Monday, April 12, 2010

Skating Off Topic


Long time blog readers will know that I enjoy playing around with political stats and numbers.

Well, I also enjoy playing around with hockey stats and numbers, so with the NHL playoffs here (for Leaf fans out there, that's what happens after the regular season ends), I've gotten together with some friends and launched PuckProb.com.

Using our statistical model, we've simulated the NHL playoffs 10,000 times, allowing us to project out the probabilities of each team winning the cup, the best scorers to pick in your playoff pool, and the odds of Denis Coderre raising a stink about something that happens in the playoffs.

We'll be adding blog posts and commentary in the coming days and in the coming weeks. So check 'er out. And forward the link on to your friends (unless, of course, you want to use the info to beat them in your pool).

Labels: ,

9 Comments:

  • A great resource for the average poolie...thanks!

    By Anonymous Anonymous, at 12:47 a.m.  

  • Dan, you have now officially become the Nate Silver of Canada!

    By Blogger Prairie Fire, at 12:49 a.m.  

  • Very cool. Just don't cut back on your blogging time at CalgaryGrit.

    By Anonymous Sean, at 9:29 a.m.  

  • If you like that stuff, check out my work on PuckProspectus.com. They're owned by the same guys that do fivethirtyeight.com.

    They've got most of the big names in hockey stat analysis: Like Gabe Desjardins, Tom Awad and Iain Fyffe.

    We also do stuff for ESPN, you can check out my stuff here:
    http://search.espn.go.com/vollman/robert-vollman/

    By Blogger Robert Vollman, at 10:21 a.m.  

  • Dan, best of luck - sounds like a cool project and it employs two of your biggest passions. Bon chance!

    By Blogger Jacques Beau Vert, at 10:23 a.m.  

  • If you have the Habs with even a 0.001 chance of winning the Cup I do not buy the numbers. Enjoy your 5 extra games Dan. Could have had a chance in the first round but you could not beat the powerful Leafs on Saturday. What are the chances of failing to get more than 2 out of six points against 3 teams that 147 times out of 246 attempts the other team got the W?

    Other than that small commentary it looks interesting. I do feel I have to put the obligatory "Dan, you are so cool, what a great idea. You are like the George Clooney of bloggers".

    By Anonymous Anonymous, at 10:47 a.m.  

  • Thanks Robert,

    I actually stumbled across some of your stuff when I was browsing hockey articles back in the fall.

    I was like, "hey, that's the guy that always comments on my blog. Small world.".

    By Blogger calgarygrit, at 10:56 a.m.  

  • There's something funny happening in your model (unless I'm misreading it, which is always possible): looking, for example, at WAS-MTL, we see MTL with a likely 7-game exit, and we see WAS likely winning the cup in 16.6 games.

    But ... a minimum of 16 games are required to win the cup (sweeping every series in 4 games) ... so the numbers don't allow for WAS to spend as many as 5 games (let alone 7) to take down MTL.

    That is, once the model shows the likely number of games it will take a team to lose a series, it will take their opponent the same number of games to win the series. But this does not appear to be reflected in the modelling.

    By Blogger Paul, at 3:03 p.m.  

  • By Blogger mmjiaxin, at 8:19 p.m.  

Post a Comment

<< Home