R Statistics Hockey Analysis Project

Discussion in 'By The Numbers' started by ChiefGonzo, May 2, 2021.

  1. ChiefGonzo Registered User

    Joined:
    Mar 3, 2020
    Messages:
    578
    Likes Received:
    303
    Trophy Points:
    49
    SB Cash:
    $ 100,000
    Gender:
    Male
    Is anyone interested in helping with some R work from datasets found on kaggle? Have a statistics project due tonight, could use a tutor.

    The three proposed questions are as follows:

    1. Is there a trend of skaters and goaltenders entering the age at an earlier age and if so what kind of contribution or impact are they having? Are specific positions more specific than others?

    2. What are the trends of professional organizations drafting in the NHL with regard to physical attributes? Are there similar deviations from the decade previous where on average these smaller athletes are producing more?

    3. My final question if I can cover the above in good manner is whether countries over the decade are having their athletes drafted more frequently and is there overall data to show that they are developing better athletes who are making more impact compared to the previous decade? If so in a specific country like the United States, what state or province are these athletes from and is there economic data to back it?

    If this is too complex, I was considering looking at something like multiple regression – Time on ice vs. points for a handful of 1-10 prospects.

    Can provide more context and info just PM me, happy to compensate for your time if this is a bit much.
     
    Bear of Bad News likes this.
  2. RationalExpectations Registered User

    Joined:
    May 12, 2019
    Messages:
    1,812
    Likes Received:
    915
    Trophy Points:
    84
    SB Cash:
    $ 100,000
    Gender:
    Male
    Depends how your dataset looks like. Happy to help you if need be.

    1. You could run a regression on all players with dependent variable number of points, explanatory variable : dummies : Age of entry in the NHL, and Draft year +x, then control variable such as TOI, Team for which they play (this should absord the effect of playing for a team which already is in good shape), position at the draft...
    2. Not sure what you mean here but if the goal is to look at trends in drafting, I believe you may want to cut your sample draft round by draft round, with an average height and control for positions and age
    3. Fixed effects analysis, and not sure this is the most interesting question to look at to be honest

    For 1-10 propsect the sample is not enough
     
    DatsyukToZetterberg likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice
monitoring_string = "358c248ada348a047a4b9bb27a146148"