Timeline for How can we judge the accuracy of Nate Silver's predictions?
Current License: CC BY-SA 3.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 7, 2021 at 23:10 | comment | added | Aksakal | The bottom line is that the accuracy of his predictions is impossible to measure. He's in election data collection business truly. He collects a ton of information and shares with clients the processed data. The actual predictions are just a tiny bit of value he creates to his customers. So if he's wrong then the impact on his revenue is immaterial. | |
| Feb 2, 2017 at 19:04 | history | edited | amoeba |
edited tags
|
|
| Oct 9, 2016 at 8:17 | history | edited | Tim |
edited tags
|
|
| Oct 9, 2016 at 2:13 | answer | added | Glen_b | timeline score: 6 | |
| Oct 8, 2016 at 22:33 | answer | added | dpritch | timeline score: 13 | |
| Oct 8, 2016 at 19:55 | history | tweeted | twitter.com/StackStats/status/784844652180627456 | ||
| Oct 8, 2016 at 19:32 | comment | added | Micah | Silver's model gives a lot more than just a probability estimate — it gives an estimated victory margin, which is derived from win probabilities and victory margins for each of the 50 states. So it's giving a point estimate and error margin for 50 different measurements (albeit with some — probably high — degree of correlation among them), not just predicting a single binary outcome. | |
| Oct 8, 2016 at 18:45 | comment | added | prince_of_pears | I think this question is fundamentally flawed. My (limited) understanding of Mr. Silver's predictions are such that he predicts the outcome of a single event. Therefore, we must wait until the event occurs before knowing if the prediction was accurate or not. Since there is only one event, it follows that accuracy can be only 0% (perfectly wrong) or 100% (perfectly correct). The suggestion by @StephanKolassa is thorny since now predictions become correlated as new information is used, and yet they still apply to a single event. | |
| S Oct 8, 2016 at 18:37 | history | suggested | Rodrigo de Azevedo | CC BY-SA 3.0 |
minor edits
|
| Oct 8, 2016 at 17:38 | comment | added | Rodrigo de Azevedo | Why can't probabilities change over time? In a sports event, don't the odds change whenever a goal is scored or a home run is hit? | |
| Oct 8, 2016 at 17:36 | review | Suggested edits | |||
| S Oct 8, 2016 at 18:37 | |||||
| Oct 8, 2016 at 15:49 | comment | added | user1566 | This is probably a "scoring rule", but, for n events, multiply his probability for those events occurring and take the nth root to get an average sort of prediction rate (we assume he never makes 0% predictions). You can consider each daily probability as a separate prediction. | |
| Oct 8, 2016 at 13:51 | comment | added | Stephan Kolassa | @TalGalili: we can say at least something, using scoring rules - just as, e.g., we can say something about unobservable parameters that we estimate in regressions. | |
| Oct 8, 2016 at 13:50 | answer | added | Stephan Kolassa | timeline score: 16 | |
| Oct 8, 2016 at 13:35 | comment | added | Tal Galili | I suspect we cannot. One needs a golden-standard for making such assessment, and the best we have is only the observations from previous elections which are hard to compare (since every election would include alternative methods of sampling and voters behavior). But I am no expert in election surveys, so I am leaving this as a comment and not an answer :) | |
| Oct 8, 2016 at 13:33 | review | First posts | |||
| Oct 8, 2016 at 13:37 | |||||
| Oct 8, 2016 at 13:28 | history | asked | Dinosaur Soup | CC BY-SA 3.0 |