March 6, 2010

The Eyes Tell Lies

Forgive me, dear reader, for I have sinned: I cannot make myself watch the Bulls' 105-96 loss to Memphis on Thursday.

Because of familial responsibilities, I was not able to see the game live. Of course, I recorded it and it's sitting on my DVR. But do I really want to force myself to stomach two-and-a-half hours of blown-17-point-lead agony? Sure, if I watch the game I'll usually re-watch it before writing. But since I didn't have to suffer through the disappointment the first time through, why would I want to do it now?

Cogent analysis, I suppose.

But who needs that? I refuse to suffer through the indignity of another soul-crushing loss. Instead, I am going to do the unthinkable and write this entirely blind. This might seem insane to the demographic of statistic-eschewing eye-believers, but one fact remains: while numbers don't lie, our eyes most certainly do.

Our vision is not a security camera, merely recording everything in front of it (most notably, like that time I crotched that pack of Big League Chew). Rather, everything is filtered through our brain. And when we watch, more often than not what we see becomes ensnared in a web of confirmation bias, which is a tendency to interpret information in way that confirms one's preconceptions.

Look no further than the case of Tyrus Thomas.

I loved Tyrus. What I'll remember most about his Bulls career are the flying dunks, the uncannily-timed blocked shots, and a much-improved mid-range jumper. However, plenty of people watched the exact same games I did, and their overarching perception of Tyrus is centered around play that was alternately out-of-control and indifferent, poor shot selection, and repeated failures to rotate defensively.

The thing is, Tyrus did all of those things, albeit to differing degrees. And I think most fans would acknowledge that. But depending largely on what you had decided Tyrus was -- either a waste-of-a-lottery-pick malcontent, or a guy who just needed consistent playing time -- you primarily perceived the moments that would confirm it.

Therefore, if you want analysis in its purest form, ignore them lyin' eyes. With that as my motive -- and due in no part to my overwhelming laziness -- I have the following observations about Thursday's game, based exclusively on the box score and play-by-play data:

The Bulls frontcourt can't stop anybody

Zach Randolph's night: 31 points (12-18 FG, 7-8 FT), 18 rebounds.

Hard to believe that this is the guy that the Grizzlies were roundly mocked for taking off of the Clippers hands this offseason, and the same one that the Bulls, even given their dearth of low post scoring, showed absolutely zero interest in acquiring at several times over the last few seasons.

Randolph was just the latest power forward to have his way with the Bulls. Over the last six games, the Bulls have yielded the following to 4s around the league*:


Blatche Aldridge Smith Randolph AVG
Min 41 44 34 40 39.8
Pts 25 32 17 31 26.2
FG 8-13 15-23 7-14 12-18 10.5-17 61.8%
FT 9-9 2-2 3-4 7-8 5.3-5.8 91.3%
Reb 11 7 18 18 13.5
Ast 2 0 4 1 1.8
Stl 2 1 2 1 1.5
Blk 2 0 2 0 1
TO 3 3 3 2 2.8
PF 2 4 4 2 3

* Luckily, there were two games involving Troy Murphy sandwiched in there, so they can kind of stop somebody who is eminently stoppable.

And this is the downside of the Joakim Noah injury, combined with the Thomas trade.

Until Noah's injury, the Bulls frontline had three of the 20 best shot blockers (on a per-minute basis) in the league. Now they are playing without the better two of those three, leaving Taj Gibson as the only shot-altering threat they have. Luol Deng's 59th-best rate (1.22/48 min) is second highest among healthy Bulls, and less than a third of Tyrus' (3.82).

While the play of Brad Miller in Noah's absence was significantly acclaimed -- and, just as significantly, has dropped off of late -- probably the most underrated reason the Bulls were able to initially weather the storm of Noah's injury was the presence of Thomas, who defensively does a lot of the same things. With Thomas now gone as well, that leaves only Gibson and his still too foul-prone ways defending the basket. Which is why we're seeing a whole lotta 4s go off on the Bulls.

Piss-poor second quarters are destroying the Bulls chances

Against the Grizzlies, the Bulls 13-point first quarter lead had shrunk to 6 by halftime. Which continued a disturbing trend: somehow, dating back to last Wednesday's Pacers game, the Bulls have been outscored by at least five points in five straight second quarters:


Q2 Score
vs. Indiana 36-21
vs. Portland 29-21
at Indiana 29-23
vs. Atlanta 25-20
vs. Memphis 26-19
AVG 29-21
Per 48 116-84

What's happening here is as clear as it is frightening: The second quarter is when the backups play the most, and the Bulls bench, the team's strength as recently as last year, is now a glaring weakness. Obviously, having Noah out hurts here too, but the problem runs deeper than that. The bench is now comprised primarily of two guys:


Hakim Warrick




vs. IND POR @ IND ATL MEM AVG
Min 15 34 17 25 25 23.2
Pts 5 15 4 10 6 8
FG 1-3 6-11 1-3 3-10 1-3 2.4-6 40%
FT 3-5 3-3 2-2 4-4 4-5 3.2-3.8
84.2%
Reb 1 4 4 7 5 4.2
Ast 0 3 0 1 0 0.8
Stl 1 0 1 0 0 0.4
Blk 0 0 0 1 0 0.2
TO 1 1 2 1 0 1
PF 0 3 2 2 1 1.6
+/- -2 -4 -3 -5 -13 -5.4

and


Flip Murray




vs. IND POR @ IND ATL MEM AVG
Min 33 21 19 31 19 24.6
Pts 16 8 2 12 4 8.4
FG 4-12 2-5 0-2 6-13 1-7 2.6-7.8 33.3%
FT 8-10 4-4 2-2 0-1 2-2 3.2-3.8 84.2%
Reb 6 3 2 3 2 3.2
Ast 0 1 0 2 2 1
Stl 3 0 0 3 0 1.2
Blk 1 0 0 1 0 0.4
TO 3 1 1 2 2 1.8
PF 3 1 3 1 2 2
+/- 3 7 -12 -5 -11 -3.6

Warrick and Murray essentially replaced two players, Thomas -- who had a loyal and vocal following -- and John Salmons, who didn't. And Murray and Warrick's collective shooting slump presents a huge problem, because while both were supposed to bring improved scoring, they are significant downgrades defensively.

The Bulls have tried to sell a Warrick-as-superior-defender-based-on-positioning line of bullshit, but the numbers tell a different story. According to baketballvalue.com, with Warrick on the court, the Bulls have a defensive rating (points allowed per 100 possessions) of 110.51. Without Warrick, they are at 103.83. That's a net of +6.68 (a positive number is not a good thing, in this case).

As for Murray, the Bulls are at 109.77 with him; without him, 103.89. He is then at +5.88.

Admittedly, there are issues with the data sets. For starters, as with all +/- stats, there are issues of multicollinearity, which I'll let ESPN.com's John Hollinger define:

Since the same players tend to play together most of the time, it's tough to tease out to what extent each is impacting the unit's results unless you have a massive number of observations.

Basically, if every time Bill Russell came out of a game, Bob Cousy did too, Cousy would have an amazing +/- rating which he (presumably) would've done little to merit. But the problems with Warrick's and Murray's numbers go beyond multicollinearity, because their off-court numbers include all of the Bulls data from the time before they were acquired. And most of that data came when the Bulls had the benefit of Joakim Noah's services, and he is arguably their best defensive player.

Or is he? Over the course of the season, the Bulls are at 105.46 with Noah on the court, and 103.01 with him off, for a net of 2.45. That shocked the hell out of me.

Still, because the sample size of on-court data is so small, I wanted to look at Murray and Warrick's numbers from before the Bulls acquired them. With Warrick on the court, the Bucks were at 105.82; without him they're 102.05, which is the kind of thing likely to happen with a power forward averaging 0.6 blocks (and 0.7 steals) per 36 min over the course of his career.

Besides, should we really be surprised that Warrick's defensive prowess was greatly exaggerated? If Warrick were an ace defender, would he really have fallen out of favor in Milwaukee with defensive-bug-up-his-ass Scott Skiles as their coach? If Warrick had some magical defensive ability that didn't show up on the stat sheet, I'm confident that Skiles would've sussed it out; instead, he had his minutes significantly reduced over the course of the season.

Murray, on the other hand, looks a little more competent in his Bobcats context. Charlotte was actually a better defensive team with him on the court, with a rating of 102.26, as opposed to 104.64 with him off. While his numbers are more encouraging, overall the swap of Warrick-Murray for Thomas-Salmons has not been a good one defensively. Because among Bulls who have played any real minutes, Thomas (-3.36) and Salmons (-4.60) had the two best net defensive ratings on the team.

Let me say that again: The two best.

Salmons was very long for a shooting guard, so while it's surprising he would've fared so well, it's not a total shock. As for Thomas, I suppose that maybe this means his positives as a defensive player -- primarily, his shot-blocking and ability to pick up steals -- outweighed his perceived failures.

Despite what your eyes may have told you.

No comments:

Post a Comment