In a new column which ran Friday, Sports Illustrated's Luke Winn made the statement that the Kansas State-Xavier game was the best of the NCAA Tournament. This isn't that controversial. It was a game that featured two overtimes, 13 ties, 17 lead changes and 197 points scored. It was an instant classic to any who watched it.
But, he writes, there is a way to also statistically say that it was the best game of the tournament. Ken Pomeroy—the numbers guru who saw Duke's offensive dominance before anyone else in the country—has created something called Win Probability figures (WP for short) which chart how teams' odds of victory changed on each possession of a game. Kansas State-Xavier shows huge swings in the probability graph; therefore, it was technically a great game.
Naturally, after reading this, I had to take a look at the best games (and by "best" I mean most fun to watch) that Duke played last year. Surprise! It ain't the one against Butler. Look at its graph: the WP for Duke rarely dipped below 70 percent, and its lowest number, which came after Kyle Singler missed a jumper with 38 seconds left, was only 64 percent. In other words, the Blue Devils victory was statistically never really in doubt.
It also wasn't Duke's hard-fought 78-71 win over Baylor March 28. The Blue Devils began the game with a 71.2 percent WP and were the favorites for the vast majority of the game's possessions. The only time the game was technically close was with five minutes left, when Quincy Acy threw down a monsterous dunk and Brian Zoubek missed a short jumper, and Baylor's WP swung to 60 percent. Two subsequent 3-pointers from Nolan Smith, though, give Duke a lead and WP that it would never relinquish.
From perusing each graph, it would seem that, of all the classics Duke played last season, its 79-72 loss to Maryland March 3 was statistically the best game it played in all year. Let's take a look at the graph and see why:
The Blue Devils, which hadn't lost a game in a couple of months and were the game's favorites despite having to play in raucous and raunchy Comcast Arena, began the game with a 63.6 percent WP. This quickly changed to a 62 percent Maryland WP after the Terrapins took a 21-8 lead at the 12-minute mark. Sean Mosley's 3-pointer with six minutes left in the first half pushed Maryland's WP to 75 percent. It didn't look good for Duke.
Then, the Blue Devils went on a remarkable run to close the half. The bar drops dramatically during this time. By the time Jon Scheyer made two free throws to give the Blue Devils a 46-42 lead, Duke's WP was at 75 percent.
Jordan Williams and Adrian Bowie gave Maryland a 54-49 edge soon after that, though, and the bar swings again with the Terrapins boasting a 70 percent WP. After Scheyer hit a 3 (what a rarity that was on that night) with 5:37 left in the game, Duke's WP moved to 70 percent.
Buckets from Bowie and Vasquez allowed Maryland to regain the lead and the WP advantage. Williams' layup with 1:40 left meant the Terrapins had a 60 percent chance of winning. That figure would never get lower. By the time Vasquez hit that lucky savvy shot with 39 seconds left, Maryland's win was statistically already in the bag. A lot of lead changes meant a lot of movement on the bar. Not surprisingly, the game is a classic.
Also, in case you were wondering whether Duke's 82-50 win over North Carolina was ever in question: Duke began the game with a 94 percent WP. It moved to the 98-99 percent range by the time there was nine minutes left in the first half, and it was at 100 percent by halftime. Can you say blowout?
Get The Chronicle straight to your inbox
Signup for our weekly newsletter. Cancel at any time.