Behind the Numbers: Who Is Killing Their Own Team?
Posted by KCarpenter on December 1st, 2011
A lot of the effort in basketball analytics goes towards the good things that players do that do not appear in the box score. This is the driving idea behind Michael Lewis’s seminal New York Times feature, “The No-Stats All-Star,” an early look at analytics in the NBA primarily focused on Darryl Morey, Shane Battier, the Houston Rockets, and adjusted plus/minus. This makes sense: finding hidden strengths is the coach’s angle while finding hidden value is the economist’s angle. As a result of the fine work of smart guys with formulas and others with a willingness to watch a lot of games closely, Charles Jenkins and Nate Wolters were household names last season. This, of course, assumes that your household is filled with basketball dorks, but you get the idea.
Faried Was An Underappreciated Star
Finding diamonds in the rough is a noble pursuit and talking up the greatness of underexposed and underrated players is a worthwhile task (Hey there, Kenneth Faried!). Sometimes, however, there is a joy in using analytics and “advanced” statistics to look for the guy who is hurting his team the most. Let’s ignore the diamonds and go straight for the rough.
How does a player hurt his team? Well, when push comes to shove, there are basically only two ways: offensively and defensively. Sadly, however, contemporary box scores assign no grade for bad defense to the individual outside of counting how many fouls (which could very well be offensive) a player commits. Our primary understanding of player’s individual defense comes only in positive contributions like blocks, steals, and defensive rebounds while the effect on an opponents shooting percentage is measured at a team level. The noble effort of Luke Winn, David Hess, and others that has sought to enact Dean Oliver’s defensive charting schemes is a good start at really quantifying individual defense, but a very small percentage of Division I games have been looked at in this way making the approach of limited use to someone who wants to look at the whole of college basketball. So, acknowledging that analytic approaches to finding bad defensive players are limited, let’s at least take a quick look at fouls.
Last year, California‘s Markhuri Sanders-Frison averaged an impressive 3.9 fouls per game or, more tellingly, 5.8 fouls per 40 minutes. In addition to leading to free points for an opponent, foul trouble would limit the playing time of Sanders-Frison, which was unfortunate as he was the best rebounder and second-best offensive option on the team. Still, Sanders-Frison never fouled out and only managed to accumulate 118 fouls last season, pathetic compared to the 139 fouls racked up by Cleveland State‘s Aaron Pogue who also fouled out six times. Tough defense is the calling card of Cleveland State, but letting opponents take free shots while routinely losing your starting center and best rebounder to foul trouble is just not a sound strategy. Of course, Pogue’s foul outs pale in comparison to Oral Roberts‘ Steve Roundtree who somehow managed to foul out a whopping eleven times last season. Roundtree was the team’s best offensive rebounder and most dangerous interior threat, averaging a team best 57.0% from inside the arc. You would think you might want to try to keep this guy on the floor. Still, perhaps, you can attribute his foul tendencies to freshman immaturity. What’s that you say? He’s already fouled out once this year? Nevermind.
Now let’s look at where analytics has a lot more to say about players hurting their teams: on the offensive end. There are two primary well-tracked ways that player can hurt their team on the offensive end and they are both pretty obvious. Players hurt their team by missing shots and turning the ball over. Both of these kill a team’s offensive efficiency. Of course, there are lots of players who are lousy shooters or bad ball handlers in college basketball, but most of them know their limitations and simply refrain from shooting or handling the ball too much. The guys who kill their team are the ones who don’t know or don’t care how bad they are hurting their team. So let’s take a look at guys who attempted at least 150 field goals last season with the worst percentage. Somewhat surprising to me, the “winner” is a starter on a power conference team that made the NCAA Tournament last year: Dash Harris of Texas A&M. Harris essentially played the role of a shooting guard who can’t shoot. His true shooting percentage comes out to 35.7%. He shot 26.8% from the field overall, 16.7% from three, and 56.2% from the line. He played the third most minutes on an NCAA tournament team. This boggles my mind.
Dash Might Want To Start Passing More (Credit: Jim Prisching / Houston Chronicle)
The other way a player can hurt his team is by turning the ball over. In low tempo games this hurts a lot more, and it’s honestly expected that the players who handle the ball the most will get the most turnovers, while post players and catch-and-shoot guards barely even have any opportunity to give the ball to their opponents. Turnovers can really hurt a team. Ask North Carolina. Although the Tar Heels eked out a three-point win against Wisconsin last night, they committed fourteen turnovers compared to four by the Badgers, who managed to slow the game down significantly. North Carolina won because they played defense fairly well and shot the ball at a reasonably good rate. Wisconsin played worse at both ends but nearly won. That’s what happens when your team gets nearly twenty more shots than your opponent. Wisconsin had sixty-four field goal attempts compared to North Carolina’s forty-five. If you do enough other good things to help your team, it doesn’t matter as much (Jimmer Fredette had the third most turnovers in Division I last year), but it can hurt.
Last year, Kevin Galloway was the unmatched master of turnovers, giving the ball away 160 times, which comes out to 5.0 turnovers per game. Kevin Galloway played the most minutes for Texas Southern and also managed 6.3 APG as a play-making point forward, which is impressive. He may have turned the ball over a lot, but certainly also made a lot of plays. A better way of looking at who was the worst ball-handler is to use our old buddy tempo-free statistics, and specifically turnover percentage. Steve Zazuri of Sacred Heart managed to give away the ball on 51.8% of possessions he received. Now, I don’t want to pick on Zazuri, as a seldom used back-up guard who rarely saw minutes, his lack of ability at holding onto the ball didn’t hurt the team as much as if he had been a starter. Low usage limited the damage he could do.
Now, suppose we want a catch-all measure of the ways that a player hurts his team the most on offense, taking into consideration shooting, turnovers, and usage. For this, “win shares” should be our go to measure. Developed by Bill James for usage into baseball, there have been attempts by different folks to translate the concept into basketball. The version that I like the best for college basketball is the one whipped up by the folks at Sports Reference. The method for calculating the measure is fairly complex, but I particularly like it because unlike Bill James model, this system points out players who negatively impact the system. Again, it’s worth pointing out that because of the way box scores work, a big chunk of defensive win shares is calculated by examining team defense factors, so for our purposes of looking for individuals who are hurting their team, we should probably limit our examination to offensive win shares.
If we look at the bottom of last year’s offensive win share chart, we get Justin Burton, a 5’11” bench player from Winthrop with -1.2 offensive win shares, who shot truly terribly last season, making 20.5% of his two pointers and 18.8% of his threes. Also with -1.2 win shares, we have our old buddy Dash Harris and Florida Gulf Coast‘s Marlon Rivera who averaged 3.8 turnovers a game and 31.8% shooting from the field. Rivera, who also averaged 5.7 RPG 5.4 APG, and 1.6 SPG contributed in other ways, but for the most part, he just didn’t do enough to offset his horrendous offensive performance.
All these numbers and figures are from last season and more than a few of these players have since graduated. What’s more interesting is the list of players who are well on their way to killing their team this year. So far Hillary Haley paces the field with -0.8 OWS already, though there are a number of power conference players and starters who aren’t far behind. Georgia‘s Marcus “No The Other One” Thornton shows up in top ten, while Maryland starter Nick Faust isn’t far behind, paced by his 32.3% true shooting percentage. Dash Harris is currently ranked 33rd, but with a little luck and concentration, he’ll find a way to slip off the leaderboard in a senior year. May he have success in failing to fail. Or something like that.
