Friday, August 2, 2013

The Secret Mitzvah of Steve Novak

This guy is going to propel the Raptors into the playoffs. I'm serious.

The Raptors's offense was pretty average last season. Literally average. They ranked 14th in offensive efficiency rating (102.9). You really can't get more average than that (except you can, like the Hawks at 15th; I'm a bit buzzed, folks).

Their semi-offensive woes weren't all that complicated when you parsed through the roster. They simply had too many inefficient shooters taking way too many shots. Demar Derozan, Rudy Gay and the (thankfully) departed Andrea Bargnani are nobody's models of efficiency. Their shitty shooting was buoyed by the marksmanship of (the sorrowfully departed) Jose Calderon, Amir Johnson and Jonas Valanciunas.  

So the solution seems easy right? Take some shots away from Demar and Gay, and hand more to Johnson and Valanciunas. Ba-da-bing, like in The Sopranos, right? (R.I.P James Gandolfini)

Well, basketball isn't that simple. Sure, Coach Dwayne Casey could tweak the offense and hand more shots to Amir and Valanciunas, but how can he do so? How many could he redistribute? How will the redistribution of shots affect efficiency? How will the increased workload affect other aspects of Amir and Valanciunas's game (ie: what if they're too tired and their defense suffers)?




Just as a quick aside; this is one of my major beefs with the efficiency vs usage debate. The book is still out on how increased usage affects efficiency (or increased efficiency affects usage). There's quite a bit of literature on either side; the side that argues that increased usage negatively affects efficiency, and the side that argues that there is no statistically significant effect. My polite beef is with the latter.

In my opinion, there's only so much an offense can produce. It's like having a 100 bucks for groceries. You go to the store, and you'd love to buy 100 bucks worth of potato chips, but you're never going to. You're going to pick up vegies, you're going to buy milk, eggs, whatever. At the end of the day, your shopping cart's composition is always more or less the same.

In the same way, an NBA offense will always have some isolation, some pick and roll, some transition, etc etc. There isn't a large discernible difference between NBA offenses because it's simply not possible. I'm sure a coach would love to run a pick and roll every time and kick the ball out to a wide-open Ray Allen in the corner. That's probably an awesome play, but there's way too much standing in the way of that happening. Defenses adapt, different players are used, shot-clock and time left impacts how much a player shoots. A million things happen in the course of an NBA game.

Consider the play-by-play breakdown of the (average) Raptors's offense, and the (excellent) San Antonio Spurs offense:






Are the differences that drastic? Not really. The player personnel could not be more different; Toronto has ball dominant wings in Demar and Gay, and San Antonio has a well-balanced offense focused around Tony Parker and Tim Duncan. Yet, the end result, in terms how how possessions are utilized, are essentially the same.

What's the difference? The players who shoot, and the environment that they shoot in. The Spurs offense is able to generate a lot of quality, open shots, and they have more accurate shooters. That's the difference, and it's a huge difference.

---

Luckily, Toronto won't have to drastically revolutionize NBA offense to become a better offensive team. They have two things working for them. First, Andrea Bargnani is no longer on the team. Last season, Bargnani took 9.6 shots per game at an abysmal 48.3% TS% (average Centre: 54.1%). Secondly, his replacement Steve Novak (yes, that pasty white man in the cover picture) will be his replacement.

Take a look at the Raptors's shot selection. Can you see where they struggled?


Yeah, that's right. The most frequently used play (spot-ups; 20.9%) was also their worst (by NBA ranking). That's not a good thing. I wonder why that is. Was it perhaps...BARGNANI? DEROZAN? GAY? 

Okay, I might have given it away. Yeah, it was those three. I organized it into another table below:


Yeah, that's pretty awful. Thankfully, the Raptors jettisoned Bargnani to New York for a bag of dicks and some draft pick. Wait...what? Steve Novak was in that bag? Who left Novak in that bag of dicks? Oh, he chose to be there voluntarily? Why was h-nevermind, we'll take him. We'll use him. What's he good for, anyway?

Well, let's see what how New York used him last season:


I tried. I really did. I basically full-screened that graph and it still refused to show the percentages on those tiny slivers. I organized the stats into a table below:


What did Novak do well? If you can't answer that, I question why I spent so much time embedding charts into this post. Novak was a spot up shooter, and was one of the very best in the game. You see how good that PPP is? It would take two spot-ups from Gay to score as much as Novak would in one. That's crazy efficiency.

The answer now seems simple, right? No? What is wrong with you? I ask a lot of questions to you non-existent blog readers. 

Look, just redistribute all of Bargnani's spot up shots to Novak. Maybe toss in some of Demar's and Gay's. That would instantly improve upon that shitty 0.94 points per play on spot ups that Toronto scored last season. The differences will be subtle; maybe it'll only be 2-3 points per game, but those 2-3 points may very well win 5-10 more games this season.

The Raptors's optimal lineup next year will probably look something like: Lowry-Gay-Novak-Johnson-Valanciunas. The first option should be a Lowry-Johnson pick and roll with Gay on the wing, Novak in the near corner and Valanciunas on the weak-side. See the following (incredibly amateur play diagram)


Here's how it's working out in my mind. First, Johnson screens Lowry's defender and creates a mis-match with Lowry going up against Johnson's defender. If this defender drops back, Lowry shoots. If he steps up, Lowry drives into the paint. Johnson darts to the basket as this happens, forcing Novak's man to rotate over to cover the rolling Johnson (who is one of the league's best roll-men). Lowry then has the option of passing one over to Johnson (who can then drive, or if Valanciunas's man rotates over, he can pass over to Val (who should be wide open). Lowry can also choose to zip it to an open Novak, who will shoot (duh!) while Johnson and Valanciunas crash the boards (both above average at offensive rebounding and proficient at putbacks). If Lowry is plugged, he can reset over to Gay at the top of the key, where he's actually pretty awesome at shooting 3's (see my Rudy Gay breakdown).

Failproof, right?

Well. no. But it really plays into every player's strengths. Novak is awesome at spot ups. Valanciunas is great at grabbing offensive boards and put backs. Johnson is an excellent roll-man. Lowry is a good three-point shooter and dishes well. Finally, Gay is given open shots where he's most efficient. Sounds like a dream, right? Step aside, Nick Nurse. I gots this (no not really; my shitty play won't do much; come back!).

Be prepared, folks. Novak is going to be a fan-favorite, and rightfully so. He's going to revolutionize the Raptors offense, and we're all going to have to get used to this (and forget about how douche-y this looks), because it's going to be awesome.



5 comments:

  1. "I ask a lot of questions to you non-existent blog readers." oh we do exist. Great analysis and fun read. While I was reading realgm I almost lost hope. Your post brought back some optimism in me

    ReplyDelete
  2. Oh my god! You guys exist?!?!?

    All kidding aside, thanks for the kind words. I really appreciate it. I'm glad you like the content. I'll aim for 3-4 posts a week. Keep coming back!

    Thanks again

    ReplyDelete
    Replies
    1. I'll definitely keep coming back man. You post really interesting stuff. On reddit (i lurk, don't have account lol) you posted wins produced projection by Arturo and he projects around 45 wins for Raptors, more than Indiana. I'm excited but skeptical at the same time(too good to be true) I was wondering if you know how close his projections were past seasons or how close wins per projections are in general.
      Thanks

      Delete
    2. The truth is that no current statistical model is any good at predicting NBA outcomes. The Wages of Wins guys even said it themselves here:

      http://www.thenbageek.com/articles/we-re-all-bad-at-predicting

      ---

      I think the reason Arturo's projections ranked Indiana so low is because of a few things. First, they lost some productive pieces of their bench (to Toronto no less) in Hansborough and Augustin. Second, he probably expects Danny Granger (not productive) to steal Lance Stephenson's minutes (is productive). The wins produced also doesn't like Scola's current performance all that much (he's getting old).

      I think Indiana is a little overrated in the minds of many because they almost beat the Heat in the ECF. However, 7 game samples are subject to a lot of randomness, and they just happened to matchup really well with the Heat (nobody could match up with Hibbert). They were the best defensive team in the league (tied with memphis), but they had a bottom-12 offense. By net rating (OffRtg-DefRtg) they were 7th.

      That being said, I think they'll hit 50 wins, no problem.

      Delete
    3. Thanks for reply! Man I hope it works out for Raptors. Keep it up man, hope to read more stuff from you.

      Delete