For any, like myself, wondering "Who is Ben Welsh" ?
Hello. My name is Ben Welsh. I'm an Iowan living in New York City.
I am a reporter, an editor and a computer programmer. My job is to use those skills, together, to find and tell stories.
I work at Reuters, the world's largest multimedia news provider, where I founded the organization's News Applications Desk. In that role, I lead the development of dashboards, databases and automated systems that benefit clients, inform readers, empower reporters and serve the public interest.
[...]
At a guess (my Telepathy/IP is weak today, I'm not reading dang at usual strength) .. the initially submitted title was "invented" for submission and didn't match the content title.
HN veers toward "the guts of the content w/out decoration" - limited additional information, framing, weasel words, perceived slanting, etc.
It's uncommon to name an author unless the author themself is an important part of "the story".
I personally have no issue with the original title, however it's not really for me (non US citizen) to judge whether the reporter in question has a name / identity that carries weight in US IT circles.
I don’t really know much about it, but remember it as being _fantastic_ journalism every time I encountered one of their articles. As a bonus, great infographics and interactive data visualizations.
I am certainly missing a lot of nuance here, but it seems to me Nate Silver managed to have his cake and eat it too. He surely got good money for selling FiveThirtyEight, and now that the buyer has erased the product, Nate can get back a huge chunk of its readers since he offers very similar analyses on his personal site. Sure, natesilver.net has less brand recognition than fivethirtyeight.com, but it's still decently well-known and can only go up from here.
I think the nuance is that it is notable historical articles about predictions and discussions of political elections, during a time when politics is quite at the fore-front of many people's minds
He may also finally shake off the comment trolls who piled onto him after 2016, seemingly blaming him for the election results (absurd but people are absurd).
After that election, a certain group would tirelessly work to discredit him any time his election predictions were not entirely one-sided.
It’s possible NS may have signed a contract saying that he cannot engage in elections prediction for X YZ months to same extent that he did with FiveThirtyEight.
> Electoral-vote.com is a website created by computer scientist Andrew S. Tanenbaum. The site's primary content was originally poll analysis to project election outcomes. Since the 2016 elections, the site also has featured daily commentary on political news stories.
Unfortunately most of the most important visualizations are broken in the archived version. Including the gun deaths visualization and I think the P-hacking interactive
It's kinda sad to know no one else will get to experience those interactive visualizations. Though its nice to see the approval comparison page still works
Curious why they're broken, as the wayback machine seems to be able to run javascript. Do the visualisations rely on a server (or some other assets not included in wayback machine's crawl)?
It's not about robots.txt but yes, the owners of 538 can just send a cease and desist letter to get them all immediately removed. Many sites that don't want to preserve history have done this already.
The robots.txt file should be used to restrict (and, in some cases, slow down) crawling at the time it is being crawled, not for SEO or for restricting access to mirrors or for any other purpose. It should never apply retroactively. (Unfortunately it is sometimes used badly despite this.)
People always use that link as reference to say that Internet Archive ignores robots.txt but it only actually says they are ignoring it for government sites. It suggests that they might do it for other sites in the future (of 2017), but does not actually say that that they have done it.
That first link is confusing; it seems to say they ended up removing the pages not because of a legal threat but because of robots.txt “automated”.
If archive.org can be manipulated to remove content either via legal threats or simple robots.txt it loses a significant portion of its societal value.
Did FiveThirtyEight really get that much right other than the 2016 election?
I remember thinking they were the best data journalists out there, and they had some nice visualizations but did their other predictions actually hold up?
I think it's the fivethirtyeight of of historical significance, and Disney is one of the largest and wealthiest companies on the planet. So it's just kinda like "whoa, this is stratospheric negligence" or "whoa, what is the reason for this... assuming they are not idiots?"
Also, they don’t any plans for the IP, and Nate would’ve paid above-market rate just to take over and preserve the content for posterity. He estimates that they deleted 200,000 hours of human labor.
This is just some Disney suits being extraordinarily petty.
Yes, just to add to this: in the article by Nate [0] he says that he tried to buy the IP but Disney refused because they were unhappy with some of his prior comments.
"I did approach Disney a year or two ago, through my agent, about acquiring the remaining IP. ...
We were told to basically get lost: ABC was annoyed with my critical public comments about their management of FiveThirtyEight. It apparently wasn’t a long conversation, so I don’t have a lot more color to report than that."
Here are some numbers roughly in the right ballpark: during the Disney era, which lasted about 10 years, FiveThirtyEight published about 20 stories a week. Let’s say that each story took about 20 hours to produce between research, writing, graphics and editing.3 Do the math, and that works out to about 200,000 person-hours of work that ABC News just deleted.
In a sense, nothing - and any other website should be archived, too.
In another sense, it's a journalistic source with information and commentary on past elections. Even aside from the political context that muddies the waters around or outright denies results, matters of public discourse on the web should not be ephemeral or subject to the decisions of the publication - they should be archived.
This is a great service for everyone who appreciates thoughtful analyses about politics. Losing FiveThirtyEight was a big loss, but this archive helps. Bravo!
But that would be a false attribution. The Internet Archive did not create the index, Ben did. And the Internet Archive is not hosting the index, Ben is.
Ah, yes, could be worded better, fairplay. Point is the Ben attribution isn't needed in that place to avoid unnecessary confusion about who that is etc.
If I wanted to get the complete WARC archive of 538 - how do you do this in a friendly way? No interest in history tracking, just want the last available version from Internet Archive.
This is why people don't really buy the "but he had Trump at 30%, you just don't understand statistics" apologist line. Sure he hedged in the dying days of the campaign (a cynic might think to try to protect his credibility), but the tone overall was of a person who comprehensively failed to understand the mood of the country from beginning to end.
Which is a problem because these election predictions are not just pure "mathematical models" and "data driven" like 538 would have had you believe. What mathematical model should be used? What data should and should not be used? At some point those things are based on the modeller's understanding of reality.
I think Nate did a phenomenal job calling out pollsters in that time. Since 538 was predominately a poll aggregator that did tricky stats to rank the reliability of each poll. I remember specifically an interview with him griping about some of the unusual data he was seeing from pollsters that made it look like, and I quote, 'Someone has their finger on the scales'
Perhaps critiquing statistical methods used by polling was something he was good at. I have no real opinion of his work there, which I didn't pay attention to.
But predicting an election requires a lot more than polling datasets and statistics textbooks. That's the problem that he made himself out to be an election prediction wizard, but really that was off the back of his good prediction in quite a bland and conventional election.
When things got slightly more spicy and reality diverged from his vaunted "models", his "data science" predictably fell in a heap. The worst thing is almost not even that he got it wrong, it's that he seemed incapable of recognizing that present reality was quite significantly different from the past data he had used to build his models. Even after being wrong in so many of these predictions. He just kept churning out these pieces about how Trump was probably finished this time.
Okay, this is clearly an LLM response, but for the sake of being polite:
> But predicting an election requires a lot more than polling datasets and statistics textbooks. That's the problem that he made himself out to be an election prediction wizard, but really that was off the back of his good prediction in quite a bland and conventional election.
> When things got slightly more spicy and reality diverged from his vaunted "models", his "data science" predictably fell in a heap
The models were correct in two elections - arguably three because a 30% chance means that an outcome will occur in thirty times out of hundred. That is not zero.
To the person who is running this LLM, please find better things to do with yourself.
He didn’t hedge at the end. Nate always writes the models before election season then doesn’t touch them apart from actual bug fixes. The model actually organically predicted 30%.
I still think that’s about accurate. Maybe it should’ve been 40%.
Everyone forgets that it was a pretty close election. Clinton could’ve won without the Comey announcement.
I think he did hedge (or "strategically bug fix"). The prediction for Trump went from IIRC around 15 to 30 in the last week or so. It was a big swing, IIRC with a lot of waffle around why it happened but not a lot of verifiable fact.
> I still think that’s about accurate. Maybe it should’ve been 40%.
It wasn't accurate. This is something people misunderstand about these predictions. If the 2016 election was held 100 times, Trump would have won 100 times. It's not the same as rolling dice.
These election predictions don't say that. They say something like "the observations I have agree with scenarios that have Clinton winning, 70% of the time". Which is fine and correct as far as his data and model goes, but none of those scenarios were the reality he was trying to predict. They are all just figments of the model though. Getting down to the brass tacks, he predicted Clinton would win, and he was wrong.
Which is fine, we just can't know anything about his process from that failure. Certainly we can't conclude that it was "accurate", since it was not. If we had a good sample of elections where he used the same process and built up a good record then sure.
That's the beauty of this brand of pseudoscience. Statistical predictions of singular events like a particular election are totally unfalsifiable. You can just say "I guess we live in 30% world" or whatever, every time.
> Statistical predictions of singular events like a particular election are totally unfalsifiable.
Yes. And the 2nd Law of Thermodynamics was just violated by millions of atoms within my lungs, that happened to increase in energy above the ambient average due to collisions. Clearly thermodynamics is pseudoscience, too!
To give you a trivial example: The simplest way I can put this is that turn out varies based on the weather[1], and turn out is skewed by party. So if it rains on election day you are going to get a different result, and that result can flip the outcome of the election if the election is close. So it’s kind of a nonsense to say. “Trump would have won 100 times out of 100”. Are you saying Nate Silvers model should have had a perfect meteorological model to predict the weather? Or are you saying the election wasn’t close? In which case you’re just wrong on the facts.
The 70% figure is saying “we know most of the information needed to determine what the outcome of the election will be but we don’t know everything so can’t be certain”. There is no process where you can know every factor that determines the result in advance with absolutely accuracy and I don’t know why people expect there would be.
It's not nonsense. What's nonsense is to say Nate's prediction for the election was accurate or correct. It trivially was not.
What it would be reasonable to say is if his model had correctly predicted the outcome of a significant sample of elections, then you could say his model has some accuracy or predictive power. But it still would never have been accurate or right in the specific instances it got wrong, that's just a misconception about how statistics and predictive models work. I hope this helps.
What are you even classifying as accurate or correct? Do you take every 51% prediction from FiveThirtyEight and if the result is a win you consider that forecast accurate? And every 49% prediction must result in a loss? This just not how statistical forecasts work.
>What it would be reasonable to say is if his model had correctly predicted the outcome of a significant sample of elections, then you could say his model has some accuracy or predictive power.
I don't know why you're couching that in a hypothetical, FiveThirtyEight has repeatedly done that exercise.
>But it still would never have been accurate or right in the specific instances it got wrong
It is core to the concept of a probability that the result is going to go the opposite way from the prediction sometimes! It's meaningless to call it "wrong".
That's where you're wrong, the election was very, very close. In fact, if roughly 40k voters (across three states) had switched from Trump to Hillary, she would have won, that's how close it was.
40k voters, that's really not very many. So it's hard to say whether Trump had a 30% chance of winning or 40% or whatever, but the election at most was a toss-up.
Many random events could have resulted in a different outcome.
You misunderstand my point. I am talking about the actual election that happened where these many random events that could have resulted in a different outcome did not happen. I was being a bit facetious maybe in my point. But the point is that the thing that is to be predicted is the actual real event that occurs in this universe. Silver made a prediction, and it was wrong.
"Oh but it was only a 70% prediction"
You can't 70% win an election. Silver's prediction was that Clinton would win, but he was not super confident about it. The prediction was wrong. He was right to not be super confident about it, but the prediction of who would win was still wrong.
Statistical likelihood is a measurement of the known data at the time. If you engage with the content otherwise then it's on you if you have the wrong takeaway. No one who makes a prediction based on a statistical model is going to be right every time. That doesn't mean it's not right to make a prediction. The statistical modeling can help you to be correct more often than not. And if you were going to be truly fair you would note that Nate in fact repeatedly said that it was still very much possible for Trump to win but that the current known polling data and other factors in his model pointed to a loss.
538's own post-mortem's on the event highlight that Trump was a very unusual candidate running in a very unusual election and as such the model was missing a lot of important information. They learned from the experience and adjusted the model going forward. Anyone complaining about that event is really just highlighting that they don't understand how statistical modeling works and are upset about how the model misled them or others which isn't Nate or 538's fault and is entirely on the consumer of their reporting. It's not like they didn't try to educate their consumers in their reporting.
Just want to say, I appreciate your pragmatic perspective on this. Nate Silver had one job: Predict who would win. And he failed at that. With lots of hand waving he can excuse himself but at the end of the day his visitors wanted an answer and he gave them the wrong answer.
For any, like myself, wondering "Who is Ben Welsh" ?
~ https://palewi.re/who-is-ben-welsh/Ben is one of my favorite people in the world of data journalism. He's the author of many excellent training courses in the field, including:
- https://github.com/palewire/first-python-notebook
- https://github.com/palewire/first-web-scraper
- https://github.com/palewire/first-graphics-app
(Submitted title was "Ben Welsh made an index of all FiveThirtyEight articles on the Internet Archive" - we've since changed it)
Cheers for the clarity, that'll help me look less weird wrt above comment to future historians of archived HN threads :-)
TBH I enjoyed looking up Ben and finding out what he's about and done in the past far more than I did just knowing there's a 538 archive on IA.
What do you think was perceived wrong with the old title?
At a guess (my Telepathy/IP is weak today, I'm not reading dang at usual strength) .. the initially submitted title was "invented" for submission and didn't match the content title.
HN veers toward "the guts of the content w/out decoration" - limited additional information, framing, weasel words, perceived slanting, etc.
It's uncommon to name an author unless the author themself is an important part of "the story".
I personally have no issue with the original title, however it's not really for me (non US citizen) to judge whether the reporter in question has a name / identity that carries weight in US IT circles.
Try megadoses of Vitamin C - organic, free-range, ethically sourced, of course.
It helped when I developed an allergy to healing crystals.
Insightful in spite of that difficulty :)
Can't believe Ben Welsh is not Welsh, and FiveThirtyEight has nothing to do with Wales
Don't click that link if you have any feelings of inadequacy about your achievements LOL
Couldn't figure out why archiving FTE aricles matters, but a quick search yields:
> Thousands of FiveThirtyEight articles seemingly vanish from the internet
https://www.editorandpublisher.com/stories/thousands-of-five...
And discussions here on hn:
ABC News has taken all FiveThirtyEight articles offline https://news.ycombinator.com/item?id=48152553
Disney erased FiveThirtyEight (article by Nate himself) https://news.ycombinator.com/item?id=48197703
I don’t really know much about it, but remember it as being _fantastic_ journalism every time I encountered one of their articles. As a bonus, great infographics and interactive data visualizations.
I am certainly missing a lot of nuance here, but it seems to me Nate Silver managed to have his cake and eat it too. He surely got good money for selling FiveThirtyEight, and now that the buyer has erased the product, Nate can get back a huge chunk of its readers since he offers very similar analyses on his personal site. Sure, natesilver.net has less brand recognition than fivethirtyeight.com, but it's still decently well-known and can only go up from here.
I think the nuance is that it is notable historical articles about predictions and discussions of political elections, during a time when politics is quite at the fore-front of many people's minds
He may also finally shake off the comment trolls who piled onto him after 2016, seemingly blaming him for the election results (absurd but people are absurd).
After that election, a certain group would tirelessly work to discredit him any time his election predictions were not entirely one-sided.
It was so funny, he mostly predicted the election correctly, but people confused probabilistic forecasting for sports lines.
It’s possible NS may have signed a contract saying that he cannot engage in elections prediction for X YZ months to same extent that he did with FiveThirtyEight.
He didn’t. He retained the IP of his election forecast models and he publishes the results from them on his newsletter, The Silver Bulletin
That’s a relief!
> Nate can get back a huge chunk of its readers
The downside is this furthers the divide between folks who pay for subscriptions and masses who get shoveled ad-powered slop.
https://electoral-vote.com is still up, running, and showing data. It has always been free and ad free.
https://en.wikipedia.org/wiki/Electoral-vote.com
> Electoral-vote.com is a website created by computer scientist Andrew S. Tanenbaum. The site's primary content was originally poll analysis to project election outcomes. Since the 2016 elections, the site also has featured daily commentary on political news stories.
He put up a substantial number of articles for free during the 2024 election, it felt less paywalled than much of the MSM these days.
He puts up a substantial number of articles today for free.
Unfortunately most of the most important visualizations are broken in the archived version. Including the gun deaths visualization and I think the P-hacking interactive
https://web.archive.org/web/20230205124354/https://fivethirt...
It's kinda sad to know no one else will get to experience those interactive visualizations. Though its nice to see the approval comparison page still works
https://web.archive.org/web/20241031232233/https://projects....
Curious why they're broken, as the wayback machine seems to be able to run javascript. Do the visualisations rely on a server (or some other assets not included in wayback machine's crawl)?
This is because whoever owns Fivethirtyeight now (ABC?) deleted the whole archive of articles on the site.
Don't we need more than an index of Archive.org because whomever controls the domain could robots.txt these out of existence if they wanted to?
It's not about robots.txt but yes, the owners of 538 can just send a cease and desist letter to get them all immediately removed. Many sites that don't want to preserve history have done this already.
Archive.org mostly ignores robots.txt
https://blog.archive.org/2017/04/17/robots-txt-meant-for-sea...
The robots.txt file should be used to restrict (and, in some cases, slow down) crawling at the time it is being crawled, not for SEO or for restricting access to mirrors or for any other purpose. It should never apply retroactively. (Unfortunately it is sometimes used badly despite this.)
People always use that link as reference to say that Internet Archive ignores robots.txt but it only actually says they are ignoring it for government sites. It suggests that they might do it for other sites in the future (of 2017), but does not actually say that that they have done it.
https://blog.archive.org/2018/04/24/addressing-recent-claims... which is a year later mentions that they have an automated process which is still following robots.txt for displaying old pages where the robots.txt was added later.
https://help.archive.org/help/using-the-wayback-machine/ does say they follow it for scraping, but this is phrased in such a way that would still be true for past sites whether or not they changed the policy. There is a page https://www.sysjolt.com/2021/archive-org-no-longer-honors-ro... which claims they don't follow it, but the site owner misspelled "robots" as "robot".
That first link is confusing; it seems to say they ended up removing the pages not because of a legal threat but because of robots.txt “automated”.
If archive.org can be manipulated to remove content either via legal threats or simple robots.txt it loses a significant portion of its societal value.
I'm not a soccer guy, but I still think the piece on Lionel Messi was awesome
https://web.archive.org/web/20140701122958/http://fivethirty...
Did FiveThirtyEight really get that much right other than the 2016 election?
I remember thinking they were the best data journalists out there, and they had some nice visualizations but did their other predictions actually hold up?
Yes, their forecasts had excellent calibration—e.g., events they said would happen 30% of the time actually happened about 30% of the time.
https://archive.is/sId82
https://github.com/fivethirtyeight/checking-our-work-data
I'm seeing a lot about this. What makes this situation different than any other website going offline?
I think it's the fivethirtyeight of of historical significance, and Disney is one of the largest and wealthiest companies on the planet. So it's just kinda like "whoa, this is stratospheric negligence" or "whoa, what is the reason for this... assuming they are not idiots?"
Also, they don’t any plans for the IP, and Nate would’ve paid above-market rate just to take over and preserve the content for posterity. He estimates that they deleted 200,000 hours of human labor.
This is just some Disney suits being extraordinarily petty.
Yes, just to add to this: in the article by Nate [0] he says that he tried to buy the IP but Disney refused because they were unhappy with some of his prior comments.
"I did approach Disney a year or two ago, through my agent, about acquiring the remaining IP. ...
We were told to basically get lost: ABC was annoyed with my critical public comments about their management of FiveThirtyEight. It apparently wasn’t a long conversation, so I don’t have a lot more color to report than that."
[0] https://news.ycombinator.com/item?id=48197703
Pride comes before the fall. Sorry, Nate.
https://www.natesilver.net/p/disney-erased-fivethirtyeight
In a sense, nothing - and any other website should be archived, too.
In another sense, it's a journalistic source with information and commentary on past elections. Even aside from the political context that muddies the waters around or outright denies results, matters of public discourse on the web should not be ephemeral or subject to the decisions of the publication - they should be archived.
This is a great service for everyone who appreciates thoughtful analyses about politics. Losing FiveThirtyEight was a big loss, but this archive helps. Bravo!
Github link:
https://github.com/palewire/fivethirtyeightindex.com
Related: Disney erased FiveThirtyEight
https://news.ycombinator.com/item?id=48197703
Love Ben but title can simply be: Index of FiveThirtyEight articles preserved by the Internet Archive
But that would be a false attribution. The Internet Archive did not create the index, Ben did. And the Internet Archive is not hosting the index, Ben is.
Ah, yes, could be worded better, fairplay. Point is the Ben attribution isn't needed in that place to avoid unnecessary confusion about who that is etc.
If I wanted to get the complete WARC archive of 538 - how do you do this in a friendly way? No interest in history tracking, just want the last available version from Internet Archive.
Those 2015-16 ones sure aged poorly, I'm reminded of this https://i.imgur.com/6Z9QQj3.jpeg
This is why people don't really buy the "but he had Trump at 30%, you just don't understand statistics" apologist line. Sure he hedged in the dying days of the campaign (a cynic might think to try to protect his credibility), but the tone overall was of a person who comprehensively failed to understand the mood of the country from beginning to end.
Which is a problem because these election predictions are not just pure "mathematical models" and "data driven" like 538 would have had you believe. What mathematical model should be used? What data should and should not be used? At some point those things are based on the modeller's understanding of reality.
I think Nate did a phenomenal job calling out pollsters in that time. Since 538 was predominately a poll aggregator that did tricky stats to rank the reliability of each poll. I remember specifically an interview with him griping about some of the unusual data he was seeing from pollsters that made it look like, and I quote, 'Someone has their finger on the scales'
Perhaps critiquing statistical methods used by polling was something he was good at. I have no real opinion of his work there, which I didn't pay attention to.
But predicting an election requires a lot more than polling datasets and statistics textbooks. That's the problem that he made himself out to be an election prediction wizard, but really that was off the back of his good prediction in quite a bland and conventional election.
When things got slightly more spicy and reality diverged from his vaunted "models", his "data science" predictably fell in a heap. The worst thing is almost not even that he got it wrong, it's that he seemed incapable of recognizing that present reality was quite significantly different from the past data he had used to build his models. Even after being wrong in so many of these predictions. He just kept churning out these pieces about how Trump was probably finished this time.
Okay, this is clearly an LLM response, but for the sake of being polite:
> But predicting an election requires a lot more than polling datasets and statistics textbooks. That's the problem that he made himself out to be an election prediction wizard, but really that was off the back of his good prediction in quite a bland and conventional election.
> When things got slightly more spicy and reality diverged from his vaunted "models", his "data science" predictably fell in a heap
The models were correct in two elections - arguably three because a 30% chance means that an outcome will occur in thirty times out of hundred. That is not zero.
To the person who is running this LLM, please find better things to do with yourself.
He didn’t hedge at the end. Nate always writes the models before election season then doesn’t touch them apart from actual bug fixes. The model actually organically predicted 30%.
I still think that’s about accurate. Maybe it should’ve been 40%.
Everyone forgets that it was a pretty close election. Clinton could’ve won without the Comey announcement.
I think he did hedge (or "strategically bug fix"). The prediction for Trump went from IIRC around 15 to 30 in the last week or so. It was a big swing, IIRC with a lot of waffle around why it happened but not a lot of verifiable fact.
> I still think that’s about accurate. Maybe it should’ve been 40%.
It wasn't accurate. This is something people misunderstand about these predictions. If the 2016 election was held 100 times, Trump would have won 100 times. It's not the same as rolling dice.
These election predictions don't say that. They say something like "the observations I have agree with scenarios that have Clinton winning, 70% of the time". Which is fine and correct as far as his data and model goes, but none of those scenarios were the reality he was trying to predict. They are all just figments of the model though. Getting down to the brass tacks, he predicted Clinton would win, and he was wrong.
Which is fine, we just can't know anything about his process from that failure. Certainly we can't conclude that it was "accurate", since it was not. If we had a good sample of elections where he used the same process and built up a good record then sure.
That's the beauty of this brand of pseudoscience. Statistical predictions of singular events like a particular election are totally unfalsifiable. You can just say "I guess we live in 30% world" or whatever, every time.
> Statistical predictions of singular events like a particular election are totally unfalsifiable.
Yes. And the 2nd Law of Thermodynamics was just violated by millions of atoms within my lungs, that happened to increase in energy above the ambient average due to collisions. Clearly thermodynamics is pseudoscience, too!
To give you a trivial example: The simplest way I can put this is that turn out varies based on the weather[1], and turn out is skewed by party. So if it rains on election day you are going to get a different result, and that result can flip the outcome of the election if the election is close. So it’s kind of a nonsense to say. “Trump would have won 100 times out of 100”. Are you saying Nate Silvers model should have had a perfect meteorological model to predict the weather? Or are you saying the election wasn’t close? In which case you’re just wrong on the facts.
The 70% figure is saying “we know most of the information needed to determine what the outcome of the election will be but we don’t know everything so can’t be certain”. There is no process where you can know every factor that determines the result in advance with absolutely accuracy and I don’t know why people expect there would be.
[1] https://www.sciencedirect.com/science/article/pii/S026137942...
It's not nonsense. What's nonsense is to say Nate's prediction for the election was accurate or correct. It trivially was not.
What it would be reasonable to say is if his model had correctly predicted the outcome of a significant sample of elections, then you could say his model has some accuracy or predictive power. But it still would never have been accurate or right in the specific instances it got wrong, that's just a misconception about how statistics and predictive models work. I hope this helps.
What are you even classifying as accurate or correct? Do you take every 51% prediction from FiveThirtyEight and if the result is a win you consider that forecast accurate? And every 49% prediction must result in a loss? This just not how statistical forecasts work.
>What it would be reasonable to say is if his model had correctly predicted the outcome of a significant sample of elections, then you could say his model has some accuracy or predictive power.
I don't know why you're couching that in a hypothetical, FiveThirtyEight has repeatedly done that exercise.
>But it still would never have been accurate or right in the specific instances it got wrong
It is core to the concept of a probability that the result is going to go the opposite way from the prediction sometimes! It's meaningless to call it "wrong".
It's so interesting to see how someone could so confidentially wrong and clearly show no knowledge of statistics.
That's where you're wrong, the election was very, very close. In fact, if roughly 40k voters (across three states) had switched from Trump to Hillary, she would have won, that's how close it was.
40k voters, that's really not very many. So it's hard to say whether Trump had a 30% chance of winning or 40% or whatever, but the election at most was a toss-up.
Many random events could have resulted in a different outcome.
You misunderstand my point. I am talking about the actual election that happened where these many random events that could have resulted in a different outcome did not happen. I was being a bit facetious maybe in my point. But the point is that the thing that is to be predicted is the actual real event that occurs in this universe. Silver made a prediction, and it was wrong.
"Oh but it was only a 70% prediction"
You can't 70% win an election. Silver's prediction was that Clinton would win, but he was not super confident about it. The prediction was wrong. He was right to not be super confident about it, but the prediction of who would win was still wrong.
Statistical likelihood is a measurement of the known data at the time. If you engage with the content otherwise then it's on you if you have the wrong takeaway. No one who makes a prediction based on a statistical model is going to be right every time. That doesn't mean it's not right to make a prediction. The statistical modeling can help you to be correct more often than not. And if you were going to be truly fair you would note that Nate in fact repeatedly said that it was still very much possible for Trump to win but that the current known polling data and other factors in his model pointed to a loss.
538's own post-mortem's on the event highlight that Trump was a very unusual candidate running in a very unusual election and as such the model was missing a lot of important information. They learned from the experience and adjusted the model going forward. Anyone complaining about that event is really just highlighting that they don't understand how statistical modeling works and are upset about how the model misled them or others which isn't Nate or 538's fault and is entirely on the consumer of their reporting. It's not like they didn't try to educate their consumers in their reporting.
Just want to say, I appreciate your pragmatic perspective on this. Nate Silver had one job: Predict who would win. And he failed at that. With lots of hand waving he can excuse himself but at the end of the day his visitors wanted an answer and he gave them the wrong answer.