskets with 30+ sec on the clock the result of offensive rebounds? When you only look at made baskets, everything gets skewed.

Jesse Newell 10 years, 8 months ago

Nope. As listed above, Texas A&M had only one second-chance basket. You can find it as the red dot on the graph above.

John Randall 10 years, 8 months ago

Point conceded – I expected there would be quite a few instances of possessions with a shot clock reset that made it seem like 'early' when it really was 'late'. The put-back is not the only way to get second-chance points, many come 8-10-12 seconds later, which may be as much as 45 seconds into the possession.

Commenting has been disabled for this item.