DFDepressed FanDepressed Fan



, all the time

The Disruptors

Defense wins. Defend to run. The best offense is a good defense. We've heard plenty of platitudes regarding defense, some are true, some are merely motivational tools to get guys to play on the end of the floor that isn't recognized or rewarded nearly enough in terms of dollars and recognition. After the jump, we're going to spend some time looking at my favorite kind of defender, the disruptive force.

(A quick note before we get started. If you're into stats, I've uploaded the spreadsheet I used for all the charts in this post, you can download it here or at the bottom of the post. I'll also explain my methodology down there.)

Defense is horribly difficult to quantify using statistics. I've yet to see an all-encompassing metric that I fully trust. Typically, when I'm looking for information as to whether a player is a good, bad or indifferent defender, I check three or four sites for various stats (including PER against, DFR, counting stats, etc), and then I apply my own eyeball test. For the most part, I'm pretty confident that I can judge whether a player is a good defender or a poor defender from this method, but there are so many layers that most of the time it's a hard argument to make when you have a difference of opinion with someone over a certain player's defensive production (Synergy Sports helps a ton with these arguments, btw).

This post isn't about overall defensive ability, it's about looking at individual defense on a different level. The ultimate defensive possession is zero points for the opponent, and a transition hoop on the other end of the floor. A four-point swing, completely fueled by a defensive play. Second best, is zero points for your opponent, plus a personal foul on them, in other words, a charge drawn. The basis of this post is a simple counting stat, Defensive Plays. It's simple, steals + blocks + charges taken = defensive plays. (The stat is available from HoopData). In a vacuum, steals and blocks don't tell a complete story. A player who racks up the steals could be gambling and leaving his man to get them, meaning open shots. Blocked shots aren't always recovered by the defensive team, meaning there are times when they don't end a possession. That being said, the stats themselves are meaningful for my purposes. Even if a guy gambles for steals and gives up a ton of easy looks and points, the steals that he actually gets still turn the ball over and end the possession. Every successfully blocked shot changes the possession, and the mindset of the opposing team, even if it doesn't end a possession. When we're talking about meaning, if the defensive play happens, it matter. That's black and white. It's the overall impact of players attempting to make these plays that's more of a gray area, and this is something we'll touch on later.

Let's get into the numbers now. Here's a look at the Sixers roster from last season (minus Iverson, Primoz Brezec and Francisco Elson, because they won't be back and I really don't care about them. I've also added in the new additions to this year's roster.)

The red column is the pure counting stat. As you know by now, counting stats are a horrible way to look at numbers, though, so we're going to dig deeper. The simple way to add a bit of meaning is to turn the counting stats into a per minute stat, or per-36 stat. Essentially, if this guy was playing starter's minutes, how many of these would he have? This method is somewhat useful, but only to certain extent. Instead, I used a pace adjustment (explained at the bottom of the post), to come up with a more solid per-36 stat, the green column.

Now that we have workable numbers for how many defensive plays each player could be expected to make, we have to add a bit more depth to the analysis. Position matters. It matters a whole lot. Centers are in position to block shots. Guards are more likely to get steals. It's obvious that Dalembert (3.49), created more defensive plays than Willie Green (1.24), but how many more should he be expected to create simply because he's a big and Willie's a small? And more importantly, how does Dalembert compare to other centers?

The next step is to set a positional baseline. What I did was take the top 20 players at each position with the most defensive plays (raw number) on the season, then ran their numbers through the spreadsheet. I added the Sixers in (along with a couple of player seasons I found interesting), then came up with an average based on these numbers. This isn't a pure average, it's an average of the players who made the most defensive plays, simply because I didn't have the bandwidth to pull league-wide numbers in. Let's start with point guards and move from small to big:

Ivey played extremely limited minutes, so I wouldn't much into his standing atop this list. It's safe to say he was a disruptive player in his limited role off the bench. The names down to Rondo don't really surprise me, but Kyle Lowry does. 51 charges drawn is a huge total for a guy who played less than 1,700 minutes. Jrue was middle-of-the-road among points.

A few things stand out to me among the SG data. First, Dwyane Wade is simply a beast. Second, seeing Battier up there in the rankings makes me feel good about the data, though I'm not exactly sure why he's listed as a shooting guard. I included three seasons of Iguodala here, two of which he measures up very well. In the third (08-09), he was merely above average. Lou is slightly above average, with Meeks falling woefully below average and Willie Green measuring out as atrocious (which is accurate). I was shocked to see Joe Johnson ranked so low. Another reason his $120M contract is simply absurd to me.

Gerald Wallace is an animal (more on this later, but check out the pace adjustment). AK47 was born for this stat. Granger is a surprise to me. The Sixers really perform poorly at this position (unless you count Iguodala at SF, where he measures up quite well). Nocioni and Thad are nearly as disappointing as Kapono, simply because both of those guys play big minutes at PF, a position where opportunities are much more plentiful to make these plays. 1.08 for Kapono, if that doesn't make you laugh, check your pulse.


Some real studs in this group, led by Tyrus Thomas (more on him later), Josh Smith, who is by far the best in the league, in my opinion, for a guy who plays heavy minutes. Camby definitely isn't a power forward, and should be grouped with the centers, where he still dominates, and I think Ibaka is probably more 5 than 4 as well, but his numbers are excellent. Jason Smith was very active in his limited minutes, Brand was above average (but you have to take some value away because he did play a lot of center last season), and I threw in Reggie Evans in his last season with the Sixers. I had a feeling he wouldn't do well, simply because he never blocks shots and gets called for blocks much more frequently than he draws charges. Evans' 1.87 was well below average for the position. LaMarcus Aldridge's place at the bottom of the list didn't surprise me at all, and he played a decent amount of time at center last season to boot.

Good for JaVale McGee, if he can stay on the floor this year, he can help the Wizards. Same goes for Oden, but what are the odds of that? Bogut is the biggest surprise to me, unbelievable year for him. Dalembert was well above average, as you probably could've guessed. Speights is also above average, thanks mostly to the charges he drew. Looking forward, Hawes and Battie were both well below average, but nowhere near as bad as David Lee who is approaching Willie Green/Jason Kapono status in this area.

Now that we have a baseline by position, let's take a look at how the Sixers stack up:

Dalembert and Iguodala performed the best against positional averages, with Hawes, Meeks, Green and Kapono all measuring out horribly. Not exactly a formula for building a team around defending to run, huh? Especially since Dalembert (+0.61) has been replaced by Hawes (-0.50). Another reason why I believe Speights should be starting at center, but that's an argument for another day.

We aren't done adding context to this analysis, though. Even after we've taken pace of play into account, we still haven't accounted for defensive systems. I'm going to show you two extremes now. Up first, is the Golden State Warriors.

The Warriors led the league in pace last season (100.3 possessions/48 minutes, league average was 92.7 possessions). Their entire philosophy seems to be to devalue individual possessions, really on both ends of the floor. They ranked in the middle of the league in offensive efficiency, and 29th of 30 teams in defensive efficiency rating. Turiaf, Randolph and a couple of other guys look awfully good using this metric, but I can't help but think it's fool's gold where they're concerned. Teams who aren't accustomed to playing fast wind up playing sloppy. A team defense based on chaos is going to create more opportunities for defensive plays. I just don't think the numbers will translate into a more sound defensive system, and we probably won't get a chance to measure that this season with Randolph and Turiaf in New York playing in a similar system.

On the opposite end of the spectrum falls the San Antonio Spurs:

Ginobili is the only regular who measures out as above average, does that mean the other Spurs simply can't make defensive plays? No, I don't think so. Gregg Popovich's defense is not predicated on forcing the team into turnovers, it's built to do one simple thing. Make the other team take bad shots. They play solid, fundamental, positional defense, and instead of gambling for a steal, they stay at home and make sure holes don't open. They push players into zones for low-percentage jumpers. They close off the lane and they stifle teams with a merciless devotion to those principles. Take Manu and put him on the Warriors and he may produce like Wade. Put a guy like Randolph on the Spurs and he probably never sees the floor. The system can matter that much.

Obviously, this metric is by no means a definitive way to look at defenders, nor is it an open-and-shut case for a player's value. At its core, it's a mathematical way to affirm whether a guy, "has a nose for the ball." How would I use it if I was an NBA decision maker? Well, I guess that's the question.

Personally, I love the transition game, and I love the transition game that's sparked by defense. I love players who can turn a game with a couple of well-timed defensive plays. On a much smaller scale, a steal that leads to a layup on the other end is like an interception returned for a touchdown in football. When you're talking about the elite players, I like to see this type of production as an ancillary indicator, one of the tools in the toolbox. The great ones usually have it.

When you're talking about rounding out your rotation, though, I think this may be a stat that's undervalued, or at least they're under appreciated. This exercise gave me a newfound appreciation for a couple of players. Foremost among them is Tyrus Thomas. Thomas was truly elite in his defensive playmaking production last season, and he created all of these defensive plays in a slow system down in Charlotte and Chicago (a bit faster than average). Thomas isn't a terribly efficient scorer, but he is a solid defensive rebounder. Whether or not he can play 30+ minutes/game remains to be seen, but I think he's worth the gamble. Unfortunately, Charlotte opened the piggy bank to re-sign Thomas this summer, but if I had to bet, I'd say he'll be on the trade market sooner rather than later. I'm definitely going to keep an eye on his situation, because I'd love to have him on the Sixers and even if he winds up being the first big off the bench, I think he'll more than earn his contract with his defensive work.

A couple other names I like:

  •  Jared Jeffries - He's a bit on the old side, and really gives you nothing on the offensive side of the ball. But when this ridiculous contract expires after next season, he may be a bargain worth looking into on the free agent market. The guy makes plays and can defend multiple positions.
  • Nick Collison - Ditto Jeffries in most ways, not much of a rebounder for a center, but does make plays. Again, if he can be had on the cheap, he's going to make plays for you. Also a free agent after this season.
  • Luc Mbah a Moute - Another free-agent-to-be, Luc may be the least talented offensive player of the group, but he'll get the ball for you.
  • Corey Brewer - He won't hit free agency for two years, but I love his defensive playmaking. If he can continue to improve on his perimeter shooting, he probably won't be a bargain once he becomes a FA. I'd look to target him in a trade.
  • Josh Smith - This is pretty obvious, and i doubt Atlanta would be willing to move him any time soon, but when ownership changes and they start looking at how much they're paying Joe Johnson, Smith could become available. He'd be the ideal defensive PF compliment to Jrue, Iguodala and Turner.

Those are just a few names that would move the needle in the right direction as far as defensive playmaking goes. As the roster stands right now, well, I'm not sure how much defending and running they're going to be doing. It's a legitimate concern. I'm just not sure how many opportunities the team will create. They're going to have to rely heavily on their back court to set the tone and make the plays.


I pulled the defensive plays stats from HoopData, uploaded them to a spreadsheet, then looked up possession numbers for each player on Basketball Value. Once had all the raw data in one spreadsheet, I first averaged the offensive and defensive possessions for each player, for a more accurate measure.

(Offensive Possessions + Defensive Possessions) / 2

Then I calculated the player's individual pace

(Possessions / Minutes played) * 48

Next I figured out the pace adjustment for each player. League-average pace is 92.7 possessions per 48 minutes.

Player's individual pace / 92.7 = Pace adjustment
From there, I figured out the player's defensive play rate (per 48 minutes) adjusted rate per 48 minutes, and then adjusted rate per 36 minutes.

{[(Defensive Plays / min)*48] * Pace Adjustment} * .75

There's a shorter route to the number, but I wanted to look at the numbers each step of the way. There's also a simpler way to adjust for pace.
Defensive Plays / Possessions = % Defensive Plays

This method would be in keeping with STL% (BLK% is actually calculated using the number of two-point FGA when a player is on the floor), but I find it easier to digest basketball stats in terms of per/36 numbers. I feel like I can wrap my mind around 3.07 DEF/36 better than a 4.4% DEF play rate for Dwyane Wade. I'd wind up doing the math in my head anyway, so I'm just removing the extra step and keeping the data in a form I'm comfortable with.

If you're interested, download the spreadsheet here. Any questions, comments, concerns or disputes, let me know in the comments.
by Brian on Aug 10 2010
Tags: Basketball | Defense | Offseason | Sixers |