One thing that has always interested me in fantasy sports, and sports as a whole, is the consistency of statistics. Given a players previous stats, what can be expected for the upcoming season? Last year I wrote a fantasy baseball article in which I compared players who are traditional first round picks versus guys who are selected in rounds three to five. Not only did the top picks produce bigger raw numbers (more RBI, more home runs), but they were also more consistent. There was less variation in their statistics, which I consider a good thing.
I have decided to apply this theory to hockey, but in a different way. I have always thought that while player’s point totals are fairly consistent barring injury, powerplay points can actually fluctuate quite wildly. I decided to approach this by looking at 20 players who have played at least 70 games in each of the past four seasons (2005-2009), and have been pretty consistent in their roles on their teams, which would minimize any obvious discrepancies in point totals or powerplay points. The players are Joe Thornton, Ilya Kovalchuk, Jarome Iginla, Alex Ovechkin, Dany Heatley, Daniel Sedin, Zdeno Chara, Nicklas Lidstrom, Brian Campbell, Thomas Vanek, Pavel Datsyuk, Marc Savard, Martin St. Louis, Eric Staal, Shane Doan, Kimmo Timonen, Jay Bouwmeester, Henrik Sedin, Milan Hejduk and Dion Phaneuf.
What I did was add up all the points each player scored as well as all their powerplay points. These players scored 42.9% of their points on the powerplay. I then divided each player’s powerplay points by their total points in each year, giving me the percentage of their points that they scored on the powerplay. I then calculated the standard deviation of this number for each player. On average, the standard deviation was .075, meaning you can expect that drafting a comparable player will result in him scoring between 34.9% and 50.1% of his points on the powerplay. In terms of raw numbers, this is a range from 26 to 37 points, which is pretty significant.
Some players are more reliable than others. For example, Joe Thornton and Martin St. Louis had very small standard deviation’s of .038 and .023, while Brian Campbell and Shane Doan were much higher at .185 and .131, respectively. In fact, Campbell scored 61% of his 44 points on the powerplay in 2005-06, and only 19% of his 48 points in 2006-07 on the powerplay. Someone who follows the Sabres more closely may be able to explain why this is, but it boggles my mind.
While I don’t think this is a groundbreaking idea, I think it is worth mentioning that is it probably not a good idea to draft a player because he is a “proven powerplay performer”. You are probably just better off drafting players based on their expected point totals, and let the powerplay points flow naturally.