Sunday, July 19

Ian Gooding


Advanced stats, Duncan Keith, and Seattle and NHL expansion

We've officially reached the point in the summer where there's not a lot of hockey news. So I'll try not to ramble on, so to speak. There won't be as much fantasy news and advice in today's Ramblings – just a few topics covered more in-depth. Neil covered the analysis of the fantasy rankings in yesterday's Ramblings with some great points to ponder. But do you know where you can get a lot more of that information? The Dobber Draft Guide, which is coming soon.

Speaking of which (teaser time), I'll be writing about 20 players to invest in for long-term payoff. These are players who probably won't be ready for fantasy stardom this season but are worth investing in for the future. Rookies won't be included in this article, as they will be in a separate piece from another writer. This is the kind of piece that will target those of you in deep keeper leagues – or maybe even those of you who simply want to get a leg up for 2016 drafts.


In Friday's Vancouver Province, long-time writer Tony Gallagher had a thought-provoking piece on stats. For those of you who are very pro-advanced stats, it may come off as yet another attack article. But Gallagher, who has an interesting way with words (even more so in radio interviews, in fact), describes the main reason for the introduction of new NHL stats over the past 40 years as being for contract negotiations. So you have agents and lawyers to thank for your stats beyond goals and assists.

So with that, I'll give you my philosophy on advanced stats. Maybe because Demetri is not submitting a Contrarian article this week, I'm going to go a little contrarian myself.

When I first started playing online fantasy hockey around 10 years ago (I participated in numerous hockey pools before that), the basic stats were all that were needed to get by. But when I began looking for that edge to determine whether a particular player appeared to be trending upward or downward, I discovered the wonders of stats such as time on ice, shooting percentage, and even points per 60 minutes. Some interesting stuff that I would attempt to use in my articles whenever it made sense to.

The last few years, the advanced stats revolution brought us new stats such as Corsi (shot attempts 5 on 5), Fenwick (unblocked shot attempts 5 on 5), and PDO (sum of a team's even strength shooting percentage and even strength save percentage). Nowadays, we have entire articles devoted to one player, which include scatter plots showing everything from impact on linemates to percentage of high danger scoring chances. Whoa. I bet your statistics professor would be proud.

There's a term in sales known as "data dump." To picture it, imagine yourself buying a computer. You want something that you can use for documents and spreadsheets, surfing the internet, and maybe playing some games. So you tell the salesperson what you need this computer to do and what your budget is. The salesperson tells you how many GB (now TB) of storage the computer has, how much RAM the computer has, and that there are full HD graphics. Great, you're sold.