In the 1982-83 season, Wayne Gretzky finished fifth in the NHL with 18 power-play goals, behind co-leaders Paul Gardner of Pittsburgh and Al Secord of the Blackhawks (20). Yet, by a completely different statistical measurement, he was the most valuable player in hockey when his team had the man advantage: Gretzky led the League in "power-play goals scored by his team while he was on the ice" with 77; Secord (56) and Gardner (53) didn't even crack the Top 10.
Fans will bend these facts in any number of subjective ways, but the great news is that they can now search for this type of arcane information more easily and efficiently than ever before on the Web. Hockey-Reference.com, the latest in a family of essential sports numbers sites, launched two weeks ago and immediately changed the game for hockey stat-heads. Looking back at those 1982-83 leaders, you can compare basic stats (goals, assists) with more sabermetric measurements like "Goals Created" and "Total Goals Against While the Player Was on the Ice" -- for a season that happened 25 years ago. It's one of those sites where you begin clicking around and then, five hours later, you wonder where the hell the day went.
Recently, I spoke with Justin Kubatko, Vice President of Sports Reference LLC and the person in charge of creating Hockey-Reference.com. "I'm a hockey fan. I wouldn't say I'm a die-hard hockey fan, but I grew up near Pittsburgh. When the Penguins won those back-to-back Cups in the'90s, I remember those well," said Kubatko, 36, who now lives in Columbus. He said Sports Reference LLC had planned on creating a hockey site for years, but needed "a good data set and a license for the data" before starting up. They finally found the right fit with Dan Diamond and Associates, publishers of Total Hockey and the annual NHL Official Guide and Record Book. Here's more from Kubatko, about launching the site and hockey numbers in general:
Q. While there are fans who dive head-first into the numbers game, are hockey stats overcomplicated? Do they scare off casual sports fans?
KUBATKO: For one, the fact that some games hand out three points and other games hand out two points is an abomination. I just hate those overtime points. It makes the standings very confusing. If you're someone who doesn't understand hockey very well, and you look at the standings, it's going to be hard to understand it. Looking backward, they've changing it, like, three times now.
Like I said, I'm a hockey fan, but I was never really, really into it. So some of the statistics that they track, the acronyms I couldn't get. I knew what "Time on Ice" was, but somebody else was looking at the Web site and had no idea what "TOI" was.
Q. I thought it was French the first time I saw it. Fantasy sports have opened up other games to the more numerically obsessed in sports fandom; could hockey stats, presented as they are on Hockey-Reference.com, attract those kinds of new fans to the NHL?
KUBATKO: You know what I think about hockey? You have to go to a game. You actually have to go to a game because, to me, it's the one sport where you have the greatest difference between seeing it on TV and seeing it at the game. I think if you could get people to go to a game, you could hook a lot of people. Maybe numbers are a way to get some people interested, but in a way I think those are the kinds of people who would enjoy looking at a site like mine but wouldn't necessarily enjoy sitting down to watch a game. For some people, it's more about having cool numbers to play around with rather than about the sport itself.
Q. What were some of the challenges in putting this site together?
KUBATKO: Lack of data is a big thing. They didn't track shots on goal until the modern era; they only go back about 40 years. Time on ice, they didn't track that until, like, 10 years ago; you'd think that would have been something they wanted to record.
Q. How far back does the database go?
KUBATKO: We should have anybody who's ever played in the NHL, so that goes back to 1917-18. For some bigger names that didn't make the NHL, we should have some minor league stats or juniors stats or other professional leagues. Like Hobey Baker, who never played in the NHL but played in some other leagues. What's great about Dan's data is that he doesn't just have Wayne Gretzky's NHL data; he has all the junior stats and international competition, too.
Q. Your other sites aren't all just about box scores and player stats pages. Your baseball site has a staggering amount of information, from career projections to measuring Hall of Fame potential.
KUBATKO: When baseball or basketball first came out, they didn't look anything like they look like now. It's a constantly evolving state. What I want to do is work on a metric that predicts a player's probability to be elected to the Hall of Fame one day. Maybe some sort of similarity scores [between players], although that's kind of hard in hockey because if you go back far enough, there isn't a lot of data to work with.
[Down the road], we'll add game summary data. There's this project called the Hockey Summary Project; it's like Retrosheet for hockey. One of the things we'll want to work on is cleaning those [box scores] up so you can search them.
Q. Finally, your site enters a marketplace where there's already a competitor that most fans believe set the standard for hockey information on the Web: The Internet Hockey Database. Did you consider that site in creating Hockey-Reference.com, and is the Web big enough for the both of you?
KUBATKO: We knew it existed, and we knew it would probably be our No. 1 competition. But if you look at our football, baseball and basketball sites, they all have a common theme and a common way to present information. We feel like we do a pretty good job of that. We never looked at any other site and said, "We have to do this to beat this site." We're not looking to put anybody out of business. We did what we normally do and hope that it turned out for the best. And I think it did.