Moneyball by Michael Lewis was the accelerating factor in baseball’s analytical revolution back in 2003. Even so, the seminal point of the book is often misunderstood to this day. People still say Moneyball was an argument for on-base percentage. It was on the face of it, but on the whole it really wasn’t. Because on-base percentage was undervalued in the market place, it was, but the idea was to buy whatever was undervalued and sell whatever was over valued. Moneyball was about value and markets.
Even though that was the point, the lasting impression still seems to be ‘on-base percentage is important in baseball.’ Understanding that baseball’s clock is outs and any player that forestalls that clock contributes towards their team scoring runs in myriad ways is central to the new understanding of baseball. The markets stuff is important if your concern is how to run a team, but understanding the value of getting on base is central to understanding the sport itself whether as a fan, a scout, or a member off a front office.
Hockey, though a much different creature, is going through a similar analytical revolution right now. I was thinking about this today while my wife was practicing her lecture on the pharmacology of the heart. (You can see how the mind might wander.) Her central point was that some drugs are miracles when applied to certain disease states but if used incorrectly, they can be deadly. This is true I think of the analytic revolution in baseball and hockey. What is true in one sport isn’t necessarily true of another, except possibly when speaking in grand strokes. Since I’m new to hockey analytics, a baseball writer, and Matthew Coller was kind enough to give me this forum, broad strokes are what I’m going to do here. I hope that looking at hockey’s statistical awakening through the lens of baseball’s will be instructive, and possibly even interesting. Here’s to hoping!
In baseball, on-base percentage is the Holy Grail. Once you grasp that central truism, you’re on the way to a deeper understanding of the sport. All other ideas, the real value of RBIs, why errors are bunk, why bunting is often dumb, the efficacy of stealing, even the value of FIP, are grasped following that gateway point. So what is hockey’s gateway stat? Possession.
It has to be possession. I was thinking about this (again during my wife’s lecture; sorry honey!) and it occurred to me that hockey is the only one of the major sports that doesn’t alternate offensive chances. In baseball one team bats while the other fields, but then they switch. Each team gets an equal amount of outs throughout the game and thus an equal amount of chances to score. In basketball as soon as a team scores, the other team gets the ball. In football you score a touchdown or kick a field goal and the next play (other than an extra point) is kicking the ball to the other team. But in hockey, you score a goal and the next play is a face-off. Despite the fact that one team just scored, the play starts with an equal opportunity for possession.
Put another way, on-base percentage is baseball’s possession. As long as players continue getting on base, the other team will literally never come up to bat. As long as a team possesses the puck, the other team will literally never score a goal. All other aspects, like shooting percentage, goalie play, shot quality, and what have you fall by the wayside. Possessing the puck can solve all a hockey team’s problems.
Of course, in baseball if you continue to get on base infinitely, the other team will never come to bat and the game will never end. That’s not so in hockey. One team can possess the puck for all of 60 minutes and as long as they’ve managed to score one single goal during that time (and not accidentally put the puck in their own net) the game will end and they will win. Every single time. Having the puck is never negative.
This new revolution in hockey seems to be placing much weight on possession, but I’m not sure it’s enough. I’m not sure it can be enough. Possessing the puck is, it seems, the greatest thing a player can do aside from score. On the other hand, the worst thing a player can do in any situation is to give up possession of the puck.
Any player that receives the puck or carries the puck is being entrusted with the essence of the game itself. This is why playing dump-and-chase is like bunting. Giving up possession for an easier zone entry in the hopes of getting the puck back is analogous to giving up an out (or trying to) in the hopes of a different batter scoring a runner further along the base path. In both cases it can work out, but in both cases the downside is significant and thus the strategy should be employed very selectively, much more selectively than current strategy dictates.
That brings up one other quick but important point. One thing baseball’s sabermetric community has learned over the last decade is the value of nuance. There is usually a time and a place even for the worst strategies. There are, for example, some times when a bunt is appropriate, either statistically speaking, or just within the game state. So too is playing dump and chase. The point is that those two strategies are, on the whole, not wining plays outside of a few specific situations, and learning what those situations are and applying those strategies only when those situations appear during the game is the reason to learn about and use analytics in the first place.
It took a long time for baseball’s statistical revolution to take hold in the mainstream, but gradually, through efforts like those from the writers at Baseball Prospectus, over time, it has. Now baseball organizations are made fun of for not having an analytics department, or even for having one that is deemed too small. The battle is over. Information, knowledge, and learning have won. Not in every newspaper columnist’s mind perhaps, but everywhere that matters.
But this isn’t so yet in hockey. Hockey is the next frontier, and that is exciting. It’s entirely possible that when you, the analytical hockey fan, look back at this time in 12 years, you’ll look back at your understanding of the game and laugh. All the great advancements in hockey analytics will be kicking around in your head and you’ll think how important the last decade – the decade we all face now – was to understanding the great game of hockey. And you’ll smile and wish you could go back and live it again. Well you can. This is it. It is now. We are living hockey’s Moneyball moment.