Bad News For Tennessee

[lidderer]

In my writings about who could win the national title this season, I have consciously avoided discussing Tennessee. This isn't really a case of not having something nice to say about them. I love what Bruce Pearl has done in Knoxville. Tennessee doesn't exactly have a foundation of tradition in men's hoops, yet soon after his arrival, Pearl raised the Vols to a level where they no longer have to wonder whether they can get into the NCAA Tournament, but rather how high their seed will be. This season, UT is clearly the class of the SEC; the distance between them and whoever is the second-best team in the league is enormous.

Obviously, I can't ignore Tennessee any longer. This is a huge week for Bruce Pearl's team. Duke's loss to Wake Forest has allowed Tennessee to assume the #2 spot in the major opinion polls, and wins over Auburn tonight and at Memphis on Saturday would move them into the #1 spot, probably unanimously. Even if that happens, the Volunteers have a glaring problem that would make them easily the most vulnerable one-seed, should they get one.

There's a glitch in Tennessee's famed pressure defense that will cause them problems against strong competition--the kind of competition that is absent in the SEC this season. The Vols' defense is heavily dependent on forcing turnovers. To put it more accurately, they rely on turnovers happening. That's a key distinction, because while every turnover has its own characteristics, ball security is more under the control of the offense than the defense.

To illustrate this, we can isolate the effect a defense has on its opponents' tendency to commit turnovers as well as the effect a team's offense has on its opponents ability to force turnovers. I'll use the first game of the season to demonstrate how this works. Maine has committed turnovers on 20.4% of its possessions this season. Against Richmond, they lost the ball on 28.0% of their possessions. Thus, Richmond's influence on Maine's turnover rate was +7.6, because Maine committed turnovers on 7.6% more of its possessions than normal. Similarly, Richmond's defense normally forces turnovers on 25.4% of its possessions, so Maine's affect on Richmond's ability to force turnovers is -2.6 for the game.

If you're still with me, this method applied to one game doesn't tell us much. However, if we do this for every D-I game and average the results for each team and we end up with a good idea of the offense's tendency to commit turnovers compared to the defense's ability to force them. Having done that for all games through Monday, here are the teams that are the most reluctant to commit turnovers:

1 Texas -9.8
2 New Orleans -9.6
3 West Virginia -9.1
4 George Mason -7.7
5 Northwestern -7.0

Texas' figure of -9.8 means that on average, their turnover percentage is 9.8% less than what the opponents' average turnover percentage would suggest. Now here are the teams that have shown the greatest ability to force turnovers:

1 Texas San Antonio +6.5
2 Tennessee +5.8
3 UC Santa Barbara +5.2
4 Nicholls St. +5.0
5 Army +5.0

No surprise that Tennessee is near the top. If we considered the strength of their opponents, I'm sure they would rate as the premier team in the nation at forcing turnovers. There's another interesting point in these two lists, though. The best teams in terms of taking care of the ball are much better than the best defenses in terms of forcing turnovers. I'm only showing the top five, but this is true throughout the rankings. The 50th-best team in forcing turnovers is not as effective as the 50th-best team in committing them, for example. This is a problem for Tennessee because it relies on opponents not having the ability to control the basketball. The correlation between its defensive turnover percentage and defensive efficiency is -0.72, stronger than any other team in the top 20 of the Pomeroy Ratings except Clemson. This means that the Volunteers depend on forcing turnovers for their defense to be good. As they play better competition, their ability to do that diminishes.

Turnovers have occurred on 26.4% of Tennessee's defensive possessions this season, which is the sixth-highest rate in the country. In SEC play, turnovers have occurred on 23.9% of their possessions, a rate which leads second-best South Carolina by more than 2%. What happens if the Volunteers' turnover rate drops? We can get a sense for how Tennessee will fare in that case based on its games so far in which few turnovers have occurred.

Here are the five games in which UT opponents have managed to cough up the ball fewer than one in five possessions:

Opponent TO% Def. Eff.
1/9 Ole Miss 14.4 119.7
11/24 Texas 17.7 132.4
12/15 W. Kentucky 18.4 107.6
1/22 Kentucky 19.3 115.9
2/16 Georgia 19.9 100.8

On average, Tennessee has surrendered 95.6 points per 100 possessions this season; in their five low-turnover games they've allowed an average of 115.3. Aside from Texas, this group doesn't contain a bunch of offensive powerhouses. This isn't a trend that started this season, either. Last season, Tennessee was involved in 10 games in which its opponent kept its TO% below 20, including all three of its NCAA Tournament games. Opponents averaged 111.1 points per 100 possessions in those games compared to 98.6 in all other games last season. In those 15 low-turnover games combined, only once did an opponent not score a point per possession, and that was Memphis last December. In that game, Memphis went a mind-boggling 17 of 49 on two-pointers, which is a deviation from what Tennessee opponents normally shoot. That's because when Tennessee isn't getting turnovers, they're yielding too many run-outs. Opponents are shooting 51.6% inside the arc this season, which ranks 280th in the nation. That figure has risen a tick to 52.0% in SEC play.

Tennessee's press is able to consistently overwhelm its conference mates who in general are more prone to turnovers than the teams they will be playing in the NCAA Tournament. Thus Tennessee's defense is able to enjoy enough empty possessions to make up for the extremely high efficiency its opponents have when they avoid losing the ball. However, that won't occur consistently when Tennessee faces better opponents.

This isn't to say that Tennessee can't beat Memphis, or some other elite team, without having it cough up the ball a bunch. After all, the Volunteers beat its in-state rivals easily last season under those circumstances. Keep in mind, though, that was only time in the last two seasons that Tennessee has had a good defensive game without forcing many turnovers. If Tennessee can play effective defense without forcing turnovers only about once every two years, their existence in March is going to be a lot shorter than their seeding would suggest.

Basketball Prospectus | Articles | Bad News for Tennessee

 

[rexvol]

classic case of overthinking a subject

 

[NEW COACH]

^^^^yup

 

[crazyguy]

an interesting way of creating some fuzzy math.

 

[lidderer]

classic case of overthinking a subject

Classic case of offering no alternative position and grandstanding behind empty platitudes devoid of any backing.

This seems a legit point of arguement vs our Vols' chances at longevity come March. You think otherwise, and why?

 

[HawaiiVol]

so he ran a bunch of numbers to tell us what we already knew. UT's halfcourt D leaves something to be desired.

Shocking.

 

[lidderer]

If you're weirdly/bizarrely intimidated or daunted by the quote fuzzy math unquote, I mean you could just probably figure out that his point is essentially: Hey Vols, in noting your horrid fg-defense, and further noting your reliance on TOs, and given that upper-tier teams tend to fare well in both categories, it seems like you're probably pretty poorly equipped to handle the rigours and quality of march opponents, at least compared to the KU's and UCLA's and Duke's of the world.

I'm not even sure what there is to argue about with that conclusion. Anyone?

 

[onevolfan]

is that GA in your avatar?

 

[lidderer]

so he ran a bunch of numbers to tell us what we already knew. UT's halfcourt D leaves something to be desired.

Shocking.

You'd be surprised how many inane assumptions we make regarding sports and how numbers often prove them wrong; there's something to be said for taking the time and effort to verify and better pinpoint what precisely it is that is 'to be desired', and if in fact it is 'desirable'.

 

[crazyguy]

If you're weirdly/bizarrely intimidated or daunted by the quote fuzzy math unquote, I mean you could just probably figure out that his point is essentially: Hey Vols, in noting your horrid fg-defense, and further noting your reliance on TOs, and given that upper-tier teams tend to fare well in both categories, it seems like you're probably pretty poorly equipped to handle the rigours and quality of march opponents, at least compared to the KU's and UCLA's and Duke's of the world.

I'm not even sure what there is to argue about with that conclusion. Anyone?

I'm not arguing with the conclusion. I'm arguing with the math that doesn't include games against the #2 TO producing team in half of the statistical analysis.

 

[utvolpj]

If you're weirdly/bizarrely intimidated or daunted by the quote fuzzy math unquote, I mean you could just probably figure out that his point is essentially: Hey Vols, in noting your horrid fg-defense, and further noting your reliance on TOs, and given that upper-tier teams tend to fare well in both categories, it seems like you're probably pretty poorly equipped to handle the rigours and quality of march opponents, at least compared to the KU's and UCLA's and Duke's of the world.

I'm not even sure what there is to argue about with that conclusion. Anyone?

To face the "elite" teams you mention would mean an elite8/final4 appearance. To say we will make an early exit kinda blows that one away right? I highly doubt we face elite teams in the 1st round

There are lies, damn lies and then there are statistics.

 

[HawaiiVol]

You'd be surprised how many inane assumptions we make regarding sports and how numbers often prove them wrong; there's something to be said for taking the time and effort to verify and better pinpoint what precisely it is that is 'to be desired', and if in fact it is 'desirable'.

you don't have to make assumptions if you watch the games. It's right there in front of you.

And no, I wouldn't be surprised at the number of inane assumptions at all.

 

[VolColts03]

is that GA in your avatar?

or possibly Broncos vs Browns?

 

[VolfanMike]

is that GA in your avatar?

Looks like the browns and broncos to me.

 

[lidderer]

heartbreaking byner fumble, it is.

 

[vol731]

now I love stats and hearing all the crazy stuff like this that comes out of stat software, but seriously? like the rest of you have said, we already knew this.

 

[lidderer]

you don't have to make assumptions if you watch the games. It's right there in front of you.

And no, I wouldn't be surprised at the number of inane assumptions at all.

Part of me agrees with what you're saying; but another part of me knows that's not entirely true. I mean unless you're some sort of savant it's basically impossible to fully tally all the various events that occur throughout a season and sift through them at the drop of a hat in order to arrive at accurate and thoroughly founded opinions about how/why a team works or doesn't.

Stats are basically an easily accessible memory that allows us to see correlations and trends and etc, and to ignore them because you 'watch the game' is to make a severe error: it is to think of numbers as somewhat separate from the game itself; it is to think of stats as an alternate story from the one you witness. This is not the case: stats are annotation to the game itself, are addition to what our eyes do see but our mind's can in no way comprehend so thoroughly.

Of course, they leave some things out, definitely, but so does simply viewing. This might be kinda why both watching the game and the cumulative boxscores is a good idea, rather than pitting one v. another.

 

[phi0129]

We have been addressing these issues all year long. And just because statistically something is suppossed to happen doesn't mean it will. Maybe we will have perfect circumstances every game of the tournament? My point is that while this article is insightful and fairly accurate from a mathematical standpoint, there are far too many other variables affecting the outcome of the game to think this statistic completely dooms us. Maybe we will catch those elite teams on a night when they are turning the ball over. Just because someone's average turnover ratio is low doesn't mean that some games they don't turn it over more. Not to mention nowhere in that article did he give us the elite teams results of his model (Memphis, Kansas, UCLA, Duke, etc.). While a good read for someone like me (math junkie, MBA in finance), I don't put a whole lot of weight on it. I agree that our chances are diminished when we can't force turnovers, but maybe we will shoot lights out that night. All I know is we have risen to the occasion almost every game this year against stiff competition, and I think that gives us just as good a chance as any to make the Final Four. It doesn't mean I think we are going to at this exact moment.

 

[phi0129]

Earlier in the year they were talking about teams we couldn't beat because of our lack of inside presence. Well so far we have disspelled those ideas as well...

 

[lidderer]

now I love stats and hearing all the crazy stuff like this that comes out of stat software, but seriously? like the rest of you have said, we already knew this.

Well I mean this is a 'topic', right? It's not like it's just some authoritative post that you can't quibble with.

perhaps what's interesting about these numbers is that, while I agree with the premise, eg. we rely far too much on turnovers, I think there is an element that the numbers fail to account for, and it might be worth exploring just what that element is.

Meaning that I actually kinda think the Vols could force, say, only 12 TOs vs Memphis and still win, and yet I wonder what kind of hunches sponsor that thought. I'm reluctant to call it 'intagibles', because that's a bit silly-seeming, but I'm not reluctant to call it something else. What that else is I'm not aware of yet.

Let the topic be about that instead, if you are tired of hearing the same old.

 

[lidderer]

We have been addressing these issues all year long. And just because statistically something is suppossed to happen doesn't mean it will. Maybe we will have perfect circumstances every game of the tournament? My point is that while this article is insightful and fairly accurate from a mathematical standpoint, there are far too many other variables affecting the outcome of the game to think this statistic completely dooms us.

To be fair, phil, the article merely suggests that this exact metric indicates a higher possibilty of a vols' letdown; it does not say we are doomed. Numbers are not as arrogant as most think. They are often just a means to predict or suggest what is MORE likely to happen.

 

[Lexvol]

Well I mean this is a 'topic', right? It's not like it's just some authoritative post that you can't quibble with.

perhaps what's interesting about these numbers is that, while I agree with the premise, eg. we rely far too much on turnovers, I think there is an element that the numbers fail to account for, and it might be worth exploring just what that element is.

Meaning that I actually kinda think the Vols could force, say, only 12 TOs vs Memphis and still win, and yet I wonder what kind of hunches sponsor that thought. I'm reluctant to call it 'intagibles', because that's a bit silly-seeming, but I'm not reluctant to call it something else. What that else is I'm not aware of yet.

Let the topic be about that instead, if you are tired of hearing the same old.

Generally speaking turnovers and transition baskets ease the reliance improve your field goal %. However, three point shooting is not an intangible. If the Vols get hot and hit their trey's they can play with anyone in the country.

 

[phi0129]

To be fair, phil, the article merely suggests that this exact metric indicates a higher possibilty of a vols' letdown; it does not say we are doomed. Numbers are not as arrogant as most think. They are often just a means to predict or suggest what is MORE likely to happen.

I know about numbers my friend. I got a 780 on the SAT in math, perfect on the ACT, a 5 on my AP Calc Exam, and I majored in computer sciences during my undergrad years. It's not phil. It's phi. Just to let you know for future reference. On my online poker account everyone calls me phil. What you said above is my point. It seemed to me that every time someone came back with something to say about that article, you defended it to the point where you thought the sky was falling. I must have misinterpreted.

 

[HawaiiVol]

Part of me agrees with what you're saying; but another part of me knows that's not entirely true. I mean unless you're some sort of savant it's basically impossible to fully tally all the various events that occur throughout a season and sift through them at the drop of a hat in order to arrive at accurate and thoroughly founded opinions about how/why a team works or doesn't.

Stats are basically an easily accessible memory that allows us to see correlations and trends and etc, and to ignore them because you 'watch the game' is to make a severe error: it is to think of numbers as somewhat separate from the game itself; it is to think of stats as an alternate story from the one you witness. This is not the case: stats are annotation to the game itself, are addition to what our eyes do see but our mind's can in no way comprehend so thoroughly.

Of course, they leave some things out, definitely, but so does simply viewing. This might be kinda why both watching the game and the cumulative boxscores is a good idea, rather than pitting one v. another.

2 different routes came to the same conclusion. I'm no savant, as I know the majority that post here aren't, but it's pretty obvious to come to the same result.

It's nice that there are guys like that, that sift through every minute detail to validate what the majority ascertain from simply watching and paying attention.

 

[rexvol]

Classic case of offering no alternative position and grandstanding behind empty platitudes devoid of any backing.

This seems a legit point of arguement vs our Vols' chances at longevity come March. You think otherwise, and why?

Obviously, I can't ignore Tennessee any longer.

I think the above statement speaks volumes about the author's motivation. If you set out to prove a certain point you can skew the numbers to say what you want. As for you, well, you seem to be the person grandstanding. To refute him I merely need to point out that despite all of his numbers, we are #2 in the nation.