Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: 1 ... 4 5 [6]  All

Author Topic: Probability paradoxes  (Read 33499 times)

0 Members and 1 Guest are viewing this topic.

theory

  • Administrator
  • *****
  • Offline Offline
  • Posts: 3603
  • Respect: +6125
    • View Profile
    • Dominion Strategy
Re: Probability paradoxes
« Reply #125 on: January 14, 2013, 09:26:42 am »
0

He got all the states right as well as their approximate margin of victory based on polling data (using past polling data as a comparison).

To be fair, many other political stats junkies came to the same result.  Nate Silver was just highest-profile because he is part of the NYT and was slammed by the far right.
Logged

jonts26

  • Margrave
  • *****
  • Offline Offline
  • Posts: 2746
  • Shuffle iT Username: jonts
  • Respect: +3671
    • View Profile
Re: Probability paradoxes
« Reply #126 on: January 14, 2013, 02:33:20 pm »
+1

What did he get 100% on for the US election though?

Because as far as I can see, the only thing he could be 100% on is predicting the states, and thats not very impressive. Especially in a thread about probability!

Saying somebody has an 80% chance of winning, and that person winning doesn't make him 100% right! He could in fact have been way off and that person could have had a 1% chance of winning, but did anyway.

As far as I can see, he presents a LOT of pretty stats, but the end result is pretty much the same as everyone else is saying and its not possible to quantify how right he was, let alone claim 100%


(His Baseball stuff may be very impressive though, have no idea on that)

It's more than just predicting which states go where, but really looking at margins of victory show him to be highly accurate. Also he did a really good job predicting senetorial and house races as well. Still, he often does get a bit more credit than probably deserves. He wasn't the only statistician to be so accurate, but after listening to the american media talk about what a horse race, toss up, and how close the election was, crap about gut feelings and momentum and a million other talking points, it's refreshing to see someone apply actual scientific models to the problem. And it's doubly nice that he gained fairly widespread recognition. And triply nice that his models performed so well.
Logged

ConMan

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1400
  • Respect: +1706
    • View Profile
Re: Probability paradoxes
« Reply #127 on: January 14, 2013, 05:44:21 pm »
0

I think the biggest reason he's celebrated now is because he strongly defended his models and their prediction of Obama's relatively easy win, compared to a large number of more conservative (in both senses of the word) commentators who seemed to be trying to drum up excitement by suggesting it was going to be neck-and-neck. And, yes, he guessed all of the states and got the margins of error pretty close, too.
Logged

ipofanes

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1439
  • Shuffle iT Username: ipofanes
  • Respect: +777
    • View Profile
Re: Probability paradoxes
« Reply #128 on: January 15, 2013, 04:01:17 am »
0

I think the biggest reason he's celebrated now is because he strongly defended his models and their prediction of Obama's relatively easy win, compared to a large number of more conservative (in both senses of the word) commentators who seemed to be trying to drum up excitement by suggesting it was going to be neck-and-neck. And, yes, he guessed all of the states and got the margins of error pretty close, too.

I think it means the margins of actual votes, though you technically can say something about the margins of errors too (estimate coverage probabilities).
Logged
Lord Rattington denies my undo requests
Pages: 1 ... 4 5 [6]  All
 

Page created in 0.042 seconds with 21 queries.