By The Numbers: A statistical look at the nation's top players
Value Add formula, created by John Pudner, rates the value of individual players
Wisconsin's Jordan Taylor should be considered for national player of the year
Despite Vanderbilt's strong offense, it may not be a top-10 team this season
Last month, John Pudner gave a guest lecture to a sports business class at his alma mater, Marquette, entitled "Winning with Sabermetrics and Micro-Forecasting." It was about the concurrent rise in popularity of advanced statistics in the realms of sports and politics. A 46-year-old political professional, Pudner ran state-level campaigns for George W. Bush in the 2000 Presidential election, as well as Mitt Romney's Wyoming effort in 2008, and is the founder of a company called Concentric Direct, most of whose current work focuses on lobbying issues for corporations. Pudner's specialty is data analysis. "I live in spreadsheets," he said. "I like looking at them, and trying to figure things out."
As a longtime fan of Bill James' baseball books and a rabid fan of Marquette basketball, Pudner told students that he'd been curious about adapting the baseball sabermetrics concept of "Wins Above Replacement (Player)," or WAR, to rate the value of individual college hoops players. The metric he settled on, called "Value Add," attempted to quantify the percentage drop-off if, say, Ohio State were to give all of Jared Sullinger's possessions to a generic, ninth or 10th man on a Division I bench.
Pudner's formula was based on the work of Dean Oliver, who wrote the foundational tome Basketball On Paper, and used possession-based stats from Ken Pomeroy's kenpom.com, including minutes played, usage rate and efficiency. The results were first published on the Marquette-centric blog Cracked Sidewalks, where Pudner posts semi-anonymously as "bamamarquettefan1." That's where I came across the study -- and a database with Value Add figures for 2,540 returning D-I players -- this offseason. Pudner's work deserves a wider audience, and is yet another case of college-hoops enlightenment coming from sources outside the game. Pomeroy is a meteorologist in Utah, and Pudner is a political consultant in Alabama; neither has a basketball background aside from being a fan.
After poring over the database, these are four things I found most interesting and/or applicable to the 2011-12 season:
1. The value of Wisconsin's Jordan Taylor is such that he needs to be considered (along with Ohio State's Jared Sullinger and UNC's Harrison Barnes) for preseason national player of the year votes.
As a Marquette die-hard, Pudner was pleased that his study was bullish on two lesser-known Golden Eagles, Jae Crowder and Davante Gardner, but disheartened that the nation's No. 1 offensive Value Add player in 2010-11 was Taylor, the point guard of their intrastate rival. Pudner's formula estimates that Taylor added 9.46 percent more to the Badgers' offense than a replacement player using the same minutes/possessions. UConn's Kemba Walker was No. 2, at 8.29 percent, and BYU's Jimmer Fredette was No. 3, at 7.56 percent. "I certainly didn't design the study to make Taylor look that good, because I don't like Wisconsin at all," Pudner said. "But he accounted for a ton of offense on a low-possession team, and traditional stats tend to cheat people on those types of teams so badly."
The top 10 returning offensive players in the country, according to Pudner's Value Add formula, are:
2. Central Connecticut's Ken Horton is the best player you've never (or barely) heard of.
The reigning Northeast Conference player of the year, who averaged 19.5 points per game as a junior, is the only low-major player to crack Pudner's five-percent-over-replacement club. And lest you think the 6-6 Horton's standing on this list is inflated because he plays in a small conference, it should be noted that Pudner's formula is stacked against low-major players. It adjusts efficiency based on the strength of defenses faced, and uses major-conference replacement-player numbers as its baseline. What the 5.06 percent next to Horton's name means is that he'd theoretically be able to enhance a major-conference team's offense by 5.06 percent.
3. Vanderbilt's offense is going to be insanely good ... but the Commodores still might not be a top-10 team.
In terms of cumulative returning Value Add, Vandy's offensive figure of 18.75 percent is by far the best in the country, with Ohio State second at 15.80 percent, and North Carolina third at 13.70. The 'Dores' quartet of John Jenkins (6.18%), Brad Tinsley (3.68%), Festus Ezeli (3.48%) and Jeffery Taylor (2.71%) is excellent. But Pudner also attempted to assess defensive Value Add, and while it's not great on an individual level -- there just aren't stats available to properly assign credit for "stops" -- it can be worthwhile to examine on a team level. And once Vandy's essentially replacement-level defense (which ranked 88th nationally in efficiency) is factored in, the 'Dores only rank 12th in Pudner's Value Add Sweet 16 projections, just behind Florida State. UNC is No. 1 by a wide margin.
4. Two very important transfers will be taking the floor in the Big Ten.
Familiarize yourself with the names Sam Maniscalco (Illinois, formerly Bradley) and Brandon Wood (Michigan State, formerly Valparaiso), because their stats suggest they can make a serious impact on their new teams. Maniscalco's numbers as Bradley's junior point guard in 2009-10 give him an offensive Value Add of 4.20 percent, by far the highest of any Illini returnee. The next-closest player is D.J. Richardson, at 1.76 percent. Wood, a shooting guard who headed to East Lansing as a fifth-year senior after graduating from Valpo, has the second-highest offensive Value Add (3.50%) of any Spartans player, just behind point forward Draymond Green (3.54%). If Wood plays at the level Pudner's formula projects, he can account for much of the offensive gap left by the graduation of Kalin Lucas and Durrell Summers. That could very well equate to a top-three finish in the Big Ten, which would exceed most pundits' expectations, including mine, and establish Value Add as a worthy way of identifying next year's sleepers.
#DearAndy: Unlimited meals, faux Pelini, and the Juicy Lucy
Chocolate Making with SI Swimsuit Models