Wednesday, April 1, 2009

AccuScore Likes 'Nova's Chances

So If you've been on ESPN at all tonight, you may have noticed a little headline on the side of the page that reads "AccuScore: 'Nova will upset North Carolina." I saw that headline and said to my blue self, "Fuck yeah, I wanna read that shit." Unfortunately its for insiders only, of course. Since I'm not an insider and don't know anyone who is an insider, I did a little research to find this article somewhere else. Since I'm the man, I did find it, on a UNC blog. I decided to just straight up steal this shit from their blog for three reasons.

Reason 1: I really don't feel like changing his words around, or interpreting all the numbers from AccuScore.

Reason 2: It really should have been our news to report anyway since it favors the 'Cats, not the broken toes or whatever they call themselves.

Reason 3: Fuck them, that's why.

So without further ado, here's the article plus some commentary from the writer for the UNC blog for all of my fellow outsiders:

Even though the Villanova Wildcats are 7.5-point underdogs in Vegas, if they win, they most likely will win by seven or more points. If the North Carolina Tar Heels beat the Cats, expect a double-digit margin (around 15 points). In other words, you can ignore the spread.

The two outcomes hinge on two stats: 3-point shooting percentage and rebounding.

When Villanova won in simulations, the Wildcats shot 38 percent from 3-point range while North Carolina was a cold 31 percent. Villanova also slightly out-rebounded the Tar Heels. That makes a lot of sense, as we just saw a similar scenario play out in the Elite Eight.

In its upset win over Pittsburgh, Villanova shot just 30 percent from 3-point range, but this was better than Pitt's 27.8 percent. The undersized Villanova squad also out-rebounded one of the best rebounding teams in the country 29-27.

Free throw percentage was a huge factor for Villanova against Pittsburgh, but it will not likely be a big factor against the Tar Heels, as both teams shoot well from the line. Turnovers always play a major role, but Villanova can win even if it commits more turnovers than North Carolina. In simulation wins, Villanova averaged 13.1 turnovers to Carolina's 12.6.

When North Carolina won in simulations, the Tar Heels had the edge in field goal percentage and 3-point percentage and, most importantly, were plus-5.3 on the boards. When you combine the rebounding edge with a 3.7 turnover edge, the Tar Heels got nine more possessions out of these categories. Greater offensive efficiency (higher shooting percentages) combined with more possessions can result in an easy Carolina win.

Spread 'Em

The Sweet 16 and Elite Eight games were filled with blowouts. Other than Pittsburgh's games against Xavier and Villanova and the Kansas-Michigan State game, every game ended with a final margin of more than 10 points. The average point spread in those 12 games was 4.4, yet the average final margin was 13.4.

We are left with four very good teams, and most experts will predict close games. But if there are close games Saturday, it will be a surprise. For the season, when a team was favored by four or fewer points, the average final margin was 8.91. This is not much less than when teams are favored by 7.5 or fewer points -- 9.34 average final margin.

How Has AccuScore Performed?

On point spread predictions, AccuScore's predictions have generated profits for both teams:

For the record, here is AccuScore's over/under prediction records for these two teams. Both of these teams have generated profits.