DFDepressed FanDepressed Fan



, all the time

Another Look At Defensive Problems

We've been watching an unwatchable defensive system/effort for basically the entire season. The rotations the coach makes are infuriating, the rotations the players on the floor fail to make are maddening. Now that we have nearly a half season under our belts, I decided it was time to take stock. A look at several advanced defensive metrics after the jump.

Let's start with a glossary of the terms we're going to be using, and brief explanation of what they mean/why they matter. All data for this post was culled from HoopData, if you haven't checked out their site, please do:

  • Pace: This is possessions per 48 minutes. Pretty simple, we've talked about it before.
  • DefEff: Defensive efficiency, points allowed per 100 possessions. This is the most important stat in my mind. Everything else is derivative of this number. The question we're trying to answer is why are they allowing x points per 100 possessions.
  • OPTS: The raw points-per-game allowed stat. Pretty much the only number you'll see in most newspapers. It's not meaningless, but it's also not very meaningful when you don't look at the other numbers.
  • OTS%: Opponent's true shooting percentage. Takes free throws and three pointers into account. Points/(2x(FGA+(0.44xFTA)))
  • OFG%: Opponent's raw shooting percentage, no weight for the type of shots taken, simply field goals made divided by field goals attempted.
  • O3P%: Opponent's three-point percentage. Made threes divided by attempted threes.
  • O%Ast: Percentage of opponents made field goals which were assisted on. This is a pretty good indication of how well teams are running their half-court offense against you. Typically, a higher rate is worse on defense, but there are exceptions (like when an opposing point guard is beating his man off the dribble and converting at the rim. Assisted field goals made/field goals made. 
  • OAR: Opponent's assist rating is a measure of how many possessions end is assists for the opponent, per 100 possessions. Probably a more accurate look than the stat above. Possessions ending in an assist * 100 / total possessions.
  • OTOR: Opponents turnover rate is a measure of how turnovers a team forces per 100 possessions. This is a much more meaningful number than raw turnovers, because pace is taken into account. Turnovers x 100 / possessions.
  • DRR: Defensive rebounding rate measures the percent of total available defensive rebounds a team collects. Again, much more meaningful than simply the number of defensive rebounds a team grabs. DREB/(Opponent's OREB + DREB). Anything above 75% I consider a strong rating here.
  • BLKR: Simply put how many shots does the team block per 100 possessions. Better than the raw block numbers because pace is taken into account. Blocks * 100 / opponent's possessions.
  • DEFR: Defensive plays per 100 possessions. A "defensive play" is a steal, a charge drawn or a block. Essentially, how many stops does a team get where the opponent doesn't even get a shot to the rim (not including 24-second violations). A great deal of these plays (minus the charges) have a high likelihood of leading to transition opportunities on the other end of the floor, so obviously it's an important stat for the Sixers. (100 x blocks + steals + charges drawn) / opponent's possessions.
  • OFTR: Opponent's free throw rate measures how much an opponent gets to the line. FTA x 100 / FGA. (I actually prefer FTM * 100 / FGA, to measure how successful they are from the line, but that's not the number included here. It doesn't make much of a difference unless you're crediting teams for fouling bad free throw shooters, and maybe you should.)
  • O3PR: Opponent's three-point rate measures what percentage of their total FG attempts come from three-point land. This stat isn't really all that meaningful to me. "Forcing" a team to take a ton of threes is good if they're hitting them at less than 33%, but if they're hitting more than a third of them, they're hurting you. 3PA / FGA.

OK, let's jump right into the numbers, refer to the glossary above if you get confused, but I'll try to explain everything to the best of my ability.  Bear in mind we're talking about three different seasons with three different coaches (for the most part) and also varying rosters. In '07-'08 they traded away Korver in December. In '08-09 they had Miller for the entire season. This year they have Brand but not Miller. There are definitely personnel differences, but overall, the core of the team has remained constant, and maybe even trended toward better defenders with the inclusion of Brand at the four, etc.

The team's pace has remained fairly consistent, relative to their pace the year before, but the league has gotten faster dropping their rank. Nothing too meaningful to be derived from that, other than the fact that part of Eddie Jordan taking us "uptown" was that he was going to turn us into a high-possession team, which has not materialized. Nor should it, really. But we'll get to that in a minute.

The defensive efficiency rating has taken a drastic turn for the worse. Two seasons ago we were top ten (according to HoopData's methodology, which differs from Basketball-Reference.com's and my own.), this season we're all the way down to #22 (our DFR has improved dramatically over the past 10 games or so). We knew this going in, now we're going to examine why.

The answer is bleak, very bleak. Take a look at the shooting percentages across he board. TS% we've dropped from middle-of-the-pack to third worst, overall field goal percentage from 18th to 26th and 3-point percentage we've dropped from middle of the road to dead last. I know we've been griping for years about the Sixers three-point defense, but the fact of the matter is they were league average for the past two seasons, this year they're beyond horrible.

Let's jump all the way down to O3PR, this is where it gets a little bit silly. Opponents' three-point attempts are down, significantly, this season vs. the past two, but that's only attempts, now let's look at the points scored per 100 possessions by the opponents:

  • '07-08: 25.74 points/100 poss. on 3PA
  • '08-09: 27.0 points/100 poss. on 3PA
  • '09-10: 27.68 points/100 poss on 3PA

So they're giving up almost 2 full points per 100 possessions more than they were two seasons ago on three pointers, but their opponents are attempting 1.3 fewer three-pointers. That's a troubling, troubling trend and I believe those numbers are significant.

If you recall, Tony DiLeo had some theories on how to make up the three-point deficit the team was sure to accumulate in any particular game. The way to do it was to attempt and make more free throws than their opponent, and to beat them in second-chance opportunities. Well, take a look at what's happened with the free throw rate. For the past two seasons, they were among the league leaders in not sending teams to the line. This year, they're slightly below average, and sending teams to the line nearly 10% of the time more than they were on average over the past two seasons.

One thing that came as a shock was the defensive rebounding rating. If I had to guess going in, I would've bet the DRR was lower this season, but it turns out it's right in line with the past two seasons. On second thought, though, you've still got Thad getting the majority of the minutes at the four, Brand and Speights are essentially swallowing up the minutes Reggie Evans, Theo Ratliff and Speights took last season. The mitigating factors have been Iguodala's increased production on the defensive glass (17.2% vs. 14.0 and 13.4 over the past two season) and Sam Dalembert putting up Rodman-esque numbers on the defensive glass (he's approaching 30% of available d-boards. Rodman's had a nine-season run of 30%+. Dalembert is second in the league to Dwight Howard so far this season.

Since we're talking about Dalembert, he's also second in the league in blocks per game, but far and away the league leader in blocks per 40 minutes (among regulars getting 25+ minutes/game). He's averaging 3.49 block/40, number two is Kendrick Perkins at 2.89). Dalembert and Brand combine to give the Sixers a tremendous team block rate, good enough to move the Sixers from 10th to 4th in that category.

Which brings us to the stat that makes me want to strangle someone. Defensive Plays Rate. As I said in the glossary, this measures the number of stops where the opposing team doesn't even get a shot up to the rim. The Sixers rank first in the league with 16.8% of possessions resulting in either a blocked shot, a steal or a drawn charge. When you look at the Sixers overall shooting numbers, you'd think these stats are somewhat hollow. Meaning, they're gambling for the big play instead of playing fundamental defense. This is true to an extent, but what's lost in that simplification is what they accomplished the previous two seasons, without Elton Brand. In the previous two seasons they were not only top five in defensive plays, but they were middle of the road in all shooting percentages. This season, they've added Elton Brand's 1.4 steals and 1.7 blocks to their defense (essentially Brand's 3.1/40 minutes is replacing Reggie Evans' 1.7/40 minutes). Brand has elevated them from a top-five team in this category to number one, and there's no reason, absolutely no reason, they shouldn't be able to keep the field goal percentages in line with the previous two seasons considering their defensive personnel. In fact, I'd make the argument that they've made significant defensive upgrades by adding Brand and Jrue, and any reasonable coach would be utilizing those guys for much heavier minutes than Jordan has.

I've probably droned on for far too long, so let me sum it up as quickly as possible. Right now, today, the Sixers have the personnel to turn this around, and turn it around in a big way. They're getting severely abused on the perimeter, they play this wacky rotational scheme or bogus matchup zone to cover for the deficiencies of Lou Williams and Allen Iverson in the back court, but they really don't need to.

If they started Brand at the four and played him 35 minutes/game, he'd be making more defensive plays to fuel the transition game. If you played Brand 35 minutes/game at PF, Thad would rarely be playing the four, he'd rarely be at a size disadvantage on the defensive end. Most importantly, if you played Brand at PF for 35 minutes/game, you'd never, ever have to have two defensive liabilities playing in the back court at the same time. Lou and Iverson could split minutes at one guard however you like, with either Jrue or Iguodala always on the floor with them in the back court. This would allow you to use a superior perimeter defender on the other team's biggest penetration threat, thus avoiding the defensive breakdowns that decimate this defense.

If you sat Eddie Jordan down and had an honest conversation with him about the matter, I believe he'd tell you Brand can't play the four because he isn't mobile enough to get out and defend the stretch fours in the league. Is this a legitimate concern? Yes. It's a legitimate concern against the Orlando Magic (Rashard Lewis), the Indiana Pacers (Troy Murphy), the Dallas Mavericks (Dirk), the Washington Wizards (Antawn Jamison) and the Minnesota Timberwolves (Kevin Love). Those are the only five teams in the league who play a stretch four who shoots above league average on three pointers. I don't really care if any big is taking 21-foot jump shots regularly. If you get them to, it's a win defensively. The only way the stretch four really kills you is with threes and we're talking about 13 games/season where it's an issue. What that doesn't take into account is that not a single one of these guys can guard Brand on the blocks on the other end.

Put as simply as I can, I believe if the Sixers were to fire Eddie Jordan tomorrow, bring in someone who would make no changes but to simply play the proper guys at the proper position, we'd be an average defensive team, at worst. If we brought in a guy who not only played the right guys, but completely scrapped Jordan's system and replaced it with a more honest, straight-up, pressuring (rather than reacting) scheme, we'd be a top-ten defensive team and a team that would only get better as Holiday matured into his eventual role as a 35+ mpg defensive stud.

Thoughts, questions, concerns, arguments in the comments, as usual.