Saw this article over on the Hockey News site today and realized that while Ken Campbell does a nice job of not completely ridiculing advanced stats, he missed out on an opportunity to define them properly.
How did the best possession team in the NHL miss the playoffs, and how did one of the worst get in?
The point he makes in the last paragraph is the one that seems to be a struggle for those arguing against the use of analytics:
Usually, analytics are right, but sometimes they’re wrong. In this case, they were wrong. The Kings should have made the playoffs and the Flames should have missed, according to analytics. But you can’t control luck. And when so many games are decided by a goal or two, luck plays a more pivotal factor. The one thing is, and the Avalanche proved that this season, that luck will change. Being a team that gives itself the best chance to win by possessing the puck and having the best chances to score has the better chance for long-term success.
See that’s the thing about statistics, they can’t ever predict the future with 100% accuracy. The possession numbers of the Kings are very good, and usually teams with good possession numbers will win more than their fair share of games. But, in the course of an individual game, series, or even season, random shit will still happen. That’s why they play the games, right? If you gave me the Kings possession numbers and played out the season 100 times, they probably make the playoffs well more than 75% of the time. (Probably higher in truth). But, thanks mostly to shootout luck, this season fell on the other side of that percentage. It happens. Underdogs win games that they would lose 99 times out of 100 if you played 100 times, yet we don’t argue that anyone who predicted the opposite outcome simply shut up, do we? Of course not, because their prediction was based on the best evidence available at the time, which is that one team is clearly better than the other. But, occasionally, randomness strikes and those teams lose. It’s why we love sports!
Analytics are no different in that regard. They can tell us what is more likely to happen, but sometimes, randomness strikes. The Kings would seem to be a victim of randomness this season. (And a complete inability to succeed in shootouts!)
So keep up with the stats and know that they will help predict what will happen over the long haul fairly accurately, and also enjoy it when something completely random happens and surprises us.
If anything, the Kings season is not evidence that we should stop looking at puck possession as much as evidence that the shootout punishes and rewards teams based on something that is not an integral part of the normal the course of a game. That’s what King fans should be angry about!
Leave a Reply