University of North Carolina Athletics
Lucas: UNC Basketball Mailbag
January 25, 2011 | Men's Basketball, Featured Writers, Adam Lucas
Jan. 25, 2011
By Adam Lucas
Straight to the questions this week. Make sure you check the Tuesday Talking Points for some interesting stats about this year's Tar Heels, including an illuminating factoid on Carolina's play under Roy Williams when coming off a week-long break (as the Tar Heels will do Wednesday night at Miami).
I was just curious as to why I have started to see people report stats per 40 minutes rather than per game. Does it not artificially inflate how well, or poor, a guy is doing overall. Is there an overall purpose to keeping stats per 40 minutes rather than per game, or is this just a modern way we are going to be recording stats. I'm a bigger fan of the old school ways of points-per-game, assists-per-game, etc. Maybe it's just me but I need an explanation, if there is one, as to what we benefit from as fans knows what these players do per 40 minutes rather than per game.
Jeremy P.
Auburn, Indiana
I'm as guilty as anyone of starting to shift more towards slightly different stats, so this felt like an important question to answer. The overall statistical shift began in baseball, where you have to consider Bill James the godfather. He was the first one to point out that many traditional baseball stats, while convenient, didn't have much relevance to determining actual value towards winning a game. And that's what we're trying to do with stats, right? We want them to give us some idea of how important a player might be to his team (or compared to his peers), or how one team compares to another. Stats are basically a way to keep us all from getting in fistfights when we say, "The Cubs have better pitching than the Cardinals." It helps to have numbers to back up those assertions.
The more cerebral stats approach began to leak over to basketball with the publication of Dean Oliver's "Basketball On Paper." The NBA began tracking plus/minus on its official site and also uses some of the per-40 stats. Those types of figures have been slower to trickle down to college basketball, partly because there are so many Division I college teams that it's difficult to calculate them for every single team. But certain sites do a terrific job with it, with my favorites being DraftExpress.com and Ken Pomeroy's terrific site.
Now, what's the value of those stats? Well, let's say I tell you that Kendall Marshall is a very good passer. You knew that already, of course. He's averaging 4.56 assists per game. That's very good, and in fact it ranks third in the Atlantic Coast Conference. But the two players above him are Nolan Smith and Reggie Jackson, who also happen to rank in the top four in the league in minutes played. So how could we balance out the minutes to make a perhaps more accurate comparison of the way they've racked up assists this year? Right, by using the assists-per-40-minutes stat. And when we do that, Marshall not only leads the ACC at 11.3 assists-per-40, he leads the entire country by a wide margin (nearly 3.0).
To me, that's a better measure of his skill set. It's not intended to imply that if Roy Williams would just play Marshall 40 minutes every game, he'd hand out 11.3 assists (that one-third of an assist would be "the way we keep them"). Instead, it's just meant to balance possible playing time discrepancies. The Marshall/Smith/Jackson comparison is a great illustration, because you're comparing a freshman to two centerpieces of other teams. Looking at the per-game stats might be deceptive, but I think if you watched Carolina, Duke and Boston College play, you'd come away with the same conclusion the per-40 stats give you: in the passing category, Marshall is having an exceptional season, and is perhaps more accomplished in that area than the other two players.
Incidentally, if your perception is that John Henson is a blocks machine, you're correct--at 4.9 blocks per 40 minutes, he ranks fourth in the nation in that category.
I do think it's possible to go overboard with the numbers. According to the per-40 stats, Patrick Crouch would average 14.5 points per game--more than both Marshall and Justin Watts--if he could just play all 40 minutes. With all due respect to Blue Steel, that seems unlikely. Somewhere there's a balance between cold numbers and a feel for the game. At some point, no matter what the numbers say, you have to go out and play the actual game, and weird things happen in every single game in America at all levels that numbers can't anticipate. But that doesn't mean we should keep using all the old stats just because that's the way it's always been. The key is finding the stats that work best for you and figuring out how to apply them.
Has there ever been a UNC team that used two PG's successfully (time on the floor 20/20ish time split)? You can determine what successful is but I would lean towards wins and post season run.
Rob Walters
Pittsburgh, PA
Under Roy Williams, this hasn't happened with the split Rob describes. The team that probably comes closest is the 2006-07 squad, which saw Ty Lawson eventually take over the starting point guard role from Bobby Frasor. Lawson played 25.7 minutes per game that season and Frasor played 10.1.
My guess, though, is that Roy Williams might tell you there is precedent for playing two point guards very successfully, and it comes from the final years of his tenure at Kansas. That's when he used Kirk Hinrich and Aaron Miles together in the same backcourt for most of his final two seasons in Lawrence, and they took him to the national championship game in 2003.
That's also why, for all the talk over the summer about Larry Drew II learning to run a team, and his passing, and all those other point guard intangibles you know so much about, the area that might ultimately determine his minutes this season is his outside shooting. No matter what you think about the situation, it appears well-established that Williams wants Drew on the floor for his defense. In fact, the junior won the coaches defensive award against Clemson. But both the numbers (that we went into above) and the "feel" of the game seem to suggest the offense is running a little more smoothly at the moment with Marshall in the game. So how do you get around that?
Right, you play them together, which the Tar Heels did in the win over Clemson. That was partly out of necessity, because Leslie McDonald's injury jumbled the perimeter rotation. But the major flaw in the Carolina offense right now is outside shooting, as the Tar Heels rank next-to-last in the league in three-point accuracy. UNC's overall shooting numbers in conference games are substandard. If moving off the ball enables Drew's three-point percentage to improve markedly from his current 20.7% figure, then it's something that would have to be explored. Basically, even as we prepare to move into February, the door is still wide open for any of a group of players--Drew, Reggie Bullock, Harrison Barnes, Dexter Strickland and McDonald--to step forward and become the consistent go-to perimeter shooters. Because it's such an important offensive need, whoever fills that role is going to get minutes regardless of any other factors.
Just to finish up Rob's question, some other Carolina teams in the shot clock era that consistently played more than one point guard: 2002 (Boone/Scott/Morrison on occasion), 1989 (King Rice and Jeff Lebo, although Lebo was playing more shooting guard by this point in his career),1986 and 1987 (Lebo and Kenny Smith, with Smith the primary ball-handler but Lebo a great second option--man, these were really fun teams and probably the best comparison to the Hinrich/Miles backcourt), and the Steve Hale/Smith combination in 1984 and 1985.
Brownlow's Down Low
There has been some talk recently about how much "tougher" this team is than last year and how we "gutted out" some wins. ... I think I would measure "toughness" as points or rebounds on the road vs. home, since road games add a whole other level of difficult. I would be curious to know who had the biggest difference in either points or rebounds on the road vs. home.
Melissa Luffman '98
Singapore
Lauren writes: In the Roy Williams era, Raymond Felton is one of the first players I think of when I hear "toughness." He averaged more points on the road (12.8) than at home (11.4), shooting better from both the three-point line and the foul line. Postseason numbers are worth considering as well. Sean May gets the obvious nod in that department with his 19.9 points and 61% shooting. But Melvin Scott is worth a mention, shooting 47.6% and nearly 40% from three in the postseason. He averaged just 5.9 points but added 2.0 assists and 0.45 turnovers. And no one will ever forget how capably he handled the team when Felton fouled out against Villanova.
But Danny Green might win this award. He shot 46.9% on the road and 40% from three (46.5% at home, 37% from three) and averaged 10.2 points (9.4 at home). He had double figures in all but two of the road games his senior year; in total, he did it in 16 of his final 23 road games (he had just six double-figure games in his first 20). Wayne Ellington is not far behind Green. His first big road moment was his game-winning three-pointer at Clemson in overtime, but in 2008 overall, he averaged more points (17.4 to 15.8) and shot better from three (43.1% to 39.5%) on the road. In his career, he shot better at home but had more points on the road (14.8 to 14.5). And his 2009 postseason speaks for itself.
In 2009, Ty Lawson cemented his ACC Player of the Year Award on the road. He hit the game-winner at FSU and had 21 points. He had 25 in a dominant performance at Duke, then 21 at Miami - including the game-winner - while being sick. His toughness was questioned during his tenure at Carolina and doubters were not silenced until his brilliant game on a bad toe against LSU.
John Henson is the closest this team has - he averaged more points (7.8 compared to 5.2) and rebounds (5.1 to 4.1) in 2010 road games. This year, he is averaging 10.5 on the road and 12.1 at home. But there have been some nice road wins this year with contributions from many players.
Adam Lucas is the publisher of Tar Heel Monthly. He is also the author or co-author of six books on Carolina basketball, including the official chronicle of the first 100 years of Tar Heel hoops, A Century of Excellence, which is available now. Get real-time UNC sports updates from the THM staff on Twitter.
