Last week both Dan and myself posted our game plans for the Blue Jackets offseason. Much theorizing has been done regarding how those rosters would fare, so I thought I would take a look at some statistics to see how they would do. Instead of just adding up goals, I checked out BehindtheNet.ca to look at the shots for and shot against rates for every player. Using stats from the last three years, I found the average rate at which each player’s team got shots on goal or allowed shots on goal. This gave me approximate shots for and against totals for the entire team over the course of the season. Using this, it was fairly easy to get a good approximation of how many goals the team would allow. To do this, I took the shots allowed at even strength and shorthanded, and multiplied those numbers by the goaltenders even strength and shorthanded save percentages. To get goals for, I used the shot rates for each player and multiplied those by their even strength and powerplay shot percentages. Shorthanded goals weren’t considered, as they are too random and too scarce for there to be a significant enough sample size (although a league-average number of shorthanded goals for and against was added to the final totals). So enough mumbo-jumbo, what does that all mean? It means a reasonably accurate portrayal of expected goals for and against for the Blue Jackets as constructed by Dan and I. The big question mark is minutes played. For every player I put together a reasonable range for games played and minutes played. Instead of an exact number, this gives us a reasonable range for next season. Without further ado, let’s see how our dream rosters would play out on the ice.
Not surprisingly, Dan and I both had a number of players that were on both rosters. Those players were RJ Umberger, Vinny Prospal, Derick Brassard, Brandon Prust (something of a surprise), Derek Dorsett, Derek Mackenzie, Ryan Russell, Colton Gillies, James Wisniewski, Jack Johnson, Marc Methot and Nikita Nikitin. That is three-quarters of our forwards and two-thirds of our defenseman that are basically playing identical roles and therefore producing identical results in this exercise.
Even beyond the identical players, Dan and I both built similarly constructed teams. Jordan Staal and David Krejci are similarly players, although there is a marked edge to Staal by the metrics used. However, that edge is almost identical to the improvement of Alex Semin over Brandon Dubinsky. Our second line is almost identical as well, with Nail Yakupov/Cam Atkinson being the only difference. Yakupov is expected to produce more than Cam, but there is also a good chance that Cam outproduces his rookie year significantly, which would bring him pretty close to the expected rookie year from a first overall pick. Our third lines are also a wash. Brad Marchand is a huge upgrade over Eric Belanger. However, Ryan Johansen is an upgrade over Derek Dorsett. This highlights a problem with this analysis, as zone starts and usage are not taken into consideration at all. Our fourth lines are also basically identical, with Dorsett and Letestu basically coming out the same as the only different players at the bottom of the forward corps.
As for the defensemen, Barrett Jackman somewhat surprisingly comes out way ahead of Marc Staal. For as much as Dan tends to avoid advanced stats, Barret Jackman is something of an advanced stat all star. Jackman’s team outshoots their opponents significantly when he is on the ice at even strength, and he does not allow very many shots on the penalty kill, all while playing against very tough opponents and with mostly defensive zone starts. Staal plays a similar role, just not nearly as effectively. However, the only other difference in our defense corps more than outweighs this difference. Fedor Tyutin looks very similar to Jackman by these metrics, and is productive on the powerplay to boot. On the other hand, John Moore struggled mightily this season. But like Atkinson, Moore could step up this year and outperform his rookie year.
The major difference between our two rosters is in net. The Harding-Clemmensen combo is better than the Niemi-Mason combo put together by Dan. Harding is slightly better than Niemi at even strength, Niemi has Harding slightly while shorthanded. Considering the much higher number of even strength shots seen by a goaltender, Harding gets the nod here. On the other hand, Clemmensen is a clear upgrade over Steve Mason. Clemmensen’s worst year over the past three is still better than the average of Mason’s last three years.
[table id=12 /]
ESGF = Even Strength Goals For, ESGA = Even Strength Goals Against, PPGF = Powerplay Goals For, SHGA = Shorthanded Goals Against, Goal Diff = Goal Differential, PTS = Expected or Actual Points over the season; NHL rank for the 2011-12 season is in brackets.
As you can see, this team would be significantly better defensively. It would not improve offensively, and would actually take a slight drop on the powerplay (however last years stats were increased due to the Jackets spending the second most time in the NHL with the man advantage). In case you were wondering, I used a variation of Pythagorean wins to calculate the points for the expected rosters. And for those skeptical of using math to estimate wins, based on the Blue Jackets performance last season their pythagorean points expectation was 68, while they finished with 65. So not a perfect formula, but accurate within 5 points. As Dan and I had predicted, these rosters would probably not put the team in the playoffs, but could have them within fighting distance. Los Angeles was the 8th seed with 95 points last year, but 9th place Calgary only had 90. These rosters should bring the Blue Jackets close to the playoffs and maybe in, if the threshold to get in lowers. Further, both of our squads were quite young with room for natural improvement in this coming season and the years after.