The last word on defense: A comprehensive title-game study
Nagging question: Did UConn play phenomenal defense or was Butler just awful?
If you refuse to credit UConn after seeing these numbers, you're being stubborn
Butler point guard Shelvin Mack: "They were able to contest every shot we shot"
Ten days have elapsed since a forgettable national championship game, providing ample time to move on to offseason matters, but I've been nagged by a question: Did UConn actually play phenomenal defense in holding Butler to 41 points on 18.8 percent shooting, or was Butler just awful?
There has been plenty of disagreement about this in the media. I chose to praise the D in the immediate aftermath, while there were prominent voices that insisted there was nothing great about the way the Huskies guarded, and that the only reasonable assessment was that the game was an abomination.
Butler-UConn was painful to watch the first time, courtside in Houston, and not particularly enjoyable to watch a second and third time on tape this week. But I reviewed it to find an answer to that nagging question, primarily by compiling comprehensive data on the difficulty of the shots the Bulldogs took. For context, I charted every Bulldogs scoring opportunity in three NCAA tournament games: the loss to the Huskies, the Final Four win over VCU in the same dome-affected setting and the second-round win over Pittsburgh, in which Butler thrived against another strong Big East defense. Each scoring opportunity was labeled in one of six ways:
1. Uncontested shot*
2. Contested shot*
3. Blocked/altered shot*
4. Forced turnover
5. Unforced turnover
(* To make this work, a contested shot was defined as one that was challenged in a way that might bother the shooter -- flying at him from a close distance with a hand raised and getting in his line of vision, or reasonably attempting to make a play on the ball. A shot was considered uncontested if a defender closed out too late to get in the shooter's sight line, or only stood in the vicinity of the shooter with a hand casually raised. On an attempt in the post, if the defender merely "walled up" [stood still with arms up], it was considered uncontested. A shot was logged as altered if the shooter had to pump, adjust in mid-air, or shoot at an unnatural angle.)
If you refuse to credit the Huskies' D after seeing these numbers, you're just being stubborn. Hard evidence emerges even in the first few minutes, when UConn sets a tone that makes Butler tentative on its shot attempts and reluctant (or unable) to attack the rim for the rest of the game.
In the other two games I studied, the Bulldogs were allowed to get great, early looks. In the win over Pitt, they scored 17 points on their first 10 possessions. None of their nine shots was blocked or altered, and they had five uncontested looks (one dunk, one lob, one short jumper and two threes). Against VCU, they only scored seven points on their first 10 possessions, but that was their own fault; seven of their first nine shots were uncontested.
The opening stage against the Huskies was distinctly different: Butler scored eight points on its first 10 possessions, during which there were 13 shots. Four of these shots (including the first two) were blocked or altered, and five more were contested. Star guard Shelvin Mack, who took the first shot of the game, a three-pointer that was blocked by long-armed freshman Jeremy Lamb, said immediately after the game, "They would contest shots that people normally wouldn't be able to contest." Later in the postgame press conference, Mack reiterated that point: "They were able to contest every shot we shot."
If you think Mack is embellishing, examine this montage of Butler shots Nos. 1, 2, 4 and 7, which include two blocks, an altered floater and a hotly contested three. Matt Howard, who's being stuffed by freshman Roscoe Smith in the bottom-right frame, went 0-of-7 from inside the arc and said, "When you're not scoring inside, especially when you're getting good looks, it can be real tough to get going."
UConn made trouble for the Bulldogs the whole game, blocking or altering an absurd 26.6 percent of the Bulldogs' shots, compared to just 3.8 percent by Pitt and 12.1 percent by VCU. In total, UConn blocked, altered or contested 60.9 percent of Butler's shots, forced four turnovers, and only had five fouls that resulted in points. A full comparison of the splits from the three games is below, in table and pie-chart form:
|Butler's Scoring Opportunities in NCAA tournament|
Butler was also atrocious in its 25 uncontested shot opportunities, making just 28.0 percent of them -- a significant drop from the other two games I charted:
|Butler's Shooting Percentage Splits in NCAA tournament|
Bulldogs coach Brad Stevens reasoned that there was a snowball effect after so many early shots were contested, saying, "I think what happens in a game like that is they guard you so well, when you start to get a few open ones, you're not feeling comfortable." But it may have been the type of uncontested shots his team was getting that mattered even more.
To truly appreciate UConn's defensive performance, don't just look at the 60.9 percent blocked-altered-contested figure. Also be aware that Smith, Alex Oriakhi and Charles Okwandu locked down the interior, allowing just five uncontested shots in or around the lane. The other 20 uncontested shots Butler had were either long twos or threes. The splits from the VCU and Pitt games were significantly different:
|Butler's Uncontested Shot Distribution in NCAA tournament|
UConn did this by design. It had focused on shutting down the paint all season, ranking seventh nationally in two-point field-goal defense, at 42.4 percent. And while their Big East counterpart Pittsburgh gave Butler a number of interior buckets off ball screens -- which the Bulldogs almost always run after their initial set -- the Huskies did no such thing. Their coaches thought they might need to "red," or double Mack off those screens, but once they realized Lamb could contain him after a brief hedge, they called off the double. The effect of this, according to associate head coach George Blaney, was that UConn remained in great position to guard the other four Butler players, and prevented Mack from ever passing the ball ahead to an open man.
The Bulldogs' only option, in many cases, was to launch jumpers, and Mack admitted as much, saying that they "started relying on the three [and] might [have] panic[ked] a little bit." In the game, 51.6 percent (33 of 64) of their shot attempts were from beyond the arc; on the season, just 38.7 percent of their attempts were threes. To make matters worse, UConn's ball pressure and hard closeouts caused Butler to take a number of deep threes; the tape shows Mack and Chase Stigall in particular taking many shots from 1-2 full steps off the line.
In the Huskies' scouting report, two of their keys to the game were:
"Need to force them to start offense farther out"
"Push them off the '3' point line"
The scouting report on Butler also said to "stay in the entire possession -- they run offense fast and hard and will not quit on a possession." Stevens has long preached the value of every possession, and his team had executed well in clutch, late shot-clock situations earlier in the tournament. Against UConn, this did not happen.
The Bulldogs took six shots with 10 seconds or less on the shot clock in the title game. Two were blocked, one was altered, and the other three were contested. Only one went in -- a Stigall three from two-and-a-half steps behind the line, with Jamal Coombs-McDaniel flying at him.
The Huskies didn't give up on a single late-clock situation, and not once did they bail out Butler with a foul. That takes discipline.
No one -- myself included -- wants another final-Monday performance like the one we witnessed last week. But it is worth noting that in a game that was lambasted as the culmination of the AAUization of college basketball, it was the savvy veterans who were launching the bad, deep threes because they couldn't get better looks. The reason they couldn't get good looks was because three freshmen, just a year removed from that awful AAU scene, were doing the lion's share of locking them down.
Lamb's ball-screen defense was exceptional, and on top of that, my chart has him down for one block, one steal and four contested shots. Backup point guard Shabazz Napier, whom UConn coach Jim Calhoun called "one of the greatest defenders I've ever seen for a freshman," was responsible for the pressure on Butler's point guards that disrupted the offensive flow. Napier also contested four shots and forced one turnover. Smith was perhaps the best of all: I credited him with four blocks, six contested shots and one forced turnover; he did a better job of harassing Howard than anyone else in the tournament. "Our defense," Smith said, "was unreal."
Butler went in the record books as the worst-shooting team in the history of the NCAA tournament title game, breaking a 70-year-old mark. The poor sightlines of Reliant Stadium were no doubt a factor in that horrific shooting percentage of 18.8, as was the fact that the Bulldogs were frigid on the big stage. But the biggest factor was the pressure UConn applied, and the type of shots it forced Butler to take. To not recognize the defense for what it was, was to miss the game altogether.