Timeline for Probability of a single real-life future event: What does it mean when they say that "Hillary has a 75% chance of winning"?
Current License: CC BY-SA 3.0
30 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 6, 2022 at 12:42 | comment | added | kjetil b halvorsen♦ | I think my answer here is relevant: Can we think of a probability in both the classical and subjective sense simultaneously? | |
| Dec 14, 2016 at 16:55 | comment | added | Aksakal | For practical purposes the statement means that the forecaster believes that Hillary's going to win strongly. The actual number 75% has no practical statistical meaning. They build a model to predict an event that will only happen once. So the probabilities are nonsense. | |
| Aug 1, 2016 at 0:54 | comment | added | icc97 | Nate Silver does a very good comparison to the 'feeling' of the 80% chance that you'll win a hand at poker youtu.be/mYIgSq-ZWE0?t=8m1s | |
| Jul 27, 2016 at 2:18 | comment | added | Maja Piechotka | @VinceO'Sullivan with frequentionist approach you're right. But in subjective, propensity or classical approach you can (though classical approach would be hard to apply). | |
| Jul 26, 2016 at 0:00 | comment | added | Chloe | See Methodology section: electionbettingodds.com/about.html | |
| Jul 25, 2016 at 23:36 | history | edited | Kodiologist | CC BY-SA 3.0 |
edited tags; edited title
|
| Jul 25, 2016 at 23:32 | comment | added | Robert Harvey | The phrase "Hillary has a 75% chance of winning" appears to derive from a CNN transcript in which Nate Silver made this declaration. His statistical methodology appears to be "the betting markets say so." Note that this is the same guy who also said, "I put Trump’s chances of becoming president at 25 percent, much higher today than a year ago I would have given a chance to win the nomination." | |
| Jul 25, 2016 at 22:36 | vote | accept | pitosalas | ||
| Jul 25, 2016 at 12:01 | history | protected | Scortchi♦ | ||
| Jul 25, 2016 at 11:41 | comment | added | Vince O'Sullivan | @immibis When talking about a one-off "thing" or event, you can't say that 75% will happen. Either they all (one of them) happen or none of them do. | |
| Jul 25, 2016 at 11:08 | comment | added | JimmyB | I'd read from that that, from polls, Clinton's expected result is to win, but the confidence interval of the numbers is such that there's as 25% chance that the actual outcome is not the same as the expected result. | |
| Jul 25, 2016 at 8:40 | comment | added | Glen_b | It depends on the source of the statement; in some cases it refers to a probability under some model, for example (as with the probability assessments on fivethirtyeight.com) but in other cases it relates to some other context it may mean something else. | |
| Jul 25, 2016 at 7:21 | answer | added | seanv507 | timeline score: 4 | |
| Jul 25, 2016 at 6:43 | answer | added | David Schwartz | timeline score: 6 | |
| Jul 25, 2016 at 2:32 | comment | added | Count Iblis | arxiv.org/abs/1212.0953 | |
| Jul 25, 2016 at 2:00 | answer | added | JTP - Apologise to Monica | timeline score: 3 | |
| Jul 24, 2016 at 23:46 | history | tweeted | twitter.com/StackStats/status/757361308933251074 | ||
| S Jul 24, 2016 at 21:22 | history | suggested | BlueRaja - Danny Pflughoeft | CC BY-SA 3.0 |
broken sentence
|
| Jul 24, 2016 at 21:12 | review | Suggested edits | |||
| S Jul 24, 2016 at 21:22 | |||||
| Jul 24, 2016 at 20:12 | answer | added | candied_orange | timeline score: 15 | |
| Jul 24, 2016 at 19:02 | comment | added | Silverfish | A key issue here is whether we can attach probabilities to unique (i.e. one-off) events, where we can't apply empirical probabilities in the manner of "if I roll a fair die a large number of times, the proportion of times I roll a six will approach one sixth". But there is an argument that mere subjective degree of belief still ought to behave in practice like a "probability" - more technically, should obey the axioms of probability. So a philosophical approach to this question might make reference to the so-called Dutch Book argument. | |
| Jul 24, 2016 at 18:06 | answer | added | Kodiologist | timeline score: 62 | |
| Jul 24, 2016 at 17:57 | answer | added | user41979 | timeline score: 2 | |
| Jul 24, 2016 at 16:10 | answer | added | Placidia | timeline score: 33 | |
| Jul 24, 2016 at 16:07 | answer | added | Matthew Drury | timeline score: 17 | |
| Jul 24, 2016 at 14:36 | history | edited | Silverfish |
add prediction tag
|
|
| Jul 24, 2016 at 14:05 | answer | added | a.powell | timeline score: 26 | |
| Jul 24, 2016 at 14:05 | comment | added | user78229 | Since the polls don't make probabilistic estimates and without further context, it sounds like that statement is based on the current results from one of the prediction markets, e.g., the Iowa Electronic Market (see tippie.uiowa.edu/iem). See their Methodology page or any of the many papers about prediction markets for deeper explanations. | |
| Jul 24, 2016 at 13:59 | review | First posts | |||
| Jul 24, 2016 at 14:36 | |||||
| Jul 24, 2016 at 13:59 | history | asked | pitosalas | CC BY-SA 3.0 |