A Tale of Two Offenses: Indiana Pacers vs. Orlando Magic

I know what you’re thinking.  The main difference between the Pacers’ offense and Magic’s is that one sucks, and the other is Orlando’s.  Well, that’s a fair point.   The Pacers provide no competition in two of the most important offensive categories.  Orlando was 2nd in eFG% and 4th in offensive efficiency (points per 100 possession), while Indiana landed 22nd and 26th in those categories, respectively.

Over the last couple of seasons, I’ve watched the Indiana Pacers struggle, and, at the same time, watched portions of the Pacer fan base struggle to understand what offense Jim O’Brien is trying to run. Or why he’s trying to run it.  As you would expect, this hasn’t been the most pleasant experience.

After watching and listening, I’ve really tried to see what Jim O’Brien’s end goal was.  I’ve tried to see if there was a roster in the NBA that O’Brien would view as ideal, and how Obie’s tactics would evolve if he had that roster.  I’ve basically come to the conclusion that Orlando is the best example.

From a roster perspective, I think it would be far and away Obie’s favorite of the 30 teams.  Dwight Howard anchors both ends as a dominating inside presence, while almost everyone else on the roster brings a nice blend of athleticism and shooting. From a stylistic perspective, there would certainly be some differences if you were to substitute O’Brien for Van Gundy, but I don’t think they’d be huge.

In order to understand them, I tried to analyze the differences this year using a couple of wonderful sites:  Synergy Sports and Hoopdata.

Nuts and Bolts

The basic design theory of the two offenses is similar — an array of shooters around an inside presence — but Orlando runs more sets, while the Pacer run (ostensibly) a motion.  The Pacers also run at a faster pace.  If you break it down specifically, you begin to find more (seemingly) significant differences.  However, I believe those differences are actually more practical than philosophical.  In other words, I believe that the offensive philosophies of Stan Van Gundy and Jim O’Brien aren’t all that different.  At the core of them is a concept that has been brought up so often in Pacerland that some people treat it as an obscenity: spacing.

However, personnel does matter, and it shows it the specific ways that Orlando and Indiana try to use the spacing created by the threat of the three in the half court.

(Note: Shots and plays for different types of actions come from Synergy Sports, and represent “finished plays.”  As such, they do not include plays where the initial action was a Pick-and-Roll (PnR) or a Post Up, but the finishing play was a spot up or a cut or offensive rebound.  Still, I believe they serve as an adequate surrogate for this discussion.)

Difference # 1: Pick-and-Roll (PnR)

Orlando runs a lot more PnRs.  They got almost 1,500 shots out of PnR action, which accounted for about 22% of their total.  The Pacers, on the other hand, only got about 800 or 12%.

The PnR, at its very core, is meant to force the defense to “pick its poison.”  The ball-handler is the crucial component in the equation, because he has to threaten the defense in three ways: (1) with the drive, (2) with the pass, and (3) with the shot.

About 78% of the Magic’s shots coming out of the PnR came from the ball-handler.  The Magic used Jameer Nelson and Vince Carter extensively as ball-handlers, accounting for about 51% of the plays and 65% of the shots coming from that type of action.  Both threaten the defense, at least with the shot and the drive.

The Pacers, on the other hand, really only have one player on the roster that can effectively run the PnR: A.J. Price.  Earl Watson and T. J. Ford are terrible in this set because they offer the defense the easy solution of simply going under the screen and letting ’em shoot.  Additionally, they’re both very poor at making the quick decisions necessary as they come off the screen.   Looking at the wings offers you no other glowing candidates.  Neither Danny Granger nor Brandon Rush have exhibited the handles or passing ability, while Dahntay is basically just a bigger T.J., offensively.  Indy’s ball-handlers are so poor in this role, that it’s almost impossible to critique the roll man.

Difference #2: Post Ups

The other area that Orlando utilizes more than Indiana is, unsurprisingly, is post up action (meaning low post).  They got 880 shots out of the low post, or roughly 13% of their total.  The Pacers got 584 shots, or 8.5% of their total.  The difference here, in my opinion, is almost entirely reflective of the difference between Dwight Howard and Roy Hibbert, both in terms of physical capabilities and overall role/development (which includes minutes played).

Howard got 62% of Orlando’s shots in the post while Roy got about 67% of Indy’s.  That’s a little incomplete in terms of actual post-play finishes, because it doesn’t factor in FTs.  Howard was able to draw shooting fouls on 18% of his plays in the post, while Roy only did it about 7% of the time.

Looking at both rosters, Howard and Hibbert are really the only major back-to-the-basket players.  Brandon Bass is, to some degree, but his 633 minutes this year was not much more than Tyler Hansbrough had for the Pacers.  This is why I attribute most, if not all, of the difference in post usage to the difference between Dwight and Roy.  Roy, in my opinion, is simply not ready to take on the load to that degree.  In fact, it’s far from a sure thing that he ever will be.

However, it’s important to put this difference in perspective.  In terms of overall usage (an estimate of the percentage of team plays used by a player while he was on the floor), the two are not all that different.  Dwight Howard’s is 23.9, while Roy’s is 22.2.  Where they differ is in terms of minutes and overall role/importance to the team.  However, it seems to me that Dwight Howard would — and should — have a much larger role on any team than Roy would, at least at this point in his career.

I took a look at where and when Roy was used in the offense and benchmarked it against a random sampling of other post players.  While far from exhaustive, it was somewhat informative.  What it tells me is that while there may be some disenchantment with how much O’Brien uses Hibbert, there can be little question that when O’Brien uses Roy, he’s using him in a classic big man role.

The chart above shows the distribution of Roy’s shots out of each different action.  Just a tick under 50% of them came from the post, which compares favorably with the likes of  Tim Duncan (41%), Carlos Boozer (21%), Andrew Bynum (49%) and Elton Brand (32%).  This distribution suffers in comparison to Dwight Howard (62%) and Shaquille O’Neal (69%), but that’s understandable on a number of levels.  Hibbert lacks the strength, athleticism and experience of those guys.  I did look at one guy who didn’t have huge physical advantages over Roy, but still got 69% of his shots out of post action: Al Jefferson.  I haven’t watched him enough to give a fully developed opinion, but he might make a good template for how to get somebody a lot of looks in the post who lacks the freakish attributes of a Howard or Shaq. (And he should also be of interest to Pacers fans who remember the Wolves offer of Al Jeff for Danny earlier this year.)

This chart shows Hibbert’s shot distribution by location.  Again, his “inside” shots (within 10 feet) are a pretty impressive 70%.  Duncan only gets 59%, while Brand is 46%.  Boozer, who spends less time in the post, took 63% of his shots in this area.  The physical beasts — Shaq and Howard — take over 90% of their shots from there.  Now, this just represents a thumbnail, and I’d need to factor in fouls drawn and free throws attempted to get a full picture, but it looks to me like Roy is being used in essentially the right manner.

The question of “how much” is a little fuzzier (and I’ll give that more attention in future project).  The short answer is that O’Brien is almost certainly not using Hibbert as much as he should, but I don’t think that it’s as drastic as some might think.  When he’s on the floor, he is not being ignored.  His usage rate is pretty solid, and he’s getting over 11 post touches per 48 minutes, which is comparable to Al Jefferson and Tim Duncan, who get between 12 & 13 per 48.

I also don’t believe the reasons for this are as dogmatic or ideological as many of O’Brien’s detractors claim.  I think it’s a function of two major things.  First is Roy’s relative inexperience and physical shortcomings.  These are valid concerns, but I think Jim is overdoing the hand-wringing here.  The second is the usage of Danny Granger.  This offense really only consistently runs plays for two positions:  small forward and center.  All of the other players on the floor are secondary or fall-back options.   When the plays for Danny are run (which is the majority of the time), Hibbert is brought up to the high post.  This makes him a passer, instead of a scorer.  Also, Roy’s looks in those sets are limited by what, in my opinion, is Danny’s disturbing propensity for “busting plays.”  (Again a topic for another post.)

Difference #3: Pace

Jim O’Brien likes an up-tempo game, but that doesn’t mean that he’s always run it.  His teams here are faster than his teams in Boston and Philly, but again, I think that’s somewhat driven by necessity.  In all honesty, I cringe when I think of this roster trying to run an offense against an established defense in the half court.  Roy’s emergence over the second half provides some promise, but it’s still a roster that lacks scoring punch in the half court.  Also, the Pacers got a little faster this year, because they were usually desperately trying to catch up.

In any case, it would be really interesting to see what O’Brien would do with pace if he had Orlando’s roster.  This season, Indy played at 97.1, while Orlando was at 92.0.  I think the speed at which the Pacers played was exaggerated by their struggles, as I mentioned above.  Orlando ran at basically the same pace as last year, which is right around the league average.  How O’Brien would deal with this probably would create the biggest differences between the two coaches.

Freaky Friday

So, what would Obie’s Pacers look like, offensively, if we had Orlando’s roster?  My wild guess for distribution of shots by action would look like this:

In this chart the shaded blue area shows the current Pacer distribution.  The black outline is Orlando’s, and the red is what I think Obie would do with Orlando’s roster.  I do believe that he would keep the pace up, so the transition stays the same.  However, I’m sure that he would scrap the motion offense in favor of more PnRs and post ups.  The post ups, as a percent, would drop marginally from Orlando’s rate due to the higher pace.

In both offenses, “iso” represents a failure.  With the possible exception of Vince Carter, there are no players on either roster that you’d want to clear out and let pound the ball.   “Other” includes cuts, hand-offs and off-screens.  The Pacers do more of these as a result of the motion offense and employing a high-post big man.  Orlando has no high post big men, and O’Brien would use more PnR than motion, given the presence of Carter and Jameer Nelson.

In the end, I think Jimmy would like a shot location distribution almost identical to the royal blue line of Orlando’s.  The shade Pacers area indicates that they’re below average at the rim, which is bad, but above average from beyond the arc, which is a positive.  Orlando does a pretty damn good job of stealing shots from the mid-range (10-23 feet) to feed threes.

In my mind’s eye, a perfect distribution would pull more from the mid-range to get the “At Rim” percentage to above the NBA norm (the red line).  The mid-range is where offenses go to die.

So What?

Yeah, this was little more than an intellectual exercise.  O’Brien will never get Orlando’s roster, and I find it exceedingly unlikely that he’ll be here beyond next year.  Still, I think there’s some value in doing this.  Levy2725 over at Indy Cornrows did an amazing analysis on Obie and coaching effectiveness a couple of weeks ago.  I didn’t agree with all of his conclusions, but it is simply spectacular work.  (I must admit, I’m a little intimidated.)

This was just an attempt to put some more context into the discussion.