One thing stands out from the soul-crushing loss to Duke in February, and it's their three-point shooting. The Blue Devils made 14 of 36 three-pointers, capping the night with Austin Rivers' game-winner that still sits like a pit in my stomach. UNC's perimeter defense has always drawn its share of criticism, and none more than after the game, which produced some truly horrible analysis:
That loss UNC suffered to Duke last night was utterly inexcusable. The Blue Devils couldn't defend my 51-year old sister, let alone somebody like Harrison Barnes or Tyler Zeller, yet because the Tar Heels would not guard the Duke 3-point gunners (which, like Florida, is practically the only way the Blue Devils can score), they snatched defeat from the jaws of victory and let Austin Rivers win the game on a last-second bomb.
You can tell by reading this that the Tar Heel fans just do not get it. Their defense is poor, or at the very least, it was in this game (and that's giving galactically-generous benefit of the doubt). That's why they lost -- they let Duke do what Duke does best. #5 ranked teams don't let rivals come on their home court and drop 14 3-pointers in their eye. If it happens in Cameron, well, that's one thing, but in the Dean Dome? UNC is far too good to have lost to the mediocre Blue Devils at home.
If I were a Tar Heel fan right now, I would be rhetorically beating Roy Williams' brains out. He has taught these players to fire up whatever crap they can get after 15 seconds in order to get the pace up, and judging from what I saw last night, the Tar Heels can't win if the other team has multiple threats from three -- they just refuse to come out and guard them for more than a few seconds. Nobody makes 14 closely-guarded threes in a 40-minute game -- if Carolina had just forced Duke to drive to the basket and make layups over John Henson and Zeller, they would have won by 20. It's called "strategy."
Yeah, I'll actually be rooting for Duke if they meet Kentucky in the tournament. My love of schadenfreude outpaces my hatred of Krzyzewski & Co., it turns out. But North Carolina under Roy Williams has always forced defenses to take more threes, often by virtues of the tremendous size they pack in the paint. In four of Williams' nine seasons the Heels have been 250th or worse in percentage of shots from their opponents being behind the arc; this season 36.3% of opponent's shots have been threes, second only to the 2005 championship team, who forced 36.7%.
The strategy actually goes back to the Dean Smith era. I first noticed it, oddly enough, in the 1995 Elite Eight game against Rick Pitino's Kentucky team. UNC, two seed to UK's one-seed, was given absolutely no chance coming into that game – Pitino's high-octane, sharp-shooting Wildcats were going to destroy the Heels. Kentucky launched 36 threes that night; they made seven. Meanwhile, Carolina's big men scoop up rebounds, while on offense they worked the ball inside, drawing fouls and making baskets. In the end, it wasn't even close – UNC won 74-61.
I chose my words carefully when I said "forced threes" a few paragraphs up. Roy Williams' teams much prefer their opponents to take a low percentage shot. He wants them defending the perimeter shooters, sure, but not at the expense of the dribble-drive. This, of course, is the exact opposite of Krzyzewski's offensive philosophy, which is to spread the defense by loading up on perimeter shooters, and then open up penetration lanes. It's no surprise Duke has one of the lowest assist-to-field goals made ratios in the country.
So Duke likes to shoot threes and opponents who overcommit to defending them. UNC is happy to oblige the former and won't be baited into the latter. Is this a defensive mistake on the Heels' part? Well it turns out Ken Pomeroy has spent most of February ruminating about college basketball defense in a series of blog posts. He began with a simple observation – the three-point percentage opponents shoot against a team in the first half of the season has no correlation to the percentage a team's opponent shoots in the second half of the season. Basically, a defense has no control over their opponent's shooting percentage behind the arc.
Now, of course, this isn't literally true. If Kendall Marshall and company just milled around the arc, hands down, listless, Seth Curry and Austin Rivers would have record-setting afternoons. But that's not going to happen. In actual basketball games, players shoot when they think they can make the shot, and the defense they face will vary on a given play between Duke levels of suffocation, and N.C. State's "let Kendall Marshall try something" unconcern. And over time, those types of opportunities will even out, and when the numbers are run, on average the defense is irrelevant.
This isn't the case with two-point shooting percentage. Teams that are good at defending inside the arc in the first half of the season generally remain so the rest of the way. And the percentage of three-point shots an opponent attempts is highly defensive-dependent. But a defense has no effect on their opponent's three-point shooting percentage. But what Pomeroy found that was really surprising is that apparently, teams in general have no influence on their own three-point shooting ability. Basically, there's no correlation between how well a team shoots the three in the first half of conference play and the second; it's all noise.
Now to be fair, this applies to college basketball as a whole. UNC isn't playing the whole of college basketball on Saturday; they're playing Duke. And Duke is 19th in the nation in three-point percentage, and it's not by virtue of being a random outlier. At 39% from behind the arc, the Blue Devils have more three-point shooting talent than most. But of their 15 conference games, six of them (40%) have found them shooting below .300 from three. And these include games at Georgia Tech and against N.C. State, some the more lax perimeter defenses around. (Yes, in the tongue picture above, Rivers is celebrating one of only two of his eight three-point attempts to find the bottom of the net.) The Blue Devils might have slightly higher odds of hitting a three than your average team, but there's still a huge amount of variation between how things will go on a given night. Pomeroy calls it playing the lottery; Krzyzewski's going all in on scratch-off tickets, while Roy is staying away from the 7-11 counter.
So if your cursing the thought of another evening of Duke raining down the threes, don't. There's smart and dumb things the Carolina perimeter defenders can do on any given trip down the court, of course, and I'm sure it will draw a bit of frustration. But overplaying the three is exactly what Duke is hoping for, and precisely the game that after nine years, Roy Williams isn't willing to play. And in the end, there's only so much a defense can do out at the arc. How do I know? Because Kentucky, with Calipari's desire to challenge every perimeter shot, and allow nothing easy, has opponents shooting 32.4% from beyond the arc. Carolina, plagued by perimeter-bombing opponents like Duke and FSU? They allow 31.4%.