clock menu more-arrow no yes mobile

Filed under:

Marshall Working On His Defense

Some Friday morning reading on the UNC point guard.

First up, Ryan Fagan at The Sporting News examines Kendall Marshall's defensive effort versus Wisconsin and explains the game plan Roy Williams had in mind for guarding the Badgers' Jordan Taylor.

It’s no secret that Marshall, the sophomore point guard at North Carolina who is a gifted passer, is a work in progress defensively. He doesn’t move fast enough laterally, and he’s not quick enough to stay with the quickest of this year’s crop of opposing point guards.

He’s just not.

Staying in front of the point guards who love to penetrate into the paint is an issue. Getting around/over screens on the perimeter is another issue. But here are three things that should make Carolina fans feel better about the player Marshall needs to become: He knows his flaws, he’s working on those issues, and coach Roy Williams is giving him the opportunity to fix those deficiencies.

There is no hiding. Not now, and if he gets better, not ever.

Fresh off a performance in the Tar Heels’ loss to UNLV that was, to be kind, subpar, Marshall was front-and-center from the beginning of the game Wednesday night against Wisconsin’s Jordan Taylor, an All-American candidate.

“I was excited, coming into this game, after not defending the way I wanted to against UNLV,” Marshall said. “I felt like I could help my team so much more just making it tougher on the other team’s point guard. I was excited for the challenge. I don’t want to shy away from anything coach challenges me with.”

The plan was for Marshall to stay with Taylor for the first half, then slide shooting guard Dexter Strickland over to mark Taylor for most of the second half. Because, while it’s important for Marshall to improve, Williams isn’t about to let one player win a game by himself by exploiting one particular matchup advantage.

Two takeaways here.

One if Marshall self-awareness of his deficiencies and this desire to improve. Marshall understands his shortcomings on defense but doesn't shy away a challenge. He also understand such challenges give him an opportunity improve on his defense. Roy Williams also understand this which is why he was willing to go with Marshall on Taylor in the first half purely for the purpose of allowing Marshall to work on his defense. The plan was to have Strickland take over for the entire second half but his three fouls prevented that so Roy waited until the game reached a critical point to make the switch. Part of UNC's rally from down 36-31 to ultimately take control of the game was fed in part by Strickland guarding Taylor and disrupting the Badgers' offense(such as it is).

This kind of move by Roy also underscores one of the reasons why UNC is not necessarily as crisp in the early part of the season as they might be down the road. Roy loves to tinker with everything from the substitution pattern to the defensive match-ups. Roy could have easily put Strickland on Taylor from the start and been done with it. Had this been a Sweet Sixteen or Elite Eight game, that is definitely what would have happened. In November Roy has one eye on winning the game at hand and one on long term team success. If he can serve both at the same time everyone wins.

Another interesting tidbit of a statistical nature from's Luke Winn who brings up the problem of assists numbers being inflated.

I wish some stat-head with a ton of time on his hands (I have a lot, but not enough for this) would start a national assist-authentication service. The No. 1 plague on America's box scores is what I call assist inflation -- the awarding of assists that are not assists according to the very extensive language in the NCAA Statistician's Manual. The manual says: "An assist should be more than a routine pass that just happens to be followed by a field goal. It should be a conscious effort to find the open player or to help a player work free." It also states very clearly that if "a player is well-guarded and has to make a move to get free," then the pass that preceded the basket not an assist. Thus normal post entries that require significant maneuvering on behalf of the post player are not assists, nor are simple perimeter passes where the guy does not catch and shoot, but instead squares up and makes a significant move before scoring.

So much subjectivity is involved, with so little oversight, that assists are not consistently credited in the same way nationwide. Consider this case study which took me a few hours to compile, comparing play-by-play data with video: The two major-conference assist leaders are North Carolina's Kendall Marshall (10.3 per game) and Pitt's Tray Woodall (8.3), but after reviewing every assist of theirs that was available on film, one player has a much higher rate of inflation. Note that this is not that player's fault, but rather the fault of statistician who's distorting the definition of an assist -- in some cases to the point of absurdity -- in order to produce inflated box-score numbers.

* Of the 69 Marshall assists that were in Synergy, I deemed 62 (or 89.9 percent) to be by-the-book assists.

* Marshall's home scorekeeper in Chapel Hill was actually more honest than the road/neutral scorekeepers for UNC games. At home, 28 of 31 (90.3 percent) of Marshall assists were legit, and in road/neutral, 34 of 39 were legit (87.1 percent).

* As for Woodall, of his 56 assists available on film, 44 (or 78.6 percent) were authentic.

* What was interesting were Woodall's home-road splits. All 12 of his road/neutral assists were legit ... but just 32 of his 44 home assists (72.7 percent) were legit. This included some egregious inflation. A few examples: In the home opener against Albany, he gets a box-score assist for a simple pass that precedes a full pick-and-roll play by two teammates; and gets another one like that against Rider in Game 2. In Game 4 against La Salle, Woodall is credited with an assist on possession in which he last touches the ball 23 seconds before Nasir Robinson scores. In Game 6 against Robert Morris, Woodall is credited with one assist in which he last touched the ball eight seconds before Ashton Gibbs scores on a two-dribble pull-up -- which came on a pass from a different teammate. And the worst one, also in the Robert Morris game: Woodall gets another assist on a possession in which he never touches the ball.

What I would like to know and Winn doesn't cover it is which assists credited to Marshall does he think are incorrect and why? We know there are many instances where Marshall leaves little or no doubt that it is an assist. I am curious which ones Winn doesn't think are assists.