Take a gander at the Sixers team page on
82games.com, I'll wait. Two particular numbers jumped out at me, opponent's production vs. Lou (17.2) and vs. Jrue (19.9). 15.0 is supposedly league-average, the lower the number, the better. Having watched every second of Sixers basketball this season, I can emphatically tell you Lou Williams isn't half the defender Jrue Holiday is, so how do you explain that number? We'll give it a try after the jump.
Obviously, I'm using Synergy Sports for the following numbers. If you don't have an account yet, I highly recommend it. Anyway, the first thing I did was take a look at the general defensive numbers for each player. Lou allowed 0.96 points-per-play, his man shot 42.3% from the floor, 40% from three, he forced turnovers on 7.3% of plays. Jrue allowed 0.93 points-per-play, his man shot 40.2% from the floor, 35.9% from three, he forced turnovers on 8% of plays. Jrue's foul rate was significantly higher (7.8% to 4.2%). Overall, Jrue's numbers are more impressive, but not drastically so.
When you look a little closer, Jrue holds a significant advantage in every defensive situation (Post-up, P&R Roll Man, Spot-Up, Off Screen, Hand Off) but one (Isolation). Jrue allowed 1.13 points-per-play on isolation plays, Lou allowed an impressive 0.6 points-per-play in similar situations.
82games.com is intentionally opaque in their methodology for calculating opponent's production, but I have to believe they're weighting isolation defense much higher, probably because it's necessarily the easiest to isolate. I figured since this was the one area Lou was superior in, we should take a closer look. The charts below depict first who Jrue was guarding for the 124 isolation plays Synergy has video of. The second is the same list, but for the 90 isolation plays they have video of for Lou.
Jrue Holiday - Isolation defense
| Player |
No |
Player |
No. |
| Derrick Rose |
9 |
Rodney Stuckey |
1 |
| Monta Ellis |
8 |
Jose Calderon |
1 |
| Jameer Nelson |
7 |
Kevin Durant |
1 |
| Dwyane Wade |
7 |
Rashard Lewis |
1 |
| Mo Williams |
5 |
Keyon Dooling |
1 |
| TJ Ford |
5 |
Terrence Williams |
1 |
| Stephen Curry |
5 |
Earl Watson |
1 |
| Mike Conley |
4 |
Paul Pierce |
1 |
| Jamal Crawford |
4 |
Joe Johnson |
1 |
| Jarrett Jack |
4 |
Jason Williams |
1 |
| Rajon Rondo |
4 |
Jordan Farmar |
1 |
| Kobe Bryant |
4 |
CJ Watson |
1 |
| Jonny Flynn |
4 |
George Hill |
1 |
| Aaron Brooks |
4 |
Manu Ginobili |
1 |
| Russell Westbrook |
4 |
Jason Kidd |
1 |
| Toney Douglas |
3 |
Jerryd Bayless |
1 |
| Ben Gordon |
3 |
Tyreke Evans |
1 |
| Mike Bibby |
2 |
Chris Paul |
1 |
| Kirk Hinrich |
2 |
Chucky Atkins |
1 |
| Danilo Gallinari |
2 |
Gilbert Arenas |
1 |
| Raymond Felton |
2 |
Nick Young |
1 |
| Mickael Pietrus |
2 |
Arron Afflalo |
1 |
| Baron Davis |
2 |
Rudy Gay |
1 |
| Brandon Jennings |
1 |
Goran Dragic |
1 |
| Carlos Arroyo |
1 |
Lester Hudson |
1 |
| Will Bynum |
1 |
|
|
Lou Williams - Isolation defense
| Player |
No. |
Player |
No. |
| Tyrke Evans |
7 |
Troy Murphy |
1 |
| Aaron Brooks |
6 |
Earl Watson |
1 |
| Rodney Stuckey |
5 |
Paul Pierce |
1 |
| CJ Watson |
5 |
Marvin Williams |
1 |
| Mike Conley |
4 |
Steve Nash |
1 |
| Earl Boykins |
4 |
Goran Dragic |
1 |
| Delonte West |
3 |
Anthony Morrow |
1 |
| Rajon Rondo |
3 |
George Hill |
1 |
| Derek Fisher |
3 |
Kevin Love |
1 |
| Derrick Rose |
3 |
Keyon Dooling |
1 |
| Baron Davis |
3 |
AJ Price |
1 |
| Brandon Jennings |
3 |
JJ Barea |
1 |
| Terrence Williams |
2 |
DeMar DeRozan |
1 |
| Jarrett Jack |
2 |
Gilbert Arenas |
1 |
| Monta Ellis |
2 |
Randy Foye |
1 |
| Jonny Flynn |
2 |
Anthony Carter |
1 |
| Wilson Chandler |
2 |
Arron Afflalo |
1 |
| Ty Lawson |
2 |
Sergio Rodriguez |
1 |
| Mo Williams |
2 |
Nick Young |
1 |
| Jason Williams |
1 |
Andray Blatche |
1 |
| Will Bynum |
1 |
Marcus Williams |
1 |
| LeBron James |
1 |
Eric Maynor |
1 |
| DJ Augustin |
1 |
Chris Duhon |
1 |
Those are too many names to easily digest, so I'll break it down for you:
Of the 126 isolation plays against Jrue, 69 of the plays were against opponents that I'll call upper-echelon offensive players, and I'm being fairly liberal with that classification. Here are the players I'm lumping into that category: Derrick Rose, Jameer Nelson, Monta Ellis, Dwyane Wade, Mo Williams, Stephen Curry, Rajon Rondo, Kobe Bryant, Aaron Brooks, Russell Westbrook, Ben Gordon, Brandon Jennings, Kevin Durant, Paul Pierce, Joe Johnson, Manu Ginobili, Jason Kidd, Tyreke Evans, Chris Paul and Rudy Gay. Jrue was defending that group of players for 56% of his defensive isolation plays.
Of the 90 isolation plays against Lou, 27 of the plays were against opponents who fall into the same group of upper-echelon players. (Tyreke Evans, Aaron Brooks, Rajon Rondo, Derrick Rose, Brandon Jennings, Monta Ellis, LeBron James, Paul Pierce, Steve Nash). Lou was defending that group of players for only 30% of his defensive isolation plays.
So, was Lou better in these situations, or was it merely a matter of Lou being gifted easier matchups, or being hidden as the case may be? Nothing is definitive, but I think the evidence is clearly pointing in that direction.
A couple of notes from the video I watched. If you do sign up for Synergy, check out these two games for Jrue:
- 12/3 @ OKC - OKC isolated Russell Westbrook on Jrue 4 times. 3 missed jumpers and 1 turnover.
- 12/15 vs. GSW - GSW isolated Monta Ellis on Jrue 7 times. 4 missed jumpers, 1 blocked layup and 2 turnovers.
One final note, this data is useful, and it's interesting to go back and see the baptism by fire Jrue underwent in his rookie season on the defensive end, but by no means is this a complete picture. This version of Synergy only records plays that end in a field goal attempt, a shooting foul or a turnover. In most of these situations, good defense is more about stopping penetration and forcing the ballhandler to give it up and reset the offense. There's no way for me to track those situations.
The one takeaway I have from this exercise is that Lou's numbers probably don't mean a whole lot. 70% of the time, he was guarding a lesser offensive talent. The fact that these lower-echelon players were going at him is both a byproduct of the defenders on the floor with him (Iguodala and Jrue, especially) and actually it's sort of a sign of a win for team defense. Think about it, when you're playing the Golden State Warriors, don't you want CJ Watson taking shots off the dribble as opposed to Stephen Curry or Monta Ellis? Earl Boykins, Derek Fisher and a host of others fall into the same category.
Looking forward, if/when we have a decent defensive mind running this team, and hopefully a defender who is fundamentally better than Lou, teams should be forced to run offense through their second and third perimeter options and it will have a huge deleterious effect on their overall offensive efficiency. Or at least that's the theory I'm going with.
Thoughts in the comments, as usual. Happy Monday.
I think the fact it's clear as a day that Jrue is a far better defender than Lou. The thing that is interesting to me is that despite guarding a lesser talent Lou posted decent defensive numbers (he is going to guard the worst perimeter threat anyway). 0.96 points per play is something i can happily live with from my worst defender in the lineup. Brian don't you think Lou may actually be worth getting 30min per night for his offensive production? From these numbers he actually looks halfway decent. Can he even improve on that with a better system and a better coach?
Jrue has better physical tools to defend and he was only in his rookie season, where as Lou has five years of experience.
Of the upper-echelon players you mentioned being guarded by Jrue, ten were SGs/SF on 28 plays, where as Lou guarded a total of nine isolations from three SG/SF's (I'm counting Evans as a SG since he played 37% of his minutes at that position as opposed to 28% at PG).
This tells me that Jrue is responsible for taking the bigger player, and often times, much bigger player. This is a defensive liability for the team and thus, I'm even further convinced that Lou should be limited to a bench role.
I don't have stats to back it up, but that's why I like a backcourt of Jrue and Andre. That pair take away easy points for opponents guards. If Thad can learn to guard (given proper coaching & motivation) they may be able to win a lot of games mainly because they can guard the perimeter in a league where the perimeter is increasingly important.
Whatever Lou!s role is going forward, I like the fact his numbers weren!t as bad as I thought they would be. He is a dangerous, unguardable player at times and fits our style of play so unless we get value back lets keep him.
Brian, with your new toy I would love to see how Lou stacked up offensively against the top shooting guards in the league. We all seem to value Thad more than Lou but he made major strides across the board offensively and whether Iggy/Lou or Thad/Iggy works better at the 3/2 is still debateable.
Yup, context is the key in defensive stats, and it's very tough to measure.
Looking at that, how impressive is Iguodala's 14.6 opponents PER?
Do you think last season Iguodala returned to being a top perimeter defender? Because I think he had taken a step back the prior season as a result of having to work so hard on offense.
Could relying more on Jrue to create things help Iguodala on both ends? Leading to better shot selection and more left in the tank to give on the defensive end?
I ask because most players at 26 are at their peak- so Iguodala might be seen as not having much more upside... But Iguodala could become much more effective and efficient if he was utilized in a role that played to hos strengths.
Actually, I thought Iguodala was even better defensively in 08-09, and the stats back it up (his 13.0 PER against was top 5 in the league, as I recall). This past year, his man scored a lot on open 3's as he would rotate and try to recover, in their crazy defensive scheme. I don't recall that happening as much last year.
It was in 07-08 and 06-07 that his defensive numbers were less impressive (though he did set a career high in steals in 07-08).
It's flat out mind - blowing, especially considering he is guarding the best opposing player unless he is a PF/C.
Unbelievably impressive. He basically took the best player on every team and held them to below-league average production. He also guarded 4 positions pretty regularly, depending on who the best player on the other team was. Very impressive numbers.
Haven't you been paying attention he's a gorssly over paid sucky player
The strength of Jrue-Dala is you can hide the defensive weaknesses of the 3rd backcourt player (Lou/Carney/Kapono/Young.) It gives the team some flexibility at that spot, and maybe helps a talented rookie make an easier transition to the pro game (Johnson?)
In terms of man defense, there is no way to consistently shut down top offensive players. All you can do is disrupt them and make them work. Ideally, you can break the other team's flow enough to hurt their overall efficiency.
For example, if the opponent always has to send their big to set picks (and you defend it properly) you have a better chance of controlling the boards. You also turn teammates into spectators if the best player has to spend extra time in breakdown mode. After watching all of those years of AI, stars are at their best when they are scoring quickly and within an offense- break that rhythm and you can win even if the still "get theirs."
btw, I LOVE the picture of Lou in his defensive stance. cracks me up every time I see him do it.
I've never seen a player whose def stance looks like his back went out once he crosses half court. Just kills me
One interesting stat that might be of some indication how good a defender a player is, is the difference between the guarded players average PER and the PER that they posted on the particular player. For example Iggy's opponents PER is 14.6. what is their average PER against any defender during the course of the season? How much has Iggy actually disrupted their play.
Does anyone know if these kind of stats can be obtained from a site?
[quote]One interesting stat that might be of some indication how good a defender a player is, is the difference between the guarded players average PER and the PER that they posted on the particular player. For example Iggy's opponents PER is 14.6. what is their average PER against any defender during the course of the season? How much has Iggy actually disrupted their play.
Does anyone know if these kind of stats can be obtained from a site?[/quote]
dMult (basketball prospectus) attempts to do that.
Thx this is very helpful indeed!!!
Williams is a third guard capable of getting hot and scoring points off the bench.