Page 14 of 64 FirstFirst ... 4121314151624 ... LastLast
Results 131 to 140 of 631

Thread: 2019 Sagarin Says

  1. #131
    Join Date
    Sep 2008
    Location
    Columbus, OH
    Posts
    15,077

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by Gully View Post
    Thanks Audit. Does this math assume the games are independent events from one another? In other words, losing or winning a particular game has no impact on the outcome of a different game?
    I think the entire data set "resets" every week. The predictor will change as the weeks progress and more and more data points are added.
    ¯\_(ツ)_/¯

  2. #132
    Join Date
    Oct 2003
    Location
    St. Paul
    Posts
    19,094

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by Gully View Post
    Thanks Audit. Does this math assume the games are independent events from one another? In other words, losing or winning a particular game has no impact on the outcome of a different game?
    Yes. Assumes games are independent events.
    I have the honor to be Your Obedient Servant - B.Aud

    We all live in stories... It seems to me that a definition of any living vibrant society is that you constantly question those stories... The argument itself is freedom. It's not that you come to a conclusion about it. Through that argument you change your mind sometimes... That's how societies grow. When you can't retell for yourself the stories of your life then you live in a prison... Somebody else controls the story. - S. Rushdie

  3. #133
    Join Date
    Oct 2003
    Location
    St. Paul
    Posts
    19,094

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by Bison 4 Life View Post
    I think the entire data set "resets" every week. The predictor will change as the weeks progress and more and more data points are added.
    This is correct also. My analysis is static. It’s performed at a point in time for all future games without updating for the “results” of simulated games. More sophisticated, chained models, like those at fivethirtyeight.com for example, update their assumptions dynamically as they progress chronologically through the simulation. So their model would simulated this Saturday’s games, feed those results back into their assessment of team strength and then use that new assessment to simulate the next week, etc. The result is a wider spread of possible outcomes, that is closer to the true distribution.

    Without the keys to Sagarin’s model I don’t think updating in that fashion is possible. With them it’s probably beyond my skills.
    I have the honor to be Your Obedient Servant - B.Aud

    We all live in stories... It seems to me that a definition of any living vibrant society is that you constantly question those stories... The argument itself is freedom. It's not that you come to a conclusion about it. Through that argument you change your mind sometimes... That's how societies grow. When you can't retell for yourself the stories of your life then you live in a prison... Somebody else controls the story. - S. Rushdie

  4. #134
    Join Date
    Sep 2008
    Location
    Columbus, OH
    Posts
    15,077

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by bisonaudit View Post
    This is correct also. My analysis is static. It’s performed at a point in time for all future games without updating for the “results” of simulated games. More sophisticated, chained models, like those at fivethirtyeight.com for example, update their assumptions dynamically as they progress chronologically through the simulation. So their model would simulated this Saturday’s games, feed those results back into their assessment of team strength and then use that new assessment to simulate the next week, etc. The result is a wider spread of possible outcomes, that is closer to the true distribution.

    Without the keys to Sagarin’s model I don’t think updating in that fashion is possible. With them it’s probably beyond my skills.
    Well, I think the rest of us would disagree with your modesty. You manage to update this weekly or nearly that based on the current data. It's literally one of the things I look forward to the most on Bville.
    ¯\_(ツ)_/¯

  5. #135
    Join Date
    Sep 2005
    Location
    Oak Ridge, NC
    Posts
    8,088

    Default Re: 2019 Sagarin Says

    Bison 4 Life, yes, I get that the predictor changes each week, I was asking about hte math given the current results.

    Thanks Audit.

    I wonder if they really are totally independent. For example, would NDSU losing on Saturday cause them to refocus, change some strategies, etc. that would increase their odds in the future? Or could a close loss be demoralizing and cause a team to spiral downward? Probably no way to know, just sharing the stream of thoughts/questions going through my mind.

    The real question we need answers to is how would beer and beyond meat sales on the concourse impact the outcomes in the FFD. I think El Chapo, Audit and GOB1SON should collaborate on an interdiciplinary research effort to answer this important question. I'll proofread the final paper as long as I'm supplied with ample Natural Light Hard Seltzer and Pizza Patrol.
    Get your BB tickets now!!!

  6. #136
    KC Bison is offline Senior Member Gets their mail at the West Parking Lot
    Join Date
    Feb 2007
    Location
    State of Missery
    Posts
    1,718

    Default Re: 2019 Sagarin Says

    If you multiply all the chances that we win each game together, you arrive at 41.1% or chance at going 12-0. To go 11-1 would somehow depend on our chance to win that game to determine our percentage.

  7. #137
    Join Date
    Oct 2003
    Location
    St. Paul
    Posts
    19,094

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by KC Bison View Post
    If you multiply all the chances that we win each game together, you arrive at 41.1% or chance at going 12-0. To go 11-1 would somehow depend on our chance to win that game to determine our percentage.
    Combinations and permutations.

    For 1 loss you need loss percentage for the games as well (1-win%). You’re looking for exactly one loss. It could be any of the 8 remaining games. So you multiply through the list 8 times substituting the loss% in for one game each time, and add up the 8 products to get the result.
    I have the honor to be Your Obedient Servant - B.Aud

    We all live in stories... It seems to me that a definition of any living vibrant society is that you constantly question those stories... The argument itself is freedom. It's not that you come to a conclusion about it. Through that argument you change your mind sometimes... That's how societies grow. When you can't retell for yourself the stories of your life then you live in a prison... Somebody else controls the story. - S. Rushdie

  8. #138
    Join Date
    Aug 2015
    Location
    Phoenix
    Posts
    88

    Default Re: 2019 Sagarin Says

    https://247sports-com.cdn.ampproject...s-136429740%2F

    I thought this article did a much better job of using metric to rank teams.

    Sent from my SM-N960U using Tapatalk

  9. #139
    Join Date
    Mar 2014
    Location
    Colorado Springs, CO
    Posts
    10,744

    Default Re: 2019 Sagarin Says

    Quote Originally Posted by Dstahn View Post
    https://247sports-com.cdn.ampproject...s-136429740%2F

    I thought this article did a much better job of using metric to rank teams.

    Sent from my SM-N960U using Tapatalk
    That would be a heck of a bracket. While JMU, SDSU, Montana State and Weber State would be on the opposite side of the Bison, NDSU would have Montana, Villanova and Towson on there's. Lots of good match-up possibilities.
    Hail the BISON!!!

  10. #140
    gumby013's Avatar
    gumby013 is offline Senior Member Gets their mail at the West Parking Lot
    Join Date
    Oct 2007
    Location
    Las Vegas
    Posts
    4,129

    Default Re: 2019 Sagarin Says

    Up to 33.

    76 - SDSU

    82 - JMU

    84 - Montana

    98 - UNI

    Has us as ~20 point favorites vs UNI at home.
    Looking good Billy Ray!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •