Henrik Sedin has been a hot pick in fantasy drafts this year. In Yahoo! Leagues, he is on average the 5th player off the board and has an Average Draft Position of 6.2. Despite the fact that he won the Hart Trophy and scoring title last year, I can’t bring myself to select him anywhere near the first round.
I am sure many other people feel the same way, and say things like “oh there is no way he can do it again.” They are right, but technically they don’t have anything to back up their argument. I do. It is a little something called shooting percentage.
My trade involves working on baseball statistics, so I am not going to pretend to know it all about hockey stats. There are some bright people out there who do that already like mc79hockey, Behind the Net and Puck Prospectus. However, there are some things I know that can be applied to hockey. In baseball, there is a statistic called BABIP (batting average on ball in play) which is essentially the rate at which balls hit in fair territory that are not home runs fall in for hits. For pitcher’s, this stat fluctuates wildly. One year Greg Maddux (one of the best pitcher’s off all time) will have one of the ten lowest BABIP’s in the league, and the next year will have one of the ten highest. There is essentially no year to year correlation in this stat for pitchers. There is very little a pitcher can do to control the rate at which balls in play fall in for hits.
Hopefully I didn’t lose you there; I promise I’m back to hockey for good. The hockey equivalent of BABIP is shooting percentage. Go ahead and pull up the career stats of your favourite NHL player and take a look at his year by year shooting percentages. I am willing to bet there isn’t much of a pattern. I ran a linear regression of 69 high scoring NHL players who played at least 41 games in each season over the last five years. The players range from Ovechkin to Horton to Erat, so we have quite a broad range. Setting 2009-10 shooting percentage as the dependant and the other four years as the independent variables, we get a r-squared of only .24, which isn’t very significant. For those of you unfamiliar with this process, it means that the previous four seasons of shooting percentage only explains 24% of the variation that we see in a player’s 2009-10 shooting percentage. The rest can be explained by quality of their linemates, quality of the shots, quality of the opposition, but mostly it is just plain, old-fashioned luck.
This brings me back to my topic of interest, Henrik Sedin. Last season, Henrik crushed his previous career high by potting 29 goals. Before last season, his shots bulged the twine 12.3% of the time. Last year, that number was the second highest of his career at 17.5%. Based on the equation from the regression, we would have only predicted a 13.4% shooting percentage (22 goals), although due to our lack of confidence, there is room for lots of variation. However, as I have been explaining, this stat fluctuates significantly from year to year. In Sedin’s case, his shooting percentage has ranged from a lowly 7.5% to 20.5%. To put that in perspective, 7.5% would have placed him among the bottom 80 forwards who played at least 50 games last season, and 20.5% would have ranked him #2 in the NHL behind Evgeny Artyukhin. Yep, Artyukhin lead NHL forwards in shooting percentage last season. He only shot the puck 42 times, so if we want to narrow our sample down to players who took at least 100 shots, our winner is Andrew Brunette at 19.4%
The fact is, you never know what you are going to get when it comes to shooting percentage*. It would be much safer to predict that Sedin will be somewhere around his career average of 13.1%, which if he shoots the puck the same number of times next season, could easily cost him 7 goals. That isn’t chump change for a guy who has only twice broken the 20 goal plateau.
*There are exceptions to the rule. For example, Derek Boogard has played 190 games over the last four seasons without scoring a goal. Due to his… um, style of play… it is likely that any future seasons in the NHL will also see a shooting percentage of 0.
Breakout seasons for players in their late 20’s are always a risk, but with Sedin it is pretty easy to see where the boost in production came from. Before you argue that a good chunk of his increased point totals came in the form of assists, just know that his two most common linemates, brother Daniel and Alexandre Burrows, had the 3rd highest and highest shooting percentages of their careers last season.
I realize that something like this takes a lot of the fun out of a sport that involves as much creativity as hockey, but I am just trying to help you out in your fantasy league. Henrik Sedin is still a fine hockey player, and could easily notch 80 points again next year, but another Art Ross Trophy is completely out of the question.