Monday, February 10, 2014

Catching Up

It’s been almost three weeks since my last post. I’ve started and then not finished a couple of long posts, the ‘Cats have played four games including two huge road wins, pundits have (nearly) declared Chris Collins’s mission accomplished, and some fans have even dared to consider NU’s NCAA tourney resume. Talking NCAAs and Northwestern in February? It’s like 2012 up in here!

I won’t bother with in-depth game recaps, but I think it’s worthwhile to post the efficiency numbers from the last four games:

Iowa 76 (1.171 PPP), NU 50 (0.771), 65 possessions, -0.40 efficiency differential
NU 65 (0.953), Wisconsin 56 (0.821), 68 possessions, +.132 efficiency differential
NU 55 (1.013 !!!!), Minnesota 54 (0.995), 54 possessions, +0.018 efficiency differential
Nebraska 53 (0.88), NU 49 (0.814), 60 possessions, -0.066 efficiency differential

For two games, NU’s offense showed some signs of life--heck, it was the offense that won the game in Minneapolis.  Those two games were bookended by two typically inept offensive performances. The defensive efficiency ranged from stellar (at Wisconsin) to atrocious (home vs. Iowa), but was more consistently good than the offense.

About the offense: It’s bad—0.86 points per possession. The “lack of shooters” explanation is not nearly enough to explain offense this abysmal, but I don’t have a more complete explanation myself. The only B1G offense that has been nearly as bad in conference play in the KenPom era was the 2004 Penn State squad that managed only 0.897 PPP. In fact, this offense is the worst in the country in conference play. There are 351 D1 teams in KenPom's rankings. 8 of them have OEff ratings below 90 in conference play.

For more on those games and NU’s efficiencies, I highly recommend Carmody Court. Now to one of the posts I had previously abandoned.

Get Lucky

About a week ago, I tweeted about NU’s KenPom luck ranking.

I found the stat interesting and thought others might, too. Some of my followers bristled.

Nick Medline’s tweet embodies what I suppose is the attitude of most fans, even those for whom the concept of “luck” isn’t anathema: I don’t care as long as my team is winning. Kevin Wallace’s subsequent tweet promulgates one of sports’ long-standing clichés: winning close games is a skill. Others offered another cliché: teams make their own luck.

I guess I shouldn’t be surprised that sportswriters and fans bristled at my invocation of luck. There may be many reasons for this, but I’d like to address two. First, I think they misunderstand what luck means. Second, even if you explain it to them it’s too disruptive to narratives that are comforting.

What is Luck?

The luck about which I tweeted is a very specific and relatively simple measurement. Ken Pomeroy defines it thusly:

“A measure of the deviation between a team’s actual winning percentage and what one would expect from its game-by-game efficiencies. It’s a Dean Oliver invention. Essentially, a team involved in a lot of close games should not win (or lose) all of them. Those that do will be viewed as lucky (or unlucky).”

To put it another way, teams should be about .500 in close games. A common definition of close games for basketball is two possession games, or games decided by 6 or fewer points. After its one-point win in Minneapolis, NU was 5-2 overall in such games (3-0 in B1G). That’s quite good. It’s also quite lucky. Jeff at Basketball Predictions noted:

“Just like we all predicted three weeks ago, Northwestern is in fourth place in the Big Ten. Just in the past two weeks, they've won at Indiana, Wisconsin and Minnesota. That said, Northwestern isn't the fourth best team in the Big Ten, or even close. The reality is that they've been pretty damn lucky to win these games. All five of their Big Ten wins have come down to the final minute, while all five losses have been blowouts. They are still last place in the Big Ten in PPP differential in conference play, at -0.13 PPP.”

He goes on to say some really nice things about the future of NU hoops and Chris Collins. You should read the whole post.

The point of this is not to say that the ‘Cats don’t deserve to win or that they haven’t played well or anything else like that. It’s just to note that the basketball gods have been kind to NU in these close games. Take the Minnesota game, for example. Tre Demps shot a horrible air ball on a non-play in NU’s final possession. (Why on earth did Drew Crawford not touch the ball at all on that possession?) The Gophers then got a very close shot and a good look at a follow-up, both of which missed. If either of those shots drops, NU loses the game. The ‘Cats played well enough to win either way, but one outcome is a win and one is a loss, and it’s basically a 50/50 situation.

How would the story change had NU lost? What you heard leading up to the Nebraska game was “won five of their last seven”, “.500 in league play”, and “won three straight road games for the first time in over 50 years”. All three statements are true and sound pretty good. If the ‘Cats get one unlucky bounce against Minnesota, then the story changes. “Won four of their last seven” doesn’t sound as good. “4-6 in league play” is the same story as last season. “Still haven’t won three straight road games in the B1G since 1960.” NU isn’t a different team in either scenario; its record would be, but its strengths and weaknesses would be identical. And yet the narrative would be dramatically different.

The thing about this kind of luck is that it usually runs out. Teams trend towards the mean over time. That means .500 in close games. Some teams have extremely lucky stretches that cover entire seasons. Take Nicholls State this year, currently #1 in KenPom’s luck ratings: The Colonels are currently 7-1 in two-possession games, including two OT wins. That’s abnormal, and so was NU’s record in close games entering the Nebraska game. With its four-point loss to the Huskers, NU has dropped to 5-3 overall/3-1 in B1G in two-possession games and fallen to 68th nationally in luck (still #1 in B1G, though). For what it’s worth, KenPom projects five of NU’s final seven games to be decided by six or fewer points. That’s five more chances for luck to creep in.

More Luck

Another major form of luck is defensive 3P%. I’ve discussed this before, and Ken Pomeroy has expounded at length on the topic. NU has been exceedingly lucky this season. I know many people want to credit the ‘Cats revamped defense with opponents’ struggles behind the arc, but I’m not buying it. I’ll need to see multiple seasons of this to be convinced otherwise. Again, I refer you to KenPom’s multi-post study of the topic; read it before you attack me for saying NU is lucky in this respect.

How lucky has NU been? I set boundaries at +/-5% of an opponent’s season 3P% for unlucky/lucky. This is not entirely arbitrary, but it does merit more examination as a threshold. This graph shows that the ‘Cats have been extremely lucky.

On the flip side, I believe that teams’ poor 2P-shooting against NU has much more to do with the ‘Cats excellent switching defense. Teams are not getting a lot of comfortable shots inside the arc, and that is accurately reflected in NU’s defensive 2P% (44.7% overall, 45% in B1G).

Even More Random Luck

Another form of luck all too familiar to long-time NU fans is the injury. This is the first season since 2009 that the ‘Cats have not lost a major player for an extended period to injury. I hesitate to say that being injury-free is lucky. It’s more that experiencing injuries to your top players is extremely unlucky.

Then there is officiating. Saturday’s game against Nebraska featured a segment with a bad goaltending call going against NU and a non-call in favor of Nebraska. In a four-point game, that’s a huge deal, and there is literally nothing NU can about it. I also think of things like odd bounces, especially the funny rebounds that just fall into an offensive player’s hands despite good rebounding position by the defense. The fact is every game is full of instances of luck. If a team plays significantly better than its opponent, then luck has a much smaller influence on the outcome of the game. As the point margin shrinks, the impact of luck becomes much larger.

This season the ‘Cats have been lucky more often than not. Unless they experience a real breakthrough on the offensive end, their slim NIT hopes hinge heavily on that luck persisting. That’s not a great position in which to be.

Here's an Awesome Song

It's your lucky day. Enjoy Richard Thompson's blisteringly sinister of the Donovan classic, "Season of the Witch".

No comments:

Post a Comment