The Roundup

Flames PDO, the worst Canuck, Kessel needs time off, the cartoon above, Justin Jultz, Johnny Goalie, Las Vegas and more in this week’s Roundup.


In the Roundup we go around the Nation Network and give you some of the best articles of the last couple weeks. Every once in a while it’s good to know what other teams are doing, or maybe you find an interesting article you wouldn’t have found otherwise.

Vegas – Maybe Vegas!!




Kessel-Phil-1-MapleLeafs-2014





1280px-Shetland_Pony_on_Belstone_Common,_Dartmoor



MoralVictories




Hockeyfights.comScrap of the week


The Pacific War Room


  • ted

    this has to be asked:
    are the oilers tanking on purpose? the evidence seems to overwhelmingly suggest that yes they are.

    If the oilers are tanking on purpose, is that ethical? Is it cheating?

    • Rockmorton65

      I think they intentionally tanked the year before they drafted Hall. Ever since, they’ve just been a poorly run, bad hockey team. Once they realize they can’t compete, (usually by mid November) as an organization they seem fold their tent and stop trying. Their management seems to think the best course of action is to draft as many top 3 players as possible and one day they’ll magically be contenders. I think there is a key distinction between tanking and intentionally not trying to get better.

      Just an honest analysis, not taking shots.

      • SSB1963

        Again and I have asked this before. How do you get professional hockey players to tank? Especially when thier performances dictate whether they get a better / another / new contract.

        • Rockmorton65

          Its not the players that tank, it’s done through coaching. Coaches say, players do. Its very subtle, actually. It can be in the strategy – a team without a lot of high end skill, for example be told to play a run and gun style with Pittsburgh or try to play physical with LA. Things that can be defended like splitting up lines that are being productive or pulling a goalie with 2-3 minutes left. Or “experimenting” with a power play unit. That kind of stuff.

        • ted

          coaching decisions, who you play and when you play them can affect the outcome of a game.
          GM decisions, not picking up NHL players off waivers for a position the team is weak in and has no depth in.

  • Igor Ulanov 55

    RE: Everything lasts forever: Flames Riding the PDO Pony

    An R-squared of 0.34 is atrocious.

    That graph looks better than it is in part because points are largely clustered around the mean for a lot of teams (ie tight playoff race surrounding marginal playoff teams). As well as the fact the data bullets are pretty huge.

    This is a good concept but more ideas need to be explored here. The R-squared has to be above .80 otherwise this is just a joke.

    i) Increase sample size by adding multiple seasons.

    ii) Maybe looking at a multiple regression, (eg, create a model using PDO + Corsi close to explain point %). Especially seeing as there are about 8 teams with a Corsi close of 50% that range from (40-70%) of possible points. When you exclude the outlier (Buffalo?) that is essential the range of the points distribution meaning you’re going to have a tough time getting a credible r-squared with just Corsi close for the independent variable.

    Also, for credibility, you should be posting the standard deviation, leaving it out suggests that it’s big enough to completely undermine the model.