In the post-mortem takes on Michigan State's last game of the season, a 2-point loss to UCLA in the first round of the NCAA Tournament, the word of the day seemed to be 'microcosm'. As in, the game was a microcosm of MSU's season, featuring a big early deficit and a spirited comeback that ultimately fell short and ended in disappointment. I'm not sure I'm completely sold on that comparison, but one player whose game against UCLA did seem somewhat microcosmic was Keith Appling. A non-factor for much of the game, he came alive late to hit a series of clutch threes, showing flashes of offensive brilliance. Still, he finished with merely 9 points and no assists in 29 minutes of action. So the game and the season left us with this question: which is the real Keith Appling? Is it the guy who played tenacious defense but had no role at all in the offense for most of the game, or the guy who lit it up for 9 points in 2-and-a-half minutes to almost pull out a miraculous comeback?
On a team featuring Kalin Lucas, Durrell Summers and a restyled jump-shooting Draymond Green, the best shooter was Keith Appling. Not the best scorer - that's a different question. But he had a shooting line of .471/.411/.895 (2pt%/3pt%/FT%). He was tops in 3pt%, second only to Mike Kebler in FT%, and better than Lucas, Green and Summers inside the arc. He also led the non-Garrick Sherman Spartans in true shooting percentage, which factors 3-point shooting and free throws into shooting efficiency, at 59.3%. The problem was this number: %Poss - 15.2.
Possession percentage, or usage rate, is a pretty reliable indicator of a player's impact on the offense. If you look at a team of 5 players, the average player will use 20% of the offensive possessions, by taking shots, making assists, getting fouled or turning the ball over. Go-to guys will be higher than that - Kalin Lucas was at 26.9% - and role players will be lower. 15.2% is pretty low. Guys with a similar number this year included Stu Douglass, Mike Bruesewitz, D.J. Byrd and Alex Marcotullio. This is some indication that Appling's role in the offense this year may have come as something of a surprise.
Here's a look at the trend in Appling's possession percentage over the course of the season (charts courtesy of Statsheet.com). Interestingly, Korie Lucious' last game was January 22 against Purdue, roughly the beginning of Appling's final downward arc.
But here's the same graph showing Appling's percentage of minutes per game.
So as his minutes climbed, his offensive role disappeared. Some of this reflects his emergence in the role of defensive specialist, but on a team as starved for offense as Michigan State was this year, this showed amazing resolve on the part of both Tom Izzo and Appling.
These numbers may be less of a concern for the past year than for coming years. With MSU losing two of the top scorers from a team that was already only #62 in Pomeroy's offensive efficiency rankings, it's a real question where the points are going to come from now. In a post from earlier this year, intended to give some context to expectations for MSU's freshmen, I commented that
Keith Appling (RSCI #34) is also making decent progress. His offensive rating of 103.4 is ahead of the curve and while you'd expect his minutes and usage to be higher (currently 41.8% and 19.4% respectively) the recent decision to begin starting him will probably bump these up. . .Well, it didn't quite work out that way. Although Appling's offensive rating remained fairly steady (he finished at 101.9), and his minutes actually climbed, his usage, as we have seen, took a significant dip. Here's Luke Winn, from the article that was the occasion for my post (remember that Appling had an average recruit ranking - RSCI - of 34):
That means the following expectations should be considered UNreasonable:
- That anyone outside the top 20 will appear in more than half his team's minutes or be much higher than a point-per-possession player.
- That anyone outside the top 10 will use possessions at the rate of a "go-to" guy, or score with All-America-level efficiency
So although Appling exceeded expectations for efficiency and playing time he was not, clearly, a "go-to" guy. In fact, he was significantly below average in usage for players of comparable ranking (21.9% usage).
The cause for concern here is that, even in spite of situations where we might expect a dramatic rise, players rarely make huge leaps in usage from the baseline they establish, even as freshmen. Ken Pomeroy took a look at this issue several years ago and produced the following graph, showing the normal range of change in usage from year one of a player's career to year two:
The blue lines indicate the range within which 75% 50% of players fell and the outer gray lines encompass 95% 90% of the cases. So even under conditions where one might expect a dramatic jump, say the departure of a high-usage point guard who carried the team for four years and whose first name rhymes with Palin, it is often less dramatic than is hoped. Pomeroy doesn't provide the raw data, but subjecting the graph to analysis under a scanning electron microscope reveals that that outer edge of Appling's possible increase looks to be around 22%. Exceptions to this are not unknown: Evan Turner emerged from the long shadow of Kosta Koufos and Jamar Butler to become a 30% usage guy. But they're extremely rare, and even Turner started at 21% his freshman year.
The good news is that MSU may not need more than that from Appling on the offensive end, especially if Green, Adreian Payne and some of the newcomers can pick up some of the slack. The pleasant surprises of Appling's first year were his effectiveness on defense and his suprising ability as a rebounder. In fact, among guys his size (6'2") in the Big Ten, only Tim Frazier, Talor Battle and Jordan Taylor posted better defensive rebounding percentages than Appling. And he was far and away the best shot blocker for a guy under 6'4" in the conference. His athleticism is undeniable and will continue to be a source of excitement in the coming years for the Spartans.
The big question going forward, of course, is how effective Appling will be as a point guard. Although I prefer to view the point guard's various offensive responsibilities as separable, Izzo doesn't seem to share my views. He has always seemed to want one guy who brings the ball up, calls the plays and generally runs the show. At the moment, that guy is probably going to have to be Keith Appling, with whatever help incoming freshman Travis Trice can give. Although it would be interesting to see Appling and others bring the ball up but run the offense through Draymond Green, for example, that is likely to be more of a fallback strategy for Izzo should Appling struggle at the point. And although there are plenty of low-usage point guards around (Aaron Craft, Lewis Jackson, Tim Frazier) Appling's low assist rate (11.5%) and high turnover percentage (26.7%) will need to come a lot closer together for him to be effective in this role.
So there are a number of ways Keith Appling's sophomore year could go. He could maintain the Chris Allen transformation into a defensive and three-point specialist or he could emerge as a primary focus of the offense. Appling will be a player to watch carefully, maybe THE player to watch, to get a read on the course the Spartans' season is likely to take in this upcoming year of uncertainty.
[This is the first in a series of posts on the major players from the past season and players we'll see for the first time in the coming year. Not all of them will be as extensive as this one. If you like this kind of thing it's available in bite-sized chunks by following me on Twitter: @connertp.]