Wednesday, January 20, 2010
The Grizzlies' Stat Geek Speaks
By Henry Abbott
Joe Murphy/NBAE via Getty Images
The stats say Marc Gasol and Zach Randolph are both much improved.
People have been noticing that the Grizzlies -- among the NBA's best over the last two months with wins over the Cavaliers, Spurs, Mavericks, Nuggets, Suns, Blazers and the like -- are a new force to be reckoned with in the competitive West. And there are many good theories as to why. Marc Gasol has lost weight. Rudy Gay is back on a path to stardom. Zach Randolph is putting up All-Star numbers. O.J. Mayo is a tenacious defender who can score at will. Every starter but Mike Conley demands double-teams. Lionel Hollins is steering the ship better than ever ...
One of the more off-beat theories came from stats expert Wayne Winston the other day, who says not to forget to credit Aaron Barzilai.
Who's that? He's the Grizzlies' quantitative analyst, newly on board this season. Barzilai -- the founder of BasketballValue.com -- advises GM Chris Wallace and the Grizzlies coaching staff. Instead of drawing on a playing, coaching or scouting career, he employs his MIT degree in engineering (where he did, in fairness, play on the basketball team) and a Ph.D. in engineering from Stanford. For most of his adult life he has been supporting himself doing quantitative analysis in sales and marketing. I spoke to him yesterday.
Tell us what you do for the Grizzlies?
What I do is a little bit of everything. Mostly I work with Chris Wallace about various things that are on his mind. Questions he has. Ideas. Maybe things I've noticed.
I started off looking mostly at player personnel, but I've started working a little bit more with the coaches, mostly just taking a look at the different lineups, and identifying potential opportunities to tweak our present rotation a little bit here or there. Nothing wholesale.
Wayne Winston's always telling us this lineup's amazing! Or this lineup's terrible! Your advice would be roughly along those lines?
Yes. In the same spirit of noticing stuff like "hey our starters are playing really well. Maybe this other lineup isn't working so well for whatever reason, maybe we should play them a little bit less." Similar to Winston we're looking at which players, pairs or combinations are working or not working.
NBA teams have all kinds of insight, right? From scouts. From their own coaches. Contacts in the business. Now there's a new group of people, people like you, with insight. How does it all get assimilated? How is it received? Do they say: Gosh, great, now we have 6,000 different lineup recommendations. Or do they say hey, cool, we'll do that tonight.
Umm, no, I'm just starting out. This is my first year. We're only halfway through the season. Like anyone on a new job I think, your influence is kind of limited. Since I work remotely it's mostly through e-mails and text messages. Sometimes I'll notice things on the court that maybe are along the lines of what I had suggested, but to be honest, I'm not certain if it's because they had noticed the same things as well, or if it was based on my suggestion.
Wayne Winston said something to me about how we have to credit you with the Grizzlies playing well. Should we be doing that?
Not at all. Since I'm just starting out I'd imagine I've had very limited influence over anything. The players clearly put in time over the summer across the board. I think that also speaks well for the coaching staff. I'm hoping at some point in the future I'll feel like I deserve some small amount of credit, but that's premature at this point.
Nevertheless, the team is playing unbelievably well. Over the last two months they're basically one of the elite teams in the West. Everyone has always made fun of the franchise. The Pau Gasol trade was laughed off as one of the worst in the NBA. Zach Randolph and Allen Iverson were the big acquisitions of the last off-season. Almost nobody liked either move. Nevertheless, Rudy Gay looks tremendous. O.J. Mayo looks tremendous. Marc Gasol looks tremendous. And they're beating everybody. With your insight: How is this happening? What's making this team so good?
It's an interesting question. A lot of people are playing better than they have in the past. Marc Gasol is clearly playing better than he has in the past. Zach Randolph is playing great, and so I think a lot of people have noticed that, but in reality he has always been almost this good. This may be his best year, but actually I think statistically it's similar to 2004-2005, and he has never really had a big drop-off since then. I think Rudy Gay is playing better than he did last year -- closer to his second season when he almost won Most Improved Player. I think he has even raised his game a little bit from there. And O.J. Mayo is even playing very well in his second season. Even Mike Conley is improving, too.
I think it's a combination of everyone improving, and the team now playing well as a unit. The strengths complement one another.
We play our starters a lot. Coach understands that that's really our strength. The results on the court -- I don't know if we're quite elite yet, but definitely in the top half of the League -- that has been successful to this point. At the same time, I've seen signs that the bench is developing. I think everyone would agree that we're looking for more production from our bench, and we recently signed Lester Hudson, (but it's too early to really evaluate that move). But I think you see that Sam Young has gotten better over the season. DeMarre Carroll has gotten better over the season and has had a bigger role in the last few games. Hopefully that's something that we can sustain.
Zach has always had big box-score numbers. But I guess one knock has been that he was something of a black hole -- if you gave the ball to Zach Randolph, he was going to shoot it. Are you aware if that was ever, or is not true?
Not something I've really looked into, but from what I've observed, I haven't seen him be that. I didn't know a ton about him. You hear a lot in the media about Josh Smith and how much better he's playing, and how he's not shooting as much from 3. You could say the same thing about Zach Randolph, although for some reason that story hasn't quite caught on. He has taken just a few 3-pointers this season.
[Note: A post-interview foray to Basketball-Reference reveals Randolph has been attempting just a fifth of the 3-pointers he shot of a year ago. His usage rate is the lowest it has been in seven years, and his true shooting percentage is the best of his career.]
He likes to shoot long 2-pointers, too.
And he has been hitting them! He likes that 18-, 19-footer from the wing. He sizes it up, realizes they're not going to challenge him, and sticks it.
You run the preeminent adjusted plus/minus website, but you're not wedded to just plus/minus based stats.
I think a lot of people have the sense, because of my website BasketballValue.com, that I am perhaps the leading advocate of plus/minus. They misunderstand. I think plus/minus is an interesting tool. I think your comments, that I've seen you post on TrueHoop, probably reflect my position. Certainly from just a handful of games it's hard to draw any real conclusions. And it's very context-sensitive, so it's very hard to say what's going to happen when Zach Randolph comes over [to Memphis from another team], for example. But over the long-term, you really need to ask the questions about why this or that player has a good or bad plus/minus, and maybe we'll be able to identify it.
But I would say that the fact I'm with the Grizzlies is certainly not a sign that we're an adjusted plus/minus shop. Not by any stretch of the imagination. There are a lot of good resources out there, like the work that John Hollinger has done, Basketball-Reference.com of Justin Kubatko and those guys, and of course Kevin Pelton. I pay attention to all those things.
We should talk about Kevin Pelton of Basketball Prospectus. He's the only national analyst I know of who predicted this Grizzlies team would make the playoffs. Most of us pundits thought it was going to be a disaster. He saw something, right?
He probably reflected improvement for Conley, Gay, Gasol and Mayo. I don't know if anyone could forecast Zach Randolph being quite as good as he has been, but if you predict him not far off that, the team still looks like a team that can definitely makes some noise.
There are various models that measure an individual player's contributions. If you take a combination of players with good ratings, and slam them together on a squad, does that mean you'll have a good team? Or, to put it another way, if Kevin Pelton projects all these players to improve, isn't it possible the players could be better by many measures, but the team would not win more?
Oh absolutely. ... I don't think you can sit at your computer and say "this is what we should do" and have the answer be obvious. I think there's always going to be the uncertainty. What does it mean if, in the first game of a playoff series, lineup X beat lineup Y 16-10. Does that mean they'd beat them 160-100? Probably not. It could even go the other way. It's really hard with some of these sample sizes.
I just think there are a lot of unknowns. No one is going to model Marc Gasol running the hills of Spain all summer and losing 25 pounds. That's not how anyone thinks. There's just way too much uncertainty. We can always make suggestions about what we kind of think might work, but the proof is in the pudding. You've got to see how it plays out on the court, and no one knows for sure.
Roland Beech is traveling with the Mavericks. Do you think it would help if you were on the bench every game?
I don't really know about that at this point. Conceivably it could. But a lot of the stats that come out are coming out on those sheets they hand out at the end of every quarter. And I don't know how much you want to change your approach, say, at half-time, based on 24 additional minutes of data. You had that post recently about the hot hand and Shane Battier passing up a shot ... Now that I'm more on the inside, I've come to appreciate more than ever the value of taking the long view. Fans, the media are always talking about making this or that change. But it has become really clear to me, watching our draft picks develop, that it takes a long time to assess where you are. I wouldn't want to over-react to 24 minutes of basketball.
Most of America has not seen the Grizzlies much. I think they have played one nationally televised game. As someone who has watched them play a lot, what's working?
When we're playing really well, you're really seeing us move the ball, snap it around the perimeter, high-post, Gasol and Randolph playing really well together ... one play, as we talked about, Randolph will step outside. On the next play, Gasol will hit that shot or dump it into Randolph posting up, where he has that nice hook shot coming through the lane.
And it really is amazing watching Zach Randolph battle on the boards. You can't even really see him battle. You just see him come up with it.
The ball's like a magnet to his hands.
It's amazing. I heard the broadcast team, Sean Tuohy and Pete Pranica, mention the other day that Randolph has hardly any dunks this season. It's amazing that a guy can get so many rebounds playing his game so close to the floor. Hope for people like you and me who can only imagine really being able to jump.
And what is Rudy Gay doing better?
He's able to jump over everybody. Every shot he takes, I'm sure, feels like an easy shot. When he's playing really well, he's creating open shots for everybody else.
The other thing we're definitely monitoring as a team is keeping our turnovers down. That has definitely been an issue that we're watching over the season. And it's no coincidence that when we're playing better we have a lot fewer turnovers.
How do you decide to keep turnovers down?
I think it really just has to be a point of emphasis throughout the whole organization. Coaches stress it. Players don't need coaches to tell them turnovers are bad. But I think it's good for everybody to keep thinking about it.
Are there All-Stars on this team?
I think Zach Randolph deserves to be considered one of the top 15 or so players in the NBA. Whether that translates to being an All-Star ... is an interesting question. I think I saw that Tracy McGrady is going to be an All-Star. That's hard to say. And, of course, the power forward position is stacked in the West. But his individual stats are tremendous.
You know I think Gasol, Gay and Mayo are all also playing extremely, extremely well.
Gasol is actually our team leader in plus/minus. He has a little quieter stats. Different kinds of contributions. But he has also played very, very well.
Has enough of the season passed for us to have a big enough sample size to believe plus/minus stats on something like that?
I think so. The numbers say it's by a pretty wide margin. So, even if the margin isn't as wide as it looks right now, he's definitely a very valuable contributor. That doesn't diminish the things everybody else is doing. It speaks more to the many different things different players on the team are contributing.
So what's the biggest difference between the Grizzlies and the MIT team you played with?
Ahh! [Laughs.] I'd say that we didn't need to practice alley-oop passes. Actually, that's really true! You and I don't appreciate how hard it is to throw a good alley-oop pass. People get all upset if one's off in an NBA game, but try throwing one in pickup ... you'll be spraying them all over the place.