Morris: Dean Smith and the rise of basketball analytics

I blame it all on Dean Smith. This new-wave obsession with statistics in college basketball carries a link to the late legendary coach. He was a mathematics major at Kansas, you know. Yet Smith also was the master of suppressing the analytical part of the game, thus indirectly leading to a deeper examination of the game’s numbers.

In some ways, Smith would love the idea that his favored points-per-possession statistic is now easily available to players, coaches and fans for every college team in the country. On the other hand, he might not appreciate that anyone can find out through kenpom.com that North Carolina ranks 231st in the country in 2-point field goal distribution among its players.

This all started with Smith.

Forgive me for lacking the exact details, but either during the 1981 or 1982 season Smith made a claim that one of his star players was a much better free-throw shooter when a game’s outcome was on the line.

The All-American James Worthy was a rather sub-par, 65 percent career free-throw shooter for his three-year UNC career. I happened to have collected all the play-by-play accounts of UNC games that season and went back to study the final five minutes of games. Worthy, as I recall, was no better or worse late in games, thus refuting Smith.

You had to understand the times. Statistics were difficult to come by. Bare-bones box scores were distributed after games and a few newspapers in the ACC area carried those. There was no Internet, and some programs only accumulated numbers they wanted distributed to the media and public.

Smith, for instance, listed players on his UNC statistics sheets alphabetically rather than the traditional manner by scoring average. He also did not allow turnovers to be listed on his stat sheets. That was considered by Smith to be a negative statistic.

So, generally, when Smith made a claim such as that of Worthy and his free-throw accuracy late in games, you simply had to nod and believe him, mostly because he was among the first coaches to accumulate stats and come up with such data as points per possession. He just kept those numbers under wraps.

A ton of information

Eventually, the NCAA began to standardize statistics gleaned from games, and every program now has the same data compiled and released from game to game and season to season. With the advancement of videotaping of games came an explosion of advanced metrics.

Bill James is widely given credit for a deeper look at baseball statistics that now goes by the name sabermetrics, which is short for statistics gathered by the Society for Baseball Research. In college basketball, Synergy Sports Technology and kenpom.com have taken the analysis of the game statistically to another level.

Every college program subscribes to the two services. Duke defenders, for instance, knew in the final seconds of their 2010 national championship game victory over Butler to push right-handed shooter Gordon Hayward to his right on the final missed shot. In its scouting report, Duke learned through statistics provided by Synergy Tech that Hayward went left nearly 70 percent of the time, according to bleacherreport.com.

“There’s no question it’s a boatload of information,” said UNC coach Roy Williams. “It’s a ton of information out there that you can get about anything.”

Nearly every program today has either a full-time employee in charge of statistical analysis or a couple of assistants assigned to studying the data and/or attempting to match the statistics with videotape.

Valuing statistics

When Kentucky coach John Calipari decided in 2014 to go with a five-man platoon system, he hired a full-time statistical analyst. Calipari’s thinking, according to espn.com, was that his NBA prospects might average as few as 20 minutes played per game. He wanted someone to project statistics for his players to present to NBA scouts.

Buzz Williams of Virginia Tech has become known as the “Rain Man” of college basketball because of his obsession with statistics. Among the statistics he most values is the number of touches an opponent gets in the lane during a game.

Not all players or coaches share the same affinity for the numbers. Anthony Gill, Virginia’s outstanding senior forward, said Wednesday that he peruses the statistics included in every pregame scouting report of an opponent but really hones in on a couple of categories.

“If there’s a big that averages a lot of offensive rebounds, I know I have to really focus in on him and try to keep him off the glass,” Gill said. “The next thing I look at is their blocked shots. If they have a shot-blocker on the team, I’ve just got to be aware of that and remember where he is on the court every time I get ready to shoot the ball.”

UNC’s Williams said he probably falls into the old-school crowd when it comes to analytics. He admits to still reading the general, less-detailed weekly statistics released by the ACC office. But he also knows that if he claims that Marcus Paige has been an outstanding offensive player this season that someone can simply look to kenpom.com and find that Paige ranks 225th nationally in offensive rating percentage.

Williams can blame Smith for that.

This story was originally published March 17, 2016 at 12:50 PM with the headline "Morris: Dean Smith and the rise of basketball analytics."