The Advancing World of Advanced Stats

Jul 15, 2014; Minneapolis, MN, USA; American League infielder Miguel Cabrera (right) of the Detroit Tigers celebrates with Mike Trout (27) of the Los Angeles Angels after hitting a two-run home run in the first inning during the 2014 MLB All Star Game at Target Field. Mandatory Credit: Scott Rovak-USA TODAY Sports

Embracing the Stats Movement has been Slow Across Other Sports

A recent article over at ESPN highlighted the growing influence of advanced statistical analysis in the world of hockey, and how it’s changing the way the sport is seen, judged and played. The last six years or so  has seen a similar explosion in basketball from the creation of ESPN’s Player Efficiency Rating to True Shooting Percentage and now the installation and public access of the SportVu motion tracking cameras in all arenas.

And of course, this all bleeds from the world of baseball, where that brainchild of Bill James Sabermetrics has caused a seismic shift in the sport in the last decade or so. Truly it’s a brave new world for sports fans, writers, execs and players.

It goes without saying that any extra information that allows fans to understand the game on the same level as insiders is only a good thing. One has to wonder though, did the outsiders get the leg up on front offices in baseball? After all, it wasn’t until a few years ago that teams started employing economists and their ilk to find players. In the movie version of Moneyball we see Jonah Hill playing Peter Brand, a real life [redacted], seemingly the guy who convinced Billy Beane to hunt on the margins for that extra little something, the inefficiency that other teams with money were ignoring.

The thing about this whole movement encroaching in on hockey, football and basketball… I just don’t know if it’ll work quite as well.

Sabermetrics, the large majority of them anyway, work because baseball is inherently a static game. The batter stands there, the pitcher throws, something happens, and repeat this 100 or more times a game.

Jul 11, 2014; Cleveland, OH, USA; Cleveland Indians starting pitcher Corey Kluber (28) pitches during the first inning against the Chicago White Sox at Progressive Field. Mandatory Credit: Ken Blaze-USA TODAY Sports

In football, there’s a big issue simply because they only have 16 games. Outside of maybe QB throws, nothing is repeated enough for a solid sample. Baseball plays 10 times more than that. Not only that, but so much of football relies on something else happening. Football Outsiders might be the biggest voice for advanced analysis of football, and even they have flaws. For instance, some of their rankings for cornerbacks fail to take into account things like the safety messing up. Say coverage is blown on a reciever – this could happen just because they’re in a cover 2 or some other zone defense, and the corner is expecting the safety to be helping over the top. If that guy’s not there, all of a sudden a wideout is running free and the CB looks like the chump.

Basketball and hockey are similar. There’s so much context to what is happening. A player could have  numbers off the charts that in reality could be garbage, or vice versa. Look at Chris Bosh – this is a guy with All-NBA levels of talent, a number one option type of player who, for the past four years, has had to play out of position and in a tertiary role because there were two guys on the Miami Heat who were objectively better, and his stats suffered. But he was integral to that team’s success.

On the flip side look at someone like Kevin Love. A brilliant player, no doubt about it. The man racks up both points and rebounds. But he doesn’t box out, he chases the rebound and that hurts his team. Meanwhile Brooke Lopez on the Brooklyn Nets is an elite box out man, it’s just that he plays with board robbers like Reggie Evans, a poor defensive player but a compiler. There’s just too many moving parts.

Hockey is the same way. It’s hard to just pull raw numbers and plug them in like in baseball. So much of what makes a guy like Patrick Kane or Sidney Crosby great isn’t what they’re doing with the puck, it’s what they’re doing elsewhere. They find holes in the defense, read guys to see where they’ll be and almost instinctually find the the right place at the right time, and also have an incredible ability to control the puck. There’s so much more that isn’t understood yet, though it will be fun to see where it goes.

Another issue I mentioned before is that the public is getting screwed in other sports when it comes to analysts.

By the time it caught on in baseball this stuff had been in the public consciousness for decades. Papers were written, books published, conferences held. There might be some things we don’t know like the Indians’ proprietary software DiamondView that they’ve used for a long time now. But by and large, we know what to look for and since the game is so static anyway it’s not hard to understand. But with basketball and hockey in particular, these analysts are scooped right up, their ideas locked away by teams to gain an edge on the opponent.

John Hollinger, for instance, was an ESPN writer and big numbers guy. He created PER for basketball. The Grizzlies brought him on and now his knowledge is gone from the public sphere and we suffer for it. He’s not the only one, I’m surprised Kirk Goldsberry of Grantland has survived in the public sphere so long. The NBA HAS allowed the release of the raw data from SportVu cameras, but all that tells us is things like Nic Batum ran further than anyone else in the league, or DeAndre Jordan had more rebound opportunities than anyone else. That second makes you think he actually underperformed, pulling down “only”13.6 per game. It’s a load of data that we need to massage.

It’s the beginning of the conversation, not the middle or end like the work being done at FanGraphs or Baseball Prospectus. Plus we must note that as the world of advanced statistical analysis has advanced in baseball, the stathead has realized the need for the scout, and the scout for the stathead. These days it’s a little bit of both that makes the happy whole. You wonder when that realization will arrive in other sports.

Teams like the NHL’s Toronto Maple Leafs were still going against the grain of modern thinking because it worked two years ago and they made the playoffs that led to an epic collapse, then they caved in last year. They poo-pooed even the rudimentary, accepted numbers like possession time and Corsi and Fenwick stats, both very real things even if they don’t have acronym names. They study shot rates at the goal when teams are at even strength, if you’re curious. They’re making the switch, and if a marquee team like that is on the advanced analysis bandwagon, it’s only a matter of time before it totally explodes. If it hasn’t already.

Stats are something that makes baseball extra fun. What’s better than a playful argument? Alright, quite a few things but most of those are illicit or indecent in Victorian company so we won’t discuss that now. The nature of the game gives us such an understanding of the abilities not just of individual players, but of all people. I have a much better idea of just WHY Mike Trout is so much better than me because things are charted. Also he works out consistently and had a drive to do something with his life before the age of 25. It’ll be an interesting next couple of years as the other sports are decoded as baseball has been. I still think there’s going to be a gulf between advanced stats in hockey and basketball, and especially football because of all the other stuff, but if the window opens just a little that’s all the better for us fans.