The Texas basketball season is not off to a roaring start. After five games, the Longhorns have a 2-3 record. This is more or less what you would expect for a team that has scored and allowed an identical number of points.
After a 2-3 start, Texas fans might reasonably ask the question: what isn't working for the Longhorns?
The tools of basketball analytics give us a variety of ways to identify where problems are coming for a team. And for Shaka Smart's team, the numbers up to this point tell us a lot.
An overview of the Texas offense and defense over the first five games
Through the first five games of the season the Texas offense and Texas defense has participated in 355 possessions. Texas has both scored and allowed 1.03 points per possession. (All data here are taken from hooplen.com.) More details for how Texas has performed at both ends of the floor are shown in the table below.
That table is helpful, but I sometimes like to look at these data in a different way. It turns out it is possible to calculate how much each statistical category contributes to offensive and defensive points per possession.
To summarize the data in this way, I like to compare a team to what a "typical" NCAA D-I team would do. A typical team scores and allows about 1.04 points per possession, and they do this by achieving particular measures in each of seven statistical categories: two point shooting percentage (2FG%), three point shooting percentage (3FG%), free throw shooting percentage (FT%), the percentage of field goal attempts that are taken from beyond the three point line (3FGA/FGA), the ratio of free throw attempts to field goal attempts (FTA/FGA), the percentage of missed shots that the offense rebounds (ORB%), and the percentage of possessions that end in a turnover (TO%).
I like to present this information using a waterfall chart. Below, I show the chart for the Texas offense. On the far left, there is a grey bar that represents the points per possession scored by a typical NCAA D-I team. Then each subsequent bar (either red or green) shows how each statistical category either increases or decreases this total. Increases are shown with green bars, and decreases with red bars (when evaluating an offense, green is good and red is bad). Finally, at the far right is the number of points per possession scored by the Texas offense.
(More details on these waterfall charts can be found by reading this article.)
A look at the chart above reveals that the Texas offense is substantially hurt by poor shooting from two point range, and is also hurt by poor free throw shooting. The strength of the Texas offense, as it has been for many years, is getting extra shots on the offensive glass.
We can make a similar plot for the Texas defense. In this case, red is good, and green is bad, as red reflects a reduction in scoring. This plot is shown below.
It should come as no surprise that Texas limits opponents shooting percentage inside the arc, as this is something that Texas has done well in recent seasons. It is largely attributable to the presence of shot blockers like Cameron Ridley, Prince Ibeh, and Connor Lammert on the Texas defensive interior.
It is also not much of a surprise that forcing turnovers has become one of the strengths of the Texas defense. This specific improvement is what Texas signed up for when it brought Shaka Smart in to replace Rick Barnes. (You can read more on this transformation here.)
But the defensive record hasn't been spotless. While doing well keeping down opponent scoring inside the arc and forcing turnovers, the Texas defense has been pretty poor at securing defensive rebounds, and has given up its share of three point shooting attempts, which opponents have taken good advantage of.
We can summarize all of this by stating that Texas' problems so far this year have largely come in four broad statistical categories, two on offense and two on defense. On offense, Texas has struggled to convert chances inside the arc, and has not shot well at the free throw line. On defense, the trouble has come on the glass and at the three point line.
Let's look at each of these four issues in greater detail.
Shooting from inside the arc
To anyone who has watched Texas play so far, the different approach to offense taken by first year coach Shaka Smart, in comparison to his predecessor, has been obvious.
Smart sets a variety of ball screens, with the primary goal of creating opportunities to attack the basket. And that is exactly what Texas has done. 45 percent of Texas' shots so far have been logged as layups and dunks, compared with 32 percent last season. (So far, the NCAA team median for this statistic is 37 percent. Last year, at season's end, it was 36 percent.) All of these drives to the basket also lead to chances to shoot free throws; only four teams in D-I have a higher ratio of free throw attempts to field goal attempts than the Texas Longhorns.
The Longhorns have been getting to the rim a lot, which is good, but have only converted on 50 percent of these attempts (The NCAA median for this number is 60 percent. Last year the Horns converted on 62 percent of their layups and dunks). Texas' problems close to the basket were the subject of this season's first Inside the Numbers post.
Only converting on 50 percent of the chances close to the basket is a problem. We can dig a little deeper to bring it into better focus.
The table below comes from the data available to subscribers at hoop-math.com. It breaks Texas chances into several categories, but I want to focus on the transition and non-transition attempts. (The rows labeled "25 Seconds into Possession" and "Putbacks" are simply subsets of the non-transition data.)
As a point of comparison, last season Texas converted on 64 percent of its transition attempts at the rim, and 62 percent in non-transition situations. Texas' substantially lower shooting percentage through the first five games of the season is heavily influenced by failure near the basket.
We can dig further into Texas' troubles. Through five games, about 13 percent of Texas' two point shot attempts have been blocked, which is worse than 288 D-I teams. This is at least in part a consequence of the Longhorns playing their first four games against good shot blocking teams (yes, even Texas A&M-CC). And two of those four games were against Washington, a team currently ranked 36th nationally in percentage of opponent twos that are blocked.
But the problems scoring at the rim are not uniformly distributed across the Texas roster. If we assume a typical shooting percentage on layups and dunks is 60 percent, then we can easily calculate how many "extra" shots close to the basket that players have made or missed.
Not surprisingly, Cameron Ridley leads Texas in extra makes near the rim. So far, the Longhorn center has made 81 percent of his layups and dunks, which means that he has made about 5 extra baskets at the rim, relative to what would be expected for a typical player with the same number of shot attempts.
No other Longhorn has done as well as Ridley -- in fact none are even remotely close.
At the bottom of the list we find Tevin Mack and Isaiah Taylor. Mack has missed about 5 more shots than would be expected, which is due to the fact that he is 0-9 on shots at the rim so far this season. Taylor also has 5 extra misses, but his percentage is far better than Mack's, at 46 percent; Taylor has only missed this many shots because he has gotten to the hoop so much, taking 61 percent of his attempts at the rim (that is good -- Taylor has ten more attempts at the rim than anyone else on the roster). Both Shaquille Cleare and Kerwin Roach have about two and a half extra misses than they would if converting 60 percent near the basket.
What is interesting when we dig further is just when Taylor and Mack have missed. Somewhat oddly, Taylor's misses have mostly come in transition. He is just 1-7 at the rim in the first ten seconds of live ball possessions. He is actually shooting a much higher percentage on shots near the rim later in possessions, connecting on 54 percent of his 28 attempts in non-transition situations. This fact makes me think Taylor's troubles in transition are almost surely a sample size fluke.
For Tevin Mack, the problems are more as you would expect. Mack is 0-8 in shots near the rim in half court situations.
Of course, it has been more than just the shots at the rim that are hurting Texas. 22 percent of the Longhorns shots have been logged as two point jump shots, and Texas has only hit on 24 percent of these attempts. The D-I team median in this measure is 36 percent.
Free throw shooting
Texas' rim attacking approach on offense is earning visits to the free throw line a lot. The Longhorns are shooting 58 free throws for every 100 field goal attempts, which is the fifth highest rate in the nation.
That is good, but the benefits of this are diminished by the fact that the Longhorns are connecting on fewer than 63 percent of their free throws so far this year, which is a lower percentage than 296 other D-I teams.
Cameron Ridley is 13-30 at the line. Last season the big guy connected on 61 percent of his free throws, while the year before he made 63 percent. Given this, we know he can shoot them better than he has so far this year.
Only Taylor, Mack, and Javan Felix have connected on greater than 70 percent of their free throws so far this season.
Free throw shooting was actually a strength for Texas last season, when the Longhorns hit 72 percent from the line. But Taylor's shooting percentage is down so far (from 84 percent to 75 percent), as is Ridley's. And Myles Turner (84 percent) and Jonathan Holmes (78 percent) aren't on the team any more.
Taylor and Ridley are the two Texas players who shoot a lot of free throws. The Texas team shooting percentage will be greatly influenced by how well these two players do.
Texas' early struggles on the defensive glass are a bit of a surprise to me. Given recent history, and the number of good upperclassmen big men on the team, poor defensive rebounding is not something that I would have anticipated.
There is something of a strength of schedule effect working here. Washington is a very strong offensive rebounding team, as is Texas A&M. Even Texas A&M-CC is above average on the offensive glass. The Michigan Wolverines are the only poor offensive rebounding team so far on Texas' schedule. (And this is largely by choice; Michigan more or less ignores the offensive glass, focusing on getting back on defense.)
I should also point out that through the first five games Connor Lammert hasn't really been pulling his weight on the defensive glass. This is disappointing, as Lammert has been a steady rebounder for Texas over his career. It is probably just a blip, but Lammert's rebounding percentage is currently the lowest it has been in his college career. Last season he grabbed 20 percent of opponent misses while in the game; this year is at 12 percent.
On the positive end, some props need to go out to Kerwin Roach, who has come down with 19 percent of opponent misses while on the floor. Those are good numbers for a 6-9 forward, and are just exceptional for a skinny freshman guard. Guards who chase rebounds down can do a lot to boost team rebounding.
Three point defense
Three point shooting defense is a tricky thing to analyze. This is in large part because over small samples any potential information derived from opponent three point shooting percentages are overwhelmed by random effects. (For anyone who hasn't read it, Ken Pomeroy's writing on this topic is essential.)
So I am not too worried that after five games Texas opponents are shooting 36 percent from three point range. This is well within normal expectations for a team that has played Texas' schedule.
Perhaps more significant is the fact that Texas opponents are shooting a fairly large number of threes. 39 percent of opponent shots are coming from beyond the arc, which is a higher percentage than 279 D-I teams.
This high rate of three point attempts, when combined with the differences in value between twos and threes against Texas, is costing the Longhorns on defense.
There are a number of different ways to construct a man to man defense. In general, you can build it to take away certain things of your choosing, but it is rare to have a defense that takes away everything. For teams that pressure passing lanes the way that Texas does, they often find that they face a greater degree of opponent dribble penetration. This either leads to more chances at the rim, or more opportunities to kick the ball out for three point attempts.
So far this season, Texas hasn't given up a particularly high number of chances at the rim, but it has given up a lot of three point attempts. Most of Smart's VCU teams skewed the other way, giving up a greater number of layups, but limiting opponent attempts from the line. This is something that I will have to watch more closely as the season progresses. I haven't specifically watched a game yet with the goal of understanding what Texas' perimeter off-ball defenders are doing: are they helping far off their men to limit penetration, exposing them to threes, or are they playing close to their men, giving up penetration but closing off catch and shoot chances? Based on the numbers so far, I would guess that they are helping more than teams Smart has previously coached.
But it is still early
Arguably, spending so much effort working through the numbers of five games looking for broader trends is a waste of time, as there are so many more games ahead, and every tentative conclusion here could turn out to be worthless three weeks from now.
But while these four issues here may or may not continue through the rest of the season, they do accurately reflect the problems Texas has had through the first five games. This is rather different than saying these are permanent and immutable laws of nature; some of this will surely even out as Texas plays more games.