NHL: Canucks Projections – The Sedins

TWIN PILLARS: The Sedins are rested, healthy and the rest of the NHL should be worried


Today, exclusively at PlayNow Sports, I am kicking off my Canucks season preview – and how sweet it is.

Last season I undertook to try and predict Vancouver Canucks goal totals in advance of the season, and I actually did really well. Just to toot my own horn, going to the spreadsheet I used and comparing it with the actual results: I had the Canucks scoring 57 power-play goals (they scored 57), I had them potting 178 goals at even-strength (they managed 177) and I had them marked down for 247 markers overall (they managed 249, on the strength of three more total shootout wins than I’d anticipated). This year, I’m doing it again and I’ve tweaked the model a bit to try and refine its ability to predict the goal totals of individual skaters (as opposed to getting the team right, but being totally wrong about how they’d get there).

Before we get into what the model expects in terms of offensive output from the Sedin twins in a lockout shortened season, let’s go over the methodology. First of all, this isn’t anything Earth shattering. The model is actually very simple: I take a player’s large sample goals/60 (usually three seasons, but occasionally four if the player has been injured) and apply it to their “expected ice-time” for this upcoming season. Obviously that requires making a lot of assumptions, but since Vigneault has been coaching the Canucks for seven years and the core has mostly stayed the same for the past three or four, it’s actually not that difficult to do. The wrinkles come in the form of younger players with little or no track record (Chris Tanev and Zack Kassian, for example). To calculate a “large sample G/60″ for those players – and this is new this year, because I was way too bearish on Cody Hodgson in last season’s projections – I simply estimated their AHL ice-time, found a G/60 rate in the AHL, applied Desjardin’s NHLE league translation number and combined it to their observable G/60 rate in the NHL: and voila.

The final wrinkle actually applied only to the Sedin twins. Last year’s model called for Henrik Sedin to score 22 goals (he managed 13) and it was easily the forecast’s biggest whiff. On reflection it’s clear that there’s a simple reason for this: basing an offensive projection for players who have won two of the past three Art Ross trophies solely on recent past performance will obviously produce a bullish forecast. For example, over the normal three year course of the “large sample” that I am using, if you look at Henrik’s even-strength goals, his Hart Trophy and Art Ross Trophy winning season in 2009/10 is a very clear outlier.

That outlier season skews Henrik’s large sample G/60 rate to such an extent that if I used it without correcting for his sh% burst in 2009/10, the model would call for Henrik to score at a 21 goal pace this season. That’s not completely outside the realm of the possible, but I figured I’d try something different in an effort to make my individual predictions more accurate for the twins. I applied this methodology to Henrik’s even-strength goal scoring and also to Daniel’s power-play goal scoring since that’s another obvious outlier season that I thought I’d best correct for. You can see for yourself why:

So the methodology here is again, very simple. I took both players shot-rate (SOG/60) from that particular season and applied it to their shooting percentage (over the past five seasons) to get a “true talent” goal tally for them in their percentage driven outlier seasons. I then plugged that new number in instead of their actual G/60 rate from their respective outlier seasons. I’m hopeful that this will give me a projection closer to their “true-talent”, iron out some of the noise and make my projections more accurate.

I’ll be rolling out these posts over the course of the week beginning with the twins, then moving onto second line forwards, third-line forwards, depth-forwards, defenseman and then team goal totals. At the end of the week I’ll show you my work in terms of the spreadsheet so you can critique and peer-review what I’ve done. (You can also bet the over/under on my projections thanks to the creative pricing team at PlayNowSports, which I’m really excited about).

Without further ado, here’s my projections for the Sedin twins goal scoring output in the 2013 season:

As you can see, even with my work to try and minimize the noise of Henrik’s percentage driven success in 2009-10, my model is still very bullish on his goal scoring output this season. Meanwhile 19 goals (or a 34 goal pace) for Daniel Sedin looks like it might be a tad optimistic – it would be the third highest goal total of his career – but it’s right around the pace he was at last season. While there are those who see the Sedins as already on the “down swing” of their careers, I remain confident that they’ve got a couple of seasons left as point-per-game players and that’s an interpretation that my model agrees with.

Are you excited about watching somewhere in the neighbourhood of 29 Sedin goals this upcoming season? Yeah, me too.