Email Story

Reference Madness!

Jason posted 7/5/2007 from ballhype.com

As I have noted several times over the past few years, Basketball-Reference.com is a terrific resource for NBA fans. Even when I'm not digging around for some kind of statistical analysis, I find myself stopping by quite frequently to refresh my memory on various topics, like which team won the 1974-75 championship, who received the 1993-94 Rookie of the Year trophy, and who was picked 3rd overall in the 2002 draft.

So, I was happy to see the Basketball-Reference curator, Justin Kubatko, win the TrueHoop Stat Geek Smackdown. Not too happy, though, because I money riding on Kevin Pelton to win (Vince Carter killed us).

Justin was gracious enough to answer a few of my questions via email (and cagey enough to avoid revealing any trade secrets). Here's the transcript:

You credited your Stat Geek Smackdown win to the fact that you stuck with your numbers and ignored your gut instincts. What can you tell us about the technique you used to predict outcomes? How did you derive probabilities for each of the possible outcomes?

I did not do a straight team vs. team analysis to make my picks. Instead, I took into account the home court advantage and estimated the probability that Team A would win in four games, the probability that Team B would win in four games, and so on. I chose the outcome with the highest probability. As to how those probabilities were derived, well, I'll have to keep that a secret since Henry will probably ask me to compete again next year.

OK, OK—you can't give away the formula. But could you at least tell us whether all of your source data was from Basketball-Reference.com? In other words, if you gave 1,000 monkeys enough time with your web site and Excel, would they eventually be able to come up with it? Also, have you considered taking your technique to Vegas?

Hypothetically, if you gave enough monkeys enough time they could re-produce what I did using just my site. I have not considered taking my technique to Vegas because — unless you're Jeff Ma — the house always wins.

You won the playoffs-picking contest, but which of the participants would win in an actual 1-on-1 tournament?

I have no idea, but I'm sure it would be ugly. I'll challenge any of them to a long-distance race, though. (I finished 49th out of 3016 finishers at the Columbus Half Marathon this past April.)

Congratulations. According to Henry Abbott's recap, you recently quit your day job to focus on Baketball-Reference.com full time. Was earning a living from the site a goal when you first launched it 3 years ago?

No, not at all. I started it mainly as a hobby. I knew Sean Forman through the Society for American Baseball Research (SABR) and I knew he had the domain name reserved. I asked him if he wanted me to develop a basketball site for him, and he was quite enthusiastic about it. I'm ecstatic that things have worked out this way, though.

You have some cool query tools in testing. Now that you're working on the site full-time, should we expect to see more of this kind of thing going forward? What other plans do you have for improving or expanding the site?

Yes, I hope to develop more tools that will allow the user to do interesting queries and comparisons. In the immediate future (i.e., in the next couple of days) I plan to release a WNBA section with league summaries, team pages, and player pages. [It's up now.] I also hope to add some older box score data to the site this summer. There are some other things I am working on at the moment, but they are basically behind the scenes projects such as improving code, etc.

What determines which players show up on the home page at any given time? I love the fact that The Whopper is up there right now.

Three players are randomly selected every hour. The player list includes players who have been selected to at least one All-Star game and/or been elected to the Hall of Fame.

Interest in the statistical aspects of basketball seem to have increased quite a bit over the past decade thanks to groups like the APBR, but baseball still enjoys much more attention from statisticians. Do you foresee basketball ever reaching baseball's level of popularity from a numbers perspective?

No, I don't, in the sense that there will always be a greater attachment to baseball's numbers than basketball's. Quick, without looking it up tell me how many career home runs Hank Aaron hit. Most serious sports fans can answer that in seconds. But now answer this question: How many career points did Kareem Abdul-Jabbar have? I can't even answer that one off the top of my head.

Do you have any favorite statistical nuggets from the site? How about favorite all-time players or teams?

I don't really have a favorite statistical nugget from the site, as every day it seems like I discover something new and interesting. My all-time favorite player is Tim Duncan. I was living in a small town fairly close to Winston-Salem when Duncan was at Wake Forest, so I was able to see him play quite a bit on television and a couple of times in person. I really like that he goes about his business without any flash or fanfare, and he seems like a great teammate. My all-time favorite team is probably the Showtime Lakers. I first started watching the NBA in the 1980s, and I immediately loved the Lakers because of their fast-break style and, of course, their success. Now, however, I can't stand the Lakers, so my current favorite is the Spurs, mainly because of Duncan. (Yes, I'm an unabashed front-runner.)

How about your top underrated/overrated players? Are there any players over the years who you think have been given more credit or less credit than they deserve?

Basketball is like most other sports in that players who have one signature skill, particularly scorers in basketball, tend to get overrated while players who can do multiple things well tend to get underrated. An active player who I think is underrated is Chris Paul. That may seem like a strange choice given that Paul has received a fair amount of attention in his first two years, but I don't think people appreciate just how good he has been. On the other hand, I think the general public has overrated Carmelo Anthony. Anthony's a good player, don't get me wrong, but he tends to get lumped in with Dwyane Wade and LeBron James because they were all drafted the same year, but he's just not in that class. I actually prefer another player in that draft class, Chris Bosh, to Anthony. Historically speaking I think Horace Grant has been very underrated, while Pete Maravich has been overrated.

Russell or Chamberlain?

Pass. (Bill James readers will get that.)

OK, last question: how many NBA players have averaged at least 1.5 steals, 1.5 blocks, 1.5 three-pointers, and 1.5 assists per game for an entire season?

One: Big Shot Bob in 1995-96.