Wednesday, December 3, 2014

Sometimes A Win Isn't A Win

Northwestern has now played approximately 20% of its basketball season. It has won five games and dropped one against the 36th easiest schedule in Division I and done so in highly unimpressive fashion. I expected the 5-1 record, but I did not expect single digit margins of victory against Houston Baptist, North Florida, Elon (in OT no less!), and Miami (OH). Winning ugly is only good in the playoffs or against similarly positioned teams. A win isn't always just a win; scraping by against bad teams, at home no less, does not speak well of a team. The 13-point road win over Brown looked impressive at the time, however, Brown has turned out to be worse than expected (dropping over 70 spots from its pre-season ranking). NU’s performance in Providence now looks like an anomaly in comparison to the rest of its games.

If you’re looking for hope, don’t look to the remainder of the schedule. The ‘Cats’ next two opponents, Georgia Tech and Butler, are considerably better than all the teams they have beaten so far. Home games against Michigans, Central and Western, are no gimmies. Considering NU’s performance in its wins, those games against Mississippi Valley St., UIC, and Northern Kentucky are not either. After that, it’s B1G season, in which Ken Pomeroy now projects a 4-14 record for the ‘Cats. They are only favored by Pomeroy’s model in one of those games.

NU has gone from 77th in the KenPom pre-season rankings to 118th entering the B1G-ACC challenge. Why is that? The simplest answer is that the offense has not improved nearly as much as Pomeroy’s model expected and that the defense has regressed somewhat. The ‘Cats have a lower AdjO efficiency than the Rutgers team that scored eight (8!!!!) points in the second half against Virginia last Saturday. While NU’s AdjD efficiency is nearly identically to its 2014 defensive efficiency (94.7 and 94.2, respectively), the D1 average efficiency has dropped from 104.3 last season to 99.3 this season. In other words, the ‘Cats are much closer to average on defense this year. There are deeper reasons for these rankings, so let’s take a look.


What’s driving the decline in NU’s defensive efficiency? Let’s start with the four factors from this season and last.

NU’s opponents are shooting moderately better this season, but they’re also turning the ball over more often, collecting fewer of their own misses, and getting to the line about as often as last season. Overall, the ‘Cats are above average in OReb% allowed (D1 average is 31.6%), still slightly above average in eFG% allowed (48.3%), and right at average in TO% and FTRate.

What’s behind that increase in eFG%? It’s all about the 3P%. Last season NU was 41st in the country in defensive 3P%. As I noted in my defensive preview there was a great deal of luck in that number and it was unlikely to stay that good. This year NU is 280th in defensive 3P%; the luck is evening out. Interestingly, the ‘Cats’ 2P defense is even better this year than last, and it was pretty damn good last year. NU is again benefitting from well-below-average opponents’ FT%. That may not hold up as the season progresses. 

Where can NU improve defensively, then? It can force more turnovers via steals. Last season the ‘Cats were well below average, committing theft on a mere 6.9% of their opponents’ possessions. This year they’re at 4.8%. For context, the D1 average this year is 9.5%. Going for steals, either through zones that jump passing lanes or man-to-man traps, is a high risk/reward strategy. Missing a steal can lead to an easy bucket for the opponent, but getting a steal can do the same for NU. Given the offense’s struggles, that is a trade-off that might make sense.


Despite Chris Collins’s claim that he has more firepower at his disposal this year than last, the ‘Cats are not much better on offense. NU finished 2014 with an AdjO rating of 96.3 (309th in D1) and through six games this year has an AdjO rating of 98.0 (203rd). That’s an improvement of 0.017 points per possession and translates to 1.06 points per game at NU’s AdjTempo of 62.2 (329th in D1 if you were wondering).

NU isn’t shooting any better than last year or getting to the line more frequently. It is turning the ball over more often but also corralling more of its own misses, a trade-off that seems like just enough to boost NU’s scoring ever so slightly. 

When we break down that eFG% to its sub-components, we see that NU is much better at 3P shooting this year. 35.3% is good enough for 129
th in D1 (average is 33.5%) and a significant improvement over last year’s 30.7%. However, NU’s 2P% has cratered to 296th in D1 (average is 47.4%). 

What’s worse is the ‘Cats are taking more of their FGAs from 2P than they were last year (as shown by the lower 3PA/FGA this season). This is not a good strategy, and despite of NU’s improved 3P% it has the same point distribution as it did last season. This team should be taking a lot more 3P shots. 

Currently NU has two players shooting above D1 average from 2P range: Sanjay Lumpkin (86.7%, 13-15, and 0 2PAs in the loss to UNI) and Johnnie Vassar (100%, 2-2), neither of whom currently plays a noticeable contributing role in the offense. Vassar is currently a turnover machine and should not get more than 7.8% of minutes he’s currently playing, but Lumpkin has shown great improvement over last season in terms of shooting and ball security. The coaches would be wise to find a bigger role for him in the offense.

The three top guards in terms of minutes, Bryant McIntosh, Tre Demps, and JerShon Cobb, have combined to shoot 34.1% from 2P range, a figure so bad it gives me a migraine just thinking about it. The rest of the roster is hitting a mere 41.1% from 2P range, which is also not good. Meanwhile, the ‘Cats have six players at or better than average in terms of 3P%: Alex Olah, McIntosh, Scottie Lindsey, Cobb, Lumpkin, and Dave Sobolewski. So, again, this team should be taking a lot more 3P shots.

Let’s look a little more closely at some individual players.

Alex Olah

There’s a widespread perception that Olah has struggled this season, and it’s absolutely right. His usage is way up (from 18.4% to 24.4%) and his efficiency is way down (from 101.2 to 95.1) compared to last year. He’s turning the ball over more (20% up from 17.8%), but he’s also generating more assists, so that’s a bit of a wash. His 3P% is about average (33.3%), but with only 6 3PAs Olah is a non-factor from the outside. His free throw rate is up (73.8 from 44.1) and is 99th in D1, which is awesome. His 67.7% FT percentage is almost average (D1 average is 68.3%), so that’s really a net positive offensive contribution. Where he has struggled is in his 2P shooting, hitting only 41.7% of his 2PAs. Last year he hit 54.5% and as a freshman he hit 43.2%, so he’s having his worst season in terms of 2P% right now.

Others have pointed out that Olah is holding the ball too low, which I think is accurate, and that he seems tentative, which is also accurate. I would also posit that he may be getting the ball too far from the basket, and that NU’s emphasis on the pick and roll probably doesn’t play to his strengths. I suspect he may also be suffering from too many dribble-drive-and-dish action by NU’s guards. Again, I think dishes off those drives may not play to his strengths. Whatever the case, Olah needs to find his fire, and the coaches need to figure out where and how to get him the ball. NU’s offense will not be good if he isn’t contributing efficiently.

JerShon Cobb

Cobb entered this year with a career 2P% of 45.8% but is now at 26.7% on the year. His TORate is the highest it’s ever been, and his ORtg is the lowest it’s ever been. He hasn’t played such a minimal role in the offense since he was a freshman, and Chris Collins noted in his UNI post-game interview on WGN that they haven’t figured out how to get Cobb more involved. He is hitting a solid (and career-high) 35.3% of his 3PAs, but that number is skewed by his 3-4 performance at Brown as he’s taking just under 3 3PAs per game. Maybe he isn’t healthy enough to do some of the things he used to do. Maybe he’s checked out a bit mentally. Regardless, I think the best approach for Cobb is one I would recommend for most players on this team: more 3PAs and more action at the rim (if not on drives then on cuts) with an emphasis on getting to the line. The fewer midrange 2PAs Cobb takes the better.

Tre Demps

I think Demps is a decent player who works hard and has come up with some big shots late in games for NU, but the narratives of Demps as clutch and a dynamic scorer are overblown. I really hope that the coaching staff doesn’t buy into them. Here’s why: he’s a woefully inefficient shooter. His 38.5% 2P average is down from his career average of 41.2% entering the year. His 3P shooting is also down, from 34.2% to 29.6%. Even his foul shooting has plummeted, from 69.7% to 57.1%. The only thing Demps really has going for him offensively is his assist to turnover ratio of 2.4 (12 ASTs to 5 TOs). Considering that he takes 29.4% of NU’s shots when he’s on the court, those shooting percentages have an outsized negative impact on the offense. If Demps weren’t so inefficient most of the game, NU wouldn’t need his late game heroics. The ‘Cats don’t seem to have a better option on the bench than Demps (Lindsey can shoot but is a walking turnover, while Taphorn and Sobolewski both seem to have no role in this system), though, so they really need him to start shooting better. Short of that, the coaches may need to instruct Demps to shoot less frequently.

Bryant McIntosh and Vic Law

I’m grouping McIntosh and Law together because they’ve been the two most efficient offensive contributors of the regulars and because they’re sort of yin and yang to one another. McIntosh is best when creating for his teammates and shooting from the outside, while Law is best when getting to the rim.

McIntosh boasts a gaudy ARate of 34.0 (67th in the country) and a passable TORate of 22.6. For reference, Sobolewski was at 19.9 and 16.6 as a freshman while Juice Thompson was at 26.6 and 17.7. Those guys played on teams where everyone created for everyone else, whereas McIntosh does not. McIntosh is also shooting a blistering 47.4% from 3P range (Juice hit 43.3% as a freshman). On the flipside, McIntosh is a mere 32.4% from 2P range and has an almost negligible FTRate of 16.1. This suggests that while he is good at facilitating for teammates, he is not very good at generating his own points. It’s no secret that I loathe the midrange 2P shot, and McIntosh would be well advised to cut back on the 2P jumpers and floaters in the lane.

Law meanwhile has a decent ARate of 19.4 that is eclipsed by a TORate of 19.5. He’s hitting only 21.4% of his 3P shots and a middling 44.8% of his 2P shots. Where he excels is getting to the line. His 51.2 FTRate is 317th amongst D1 players, and he’s hitting a phenomenal 86.4% of his FTAs. Still, Law settles far too often for the dreaded midrange jumper. His value lies in his ability to get to the rim and the line. The more aggressive Law becomes, the better this offense will be. I would much rather see him get his shot blocked or turn the ball over attacking the basket than taking 13-foot baseline turnaround jumpers. In fact, if Law and McIntosh combine to take three 12-foot to 19-foot shots per game that’s at least two too many.

The Rest of the Guys

I’ve already mentioned that Lumpkin deserves a bigger role. I don’t think Vassar is at all ready for D1 ball. The coaches seem to have written off Sobolewski and Taphorn. Lindsey turns the ball over way too much for a guy who doesn’t create for his teammates, but his shooting skill may outweigh that shortcoming. I expect to see more and more of him as the season progresses. Skelly is a fantastic rebounder (15.2 OR% and 30.6 DR%) but turns the ball over more than anyone not named Vassar or Kreisberg and doesn’t look like a scorer yet. Finally, Kreisberg doesn’t seem to have any role offensively. He’s only taken two FGAs and turns the ball over at a 47.1 rate (yikes!). Skelly seems to be getting Kreisberg's minutes, and I don’t think that’s going to change.

So, What Does All This Mean?

My prescription for the offense is not much different than it was before the season. Minimize (or eliminate) the midrange jumper, get guys open threes from their best spots, get to the line, and crash the offensive glass. Right now, NU isn’t doing any of those things particularly well. To me that suggests the coaches do not have the right strategy for the players they’ve got. They showed the flexibility to change up the defense in 2014, to great effect. Can they do the same for the offense this year?

And A Phenomenal Song

Built to Spill - "You Were Right"


  1. Sometime a win is not a win, but a loss is always a loss.
    Thankfully they scheduled lightly in order to build confidence....unlike the disastrous nonconference scheduling for football, which drained fan interest in September and eventually cost us a bowl game.

    I predict NU goes to the CBI with a record slightly above or below .500 record.

  2. A loss is definitely a loss. I'm not sure this team will win more than 12 games, so even the CBI will be a stretch. Next year the CBI or NIT will be a must. By 2017 Collins should be getting bubble talk in February or his fate may be similar to every other NU basketball coach in history. And I won't be blaming him for that.