If you don't know who Jeff Ma is, you haven't been paying attention. Ma is the mind behind the famous card-counting MIT blackjack team that won millions of dollars in Las Vegas, the team that had a book written about them and a movie made about them.
But if you didn't know that Jeff Ma has consulted with the Portland Trail Blazers regarding advanced statistics for years, you are forgiven. For whatever reason, the second most famous person with recent ties to the organization (trailing only Paul Allen, of course) has flown almost totally under the radar here in Rip City, even though he's a mainstay in the advanced stats community, a co-founder of Citizen Sports and mentioned regularly on TrueHoop.
It just so happens that Ma wrote a book this summer called, The House Advantage: Playing the Odds to Win Big In Business. He discussed portions of the book, including a great exchange with Jerry West, with Henry Abbott here.
Another great exchange in the book comes when Ma meets with Kevin Pritchard, Tom Penn and the Blazers scouting staff in Portland, shortly after Pritchard was promoted to GM. Ma writes that he "realized what an important, delicate meeting it would be, as I was sure the people in that room would be predisposed to doubt all that I was prepared to sell." He then describes being pelted with questions about his analysis which led to Pritchard stopping the meeting to ask, "How do we get your rankings to look more like ours?"
Here was a guy who had been involved with the game at many levels, as a player and an executive. He'd played in the NBA for many years and had coached. He was now one of the hottest young general managers in the game. All told, he'd been in basketball for more than twenty years as a high-level player, coach, or executive. He clearly knew the game backwards and forward and had confidence in his opinions.
I didn't doubt Pritchard's knowledge, but this question seemed proof that he didn't really get what we were trying to do or what our value was in this process. He wanted the numbers to be comforting, and, although he appreciated their value, he didn't want them to conflict with what his gut and scouts were telling him.
"We don't want our rankings to look like yours," I answered quickly, and then continued, "Our goal is to measure things in a different way than your scouts might. We hopefully are seeing different things in the numbers than you can with your eyes. And the reality is, you guys will see different things that the numbers can't. So we should be happy that our rankings don't look exactly alike.
He later writes, "Pritchard understood what I was saying, and the meeting ended on a positive note."
This vignette, which covers roughly four or five pages, is a fascinating behind-the-scenes look at the development of Portland's famed analytics department. Given this passage and the fact that The House Advantage is a book primarily about ways companies can become more functional, I thought it made sense to get Ma's perspective on the end of the Pritchard era and all the corporate dysfunction that surrounded it.
Click through for a transcript of our full conversation, which ran roughly 30 minutes this morning. Ma discusses Pritchard, Paul Allen, the consulting work he provided to the Blazers, the team's emphasis on pace, Pritchard's emphasis on a "big three" and, of course, who his numbers recommended drafting in 2007: Greg Oden or Kevin Durant.
"If people that use analytics to predict player performance in the NBA, using performance analytics, meaning what they did in college, and they tell you they had Oden ranked higher than Durant, they are full of crap," Ma said this morning. "There are very few statistical measures that would have rated Oden's numbers in college better than Durant's. Oden was injured his entire career, that one season at Ohio State. He had to shoot free throws left handed, was not efficient, didn't have a great statistical season.
"Our numbers absolutely said they should pick Durant. It wasn't even close."
-- Ben Gollier | email@example.com | Twitter
Here's the transcript of my conversation with Jeff Ma. My questions are in bold. His answers are block-quoted.
Having read Bringing Down The House and now The House Advantage I've been familiar with your story for awhile. But for whatever reason your work with the Blazers has flown under the radar. When did you start working with the Blazers and what kind of work did they have you doing?
It all started about five years ago. There was a guy by the name of Hoon Cho who was working with Paul Allen at Vulcan. He was actually on the board of both the Seahawks and the Blazers. He was a friend of mine. He's still a friend of mine. He suggested that I start talking with those teams a little bit because at that time my company, ProTrade, was pretty focused on trying to find better analytical ways for teams to make decisions. We were looking at advanced metrics in sports. Hoon thought this would be a great opportunity for me to break into the sports world and also help Paul's teams make better decisions.
He introduced me to Kevin Pritchard. Kevin at that time was just the Director of Player Personnel. Kevin, because he had had a stint outside of basketball for a brief time, I believe in the private equity world for about six months, he certainly had an appreciation for data and how to use analytics to guide decision-making. I came up [to Portland] and talked to him for awhile and went to lunch and hit it off well enough to the point where he wanted me to come back up and talk to the staff. At that time it was [General Manager John] Nash and [President Steve] Patterson. We met with those guys and told them what we thought we could do, and then eventually they retained us to do some projects for them.
The project we did for them was how to use college statistics to project the success rate of different players within the NBA.
So you're feeding them a list of data regarding an incoming draft class right before the draft? Or was this a year-round thing as well?
We run it a couple of times. We'll run it somewhere in the middle of the season but really at the end of the season is where we do the most work. At the end of the college season we'll run the stats for that year and the Blazers will give us a list of all the guys that they are interested in or that they think will be drafted and we'll put those guys into our model. Then it will spit out a number for each player based, also, on what position they'll play in the NBA and what their chance of success and becoming an elite NBA player will be.
We don't really rank. What we don't do, and this is a little bit misunderstood, we don't rank for them who we think will be the best player in the draft. We actually rank which player we think has the best chance of becoming an elite player, meaning a top 10% guy at his position.
Correct me if I'm wrong, but the use of college translations has really taken off over the last few years in the NBA. When you guys started doing this, back in 2005, did you get the sense you were on the cutting edge of how stats were being used at the time?
This was right before [Denver Nuggets quant guy] Dean Oliver and all those guys got jobs in the NBA. It was definitely a little bit novel at the time. Certainly it helped the work that we did, it helped other people get jobs with the Blazers... There were only a handful of teams working on it. This is right before [Houston Rockets GM] Daryl Morey had any kind of visibility within the league. It was pretty early on.
Kevin Pritchard was a hugely popular figure here in Portland. He's often described as a charismatic figure. In the book, there's a passage about a meeting with Pritchard, Penn and the scouts where he isn't quite understanding what you're trying to do with analytics and you write something along the lines of, "He doesn't really get it." I imagine that you've seen all sorts of different executives over the last decade or so. What were your initial impressions of Kevin Pritchard, the executive and person?
First of all, the idea of Kevin being charismatic and what not is pretty dead on. Kevin is a good guy. He welcomed us and welcomed me into this [basketball] world and was accepting of trying to find new ideas that might help him make better decisions. He went on to build out a whole analytics team over there where he had, I want to say, four or five people working on this kind of stuff. From a personal level, I think Kevin is very genuine, wants to win, wants to work hard. He's the right kind of person that I would want to run my team that I was a fan of.
From an analytics standpoint, you can separate the sports executives into three buckets. There are the guys that do analytics themselves. That could sit down at a spreadsheet and crunch out the numbers themselves. There's only a handful of those. Daryl Morey, [Rockets executive] Sam Hinkie, [San Diego Padres executive] Paul De'Podesta in baseball, there's a handful of those guys. Then there are the guys in the middle who have an appreciation for analytics, they don't necessarily understand how to do it or understand a lot of the basic principles but they understand it has value in what they're doing. Then there's a third group that doesn't think there's any place for it in sports.
Kevin is in that middle group. In some respects I think that's the best group to be in because you have a great appreciation for what the scouts do and the process of scouting. I think that's important, that's valuable, that's part of this whole thing. You don't want to have too limited of a viewpoint. I'm not saying Daryl and those guys have too limited of a viewpoint but being in that middle group is the best place to be in some respects because you do allow yourself the ability to make decisions in a lot of different ways, which I think is important when you're making decisions.
In the book you joke that when you were going to sit down with the team's scouts you were worried they were going to shoot down everything you had done right from the jump. It's funny because these days analytics and numbers are one of the first things that the scouts talk about when it comes to the draft process, they talk about an "eyes, ears and numbers" approach. Would you say that your work caused that impact or that it was a development of the program? Does that philosophy shift surprise you given your early trepidation?
I'd give that credit to Kevin. I'd like to think I had a little bit of an impact on how Kevin thought about things. You can see that in the passage that I wrote about in the book, there's this whole thing with Kevin where Kevin doesn't necessarily agree with what we're seeing. I spent a lot of time explaining to Kevin - he definitely had a hard time at first, and I think this is true with a lot of people as they start to understand analytics - you have to become comfortable with the fact that information is going to tell you something different than you might think otherwise. If it doesn't tell you something different, then it's not useful information.
It's nice for me to hear from you that that is the way their scouts think, because that is certainly how I'd want their scouts to think. I don't spend weeks at a time with their scouts. I've spent time most recently with Chad Buchanan and Mike Born and I've spent time with Kevin, there were a couple of times I would come up there for an all day session. I would like to say, "yes, I definitely feel like I had an impact," but I think that mostly comes from Kevin and what he was preaching to his guys.
Another side to The House Advantage is these business lessons that you've applied from your blackjack success. From the book I get the sense that you have a pretty clear business philosophy. You talk about the benefits of working in a start-up environment. As we know, the Blazers aren't a start-up given their bureaucracy and chain of command. You write about how businesses can be more effective and efficient as an organization, aligning your incentives, making sure everyone has equal buy-in to the team goal.
Unfortunately the Blazers went through a pretty tough summer -- with Pritchard being fired, Tom Penn being fired, back room drama. As you said, you weren't there on a daily basis but you're having somewhat regular contact with them on the analytics side. Did all of the drama take you by surprise? Were you reading these headlines thinking, "What the heck is going on?" Or did you have some hints of the dysfunction that ended up coming out this summer?
I think in a lot of ways it came from the owner. And I've never spent time with the owner. That part of it I didn't see. I think it came as a little bit of a surprise to me because I don't spend a lot of time looking at the politics of the organization. My goal when I'm there is to give them good information to make better decisions and they can use it however they want. As an outsider looking in, the last thing you want to do is poke your nose where it doesn't belong.
That being said, this is an organization that has gone through a lot of turnover even in the time that I was there. In the last five years, think about all that has happened. President, GM, Kevin.
Kevin's quick rise to GM didn't surprise me because I thought he certainly was a superstar but it surprised me because it happened really fast. I think you don't really look at the Blazers as a pillar of stability. What ended up happening there was certainly surprising because you look at Kevin and you think he's one of the top, young GMs and who would want to get rid of him? Obviously the timing of what had happened seemed rather odd. But once you started hearing bits and pieces of what happened it all kind of made sense.
What do you make of Rich Cho, Kevin's replacement? Do you think he was a smart hire?
I don't know Rich at all. I've heard about him and certainly he's got a pretty good reputation. It probably is even more of a step towards analytics, I would say. It's hard for me to comment without knowing him very well. I wish I could give him a ringing endorsement but I don't know him well enough to do that.
In the book you investigate football statistics a little bit, explaining that looking at turnovers is a more reliable way to predict victories than something like rushing attempts or yards. Dean Oliver has laid out his "Four Factors" that guide success in basketball. Are you in lockstep with those factors or is there another stat or stats that you look at as more predictive of winning in basketball that the casual fan might overlook?
When I look at a game, from the standpoint of what's going on in that game, I tend to look at things like rebounds. When you focus too much on the score, you get bogged down into things that might be impacted by variance.
In sports the best way to look for inefficiencies or where the score might be lying in terms of what's actually happening is to look for places where variance or luck might play a factor. If you look at how a team is shooting from three point range especially. Whether a team is getting to the line, how many foul shots.
Let's say the Orlando Magic are up on the Blazers by 12 points at the half. You look at the game and you're like, "Wow, are the Blazers out of this game?" But you look up and see the Magic are 6-7 from three point range. The Magic are a great three point shooting team but 6-7 is ridiculous, right? Realistically they are probably a 3-7 type team. The Blazers are a pretty good defensive team too. So it's probably a product of the Magic getting kind of lucky. In my mind, I'm saying that the Magic should really only be up by three and that's a chance for the Blazers to have a good shot in the second half. Once the Magic shooting goes back to normal -- and I'm not saying the Magic are going to have a 0-7 or 1-7 half, because there's no evidence from the first half that would tell you that -- but I am saying their lead has been built on variance in three point shooting.
One other thing the Blazers have seemed to emphasize is limiting turnovers and, in turn, slowing the pace all the way down so that they're the slowest paced team in the league. Was that a directive from the coaching staff or was management involved there too? When they had you do the college translations were they asking you, "Which players do you think will work best in a slow-paced offense?"
They didn't include me in that type of decision. But I think that was a top to bottom decision. I think they realized given the types of players that they had that they wanted to play a slow pace, they wanted to play good defense, that's just the way everyone from Kevin to Nate down the line, that's the way they wanted to run the organization.
Kevin was very big on the Spurs model, having three superstars and having seven guys that would run through the wall for those superstars. That was one of the things he said to me early on. If you look at what he was trying to do, he had Brandon [Roy], he had LaMarcus [Aldridge], he was trying to find that third guy to round them out. And then fill the team with those good character, high effort guys. In terms of a pace thing he realized he didn't have that third superstar yet so they realized they needed more shooters. They realized they probably couldn't play up in the 100s, they needed to keep the ball and keep the game under 100 points.
You talked about finding that third superstar and they had hoped that would be Greg Oden. Do you remember what your translations said about Oden and Kevin Durant going into that 2007 draft? Did you have both translating really well to the NBA game? Did you have a preference one way or another?
Yes. A very big preference actually. If people that use analytics to predict player performance in the NBA, using performance analytics, meaning what they did in college, and they tell you they had Oden ranked higher than Durant, they are full of crap. There are very few statistical measures that would have rated Oden's numbers in college better than Durant's. Oden was injured his entire career, that one season at Ohio State. He had to shoot free throws left handed, was not efficient, didn't have a great statistical season.
Our numbers absolutely said they should pick Durant. It wasn't even close.
But that kind of decision is never that cut and dry. I would never want the Blazers to make the decision so cut and dry. The thinking they had was that this elite center is very rare and the ability to get that guy was staring them in the face and that's what they went after. The sad thing is that when you ignore the numbers, the numbers often tell you something regardless of what you're ignoring. The numbers in this case were ignored because Oden was hurt but what have we seen in Oden's career? A propensity to get hurt.
I felt like they should have drafted Durant and said they should have drafted Durant but I think it's really easy to look at this with hindsight. If you had polled NBA executives and even statheads at that time, who they should pick, I think at least 2/3 of them would have said Oden.
How did you react when they announced they picked Oden? You talk a lot in the book about removing the emotional element from key decisions so that you can see clearly and make the right decision. Did you have any emotional investment at all when you're making these recommendations on players?
I was disappointed about the whole Oden thing. You work on something like this, you would like... I think one of the things... We were really just a small piece in their process, which is the right way for us to be because we're not there on a day-to-day basis. But in the Oden case, I did, I wanted to see them do what was the unconventional decision, what was the harder decision.
If you look at my book, and this is a good point to bring back, this is a perfect example of using the numbers to make an unconventional or difficult decision. Conventional wisdom would tell you to take the big man. It wouldn't tell you to take Durant. It's the old thing about no one gets fired for buying IBM. No one ever gets fired for drafting a big man. If things had flipflopped and Oden had been amazing and Durant had just been a high-volume shooter and high-volume scorer, everyone would second-guess that just as much, if not more.
My disappointment was that, in a lot of ways, [the Blazers] made the safe choice versus what I thought was the right choice.
I think Kevin always made decisions that he thought were the right decisions. He could do some unconventional things. But I think a lot GMs would have made that same decision and a lot of them would have made it because they thought it was the safe decision.
Just think about the headlines or the buzzwords. "How can you give up a chance to get a 7 footer? Bla Bla Bla." I'm sure that's what would go through a lot of executives' minds if faced with a similar situation.
You've set up your company. You've written a book. You write that you have a huge passion for sports and always have. Do you have a dream job working in sports?
I think that there was a time where I really wanted to be a general manager but now I'm kind of thinking more about being the president of a team. The big thing is... regardless of what I do for a team, if I was going to work full time, I would need to be really, really close to the owner. Ultimately the owner is the one who has the most incentive. Daryl Morey and I have talked about this: the owner has the best perspective or viewpoint on the team. If you're not close to the owner and you work for the team, you don't have a ton of ability to effect change and get your feelings heard.
If you think about all the executives in sports, if you're always worried about your job or you're always worried about pleasing your owner, you're not going to be able to effectively do your job. You need to have a really strong relationship where they respect you and trust you, and you make a decision that may be unpopular with them and for whatever reason doesn't work out, they need to give you a little bit of room to continue to make those difficult decisions. Because every decision is not going to work out positively.
Follow the link to purchase The House Advantage: Playing the Odds to Win Big In Business on Amazon.
-- Ben Gollier | firstname.lastname@example.org | Twitter