Data journalism
March 17, 2014 9:14 AM Subscribe
Nate Silver's FiveThirtyEight re-launched this morning. Opening manifesto. Building an NCAA bracket. An article about a computer program to count how many lines each pair of characters in “Romeo and Juliet” spoke to each other. Toilet seat covers. 2014 midterms. And why this winter is so miserable. Among other gems.
Interesting choice to jump straight into the "we're the other Fox news" branding thing.
posted by yoink at 9:20 AM on March 17, 2014
posted by yoink at 9:20 AM on March 17, 2014
Yay! This is great news; thank you very much.
(Who else among MeFites read Signal And Noise: Why Some Predictions Fail and Others Don't? I found it highly enjoyable---opinions?)
posted by seyirci at 9:23 AM on March 17, 2014 [7 favorites]
(Who else among MeFites read Signal And Noise: Why Some Predictions Fail and Others Don't? I found it highly enjoyable---opinions?)
posted by seyirci at 9:23 AM on March 17, 2014 [7 favorites]
Is anyone else getting a google can't find 538 message?
posted by Carillon at 9:24 AM on March 17, 2014
posted by Carillon at 9:24 AM on March 17, 2014
Nate Silver arranges his books by color. Whoa! This changes everything.
posted by yoink at 9:30 AM on March 17, 2014 [3 favorites]
posted by yoink at 9:30 AM on March 17, 2014 [3 favorites]
I had to refresh 4-5 times before it showed up. I do that because I've been having slow DNS recently, but maybe it's not just me. Mash that F5 like Tina Belcher on Facebook!
posted by benito.strauss at 9:31 AM on March 17, 2014 [1 favorite]
posted by benito.strauss at 9:31 AM on March 17, 2014 [1 favorite]
While I hit the site okay, the links in the post say "Page doesn't exist". Not quite ready for guests yet?
posted by benito.strauss at 9:33 AM on March 17, 2014
posted by benito.strauss at 9:33 AM on March 17, 2014
What does the fox know?
Ring-ding-ding-ding-dingeringeding!
As a fan of both Nate Silver and foxes (except Faux News), I am delighted with this relaunch.
posted by tilde at 9:36 AM on March 17, 2014
Ring-ding-ding-ding-dingeringeding!
As a fan of both Nate Silver and foxes (except Faux News), I am delighted with this relaunch.
posted by tilde at 9:36 AM on March 17, 2014
Oh man, the bracket predictions here and accompanying table are fun as hell.
posted by MCMikeNamara at 9:37 AM on March 17, 2014 [1 favorite]
posted by MCMikeNamara at 9:37 AM on March 17, 2014 [1 favorite]
An exciting bit from the manifesto: "Within a couple of months we’ll launch a podcast, and we’ll be collaborating with ESPN Films and Grantland to produce original documentary films."
Cannot wait!
posted by suprenant at 9:37 AM on March 17, 2014 [1 favorite]
Cannot wait!
posted by suprenant at 9:37 AM on March 17, 2014 [1 favorite]
Odd - it looks like DNS changes haven't propagated correctly yet. I'm still getting "can't find the server at www.fivethirtyeight.com" for all the links.
Looking forward to this when I can actually reach it.
posted by RedOrGreen at 9:37 AM on March 17, 2014
Looking forward to this when I can actually reach it.
posted by RedOrGreen at 9:37 AM on March 17, 2014
Wow, the guy who built the healthcare.gov website must be a very charismatic individual.
posted by any major dude at 9:39 AM on March 17, 2014 [4 favorites]
posted by any major dude at 9:39 AM on March 17, 2014 [4 favorites]
Sounds like DNS hasn't propagated completely; I'm not having any problem here.
And I encourage everyone to read the manifesto; it is quite good.
posted by Mitheral at 9:41 AM on March 17, 2014 [1 favorite]
And I encourage everyone to read the manifesto; it is quite good.
posted by Mitheral at 9:41 AM on March 17, 2014 [1 favorite]
Not working for me either.
posted by cell divide at 9:41 AM on March 17, 2014
posted by cell divide at 9:41 AM on March 17, 2014
The site only reveals itself to those it considers worthy.
posted by yoink at 9:43 AM on March 17, 2014 [13 favorites]
posted by yoink at 9:43 AM on March 17, 2014 [13 favorites]
Works for me. But then I also have bought and read his book.
posted by Roger Dodger at 9:55 AM on March 17, 2014 [3 favorites]
posted by Roger Dodger at 9:55 AM on March 17, 2014 [3 favorites]
From "What the Fox Knows": "You may have heard the phrase the plural of anecdote is not data. It turns out that this is a misquote. The original aphorism, by the political scientist Ray Wolfinger, was just the opposite: The plural of anecdote is data."
Which in turn links to this:
Which in turn links to this:
I've used the quotation "The plural of anecdote is not data" in various talks over the years, never knowing the original source. I searched the usual places (though clearly not hard enough!), but never figured out whom it should be attributed to. So I was pleased to learn that John Myles White had discovered the source: Raymond Wolfinger (presumably the political scientist from Berkeley). This attribution comes in this 2004 email from Fred Shapiro, editor of the Yale Dictionary of Quotations:posted by MonkeyToes at 9:57 AM on March 17, 2014 [10 favorites]
I [Shapiro] e-mailed Wolfinger last year and got the following response from him:
"I said 'The plural of anecdote is data' some time in the 1969-70 academic year while teaching a graduate seminar at Stanford. The occasion was a student's dismissal of a simple factual statement -- by another student or me -- as a mere anecdote. The quotation was my rejoinder. Since then I have missed few opportunities to quote myself. The only appearance in print that I can remember is Nelson Polsby's accurate quotation and attribution in an article in PS: Political Science and Politics in 1993; I believe it was in the first issue of the year."
Ha. Was just going to quote the same thing MonkeyToes did.
The article about polling in Crimea and Ukraine is interesting, too. This is an exciting launch. And Fox News doesn't actually use a fox, do they? It's a bit of a poke, I guess, but great that 538 is claiming one as its mascot.
One quibble with Silver's NCAA bracket-building piece. Early on Silver points out that one of the ratings systems he uses, the Moore ratings, "place a lot of emphasis on recent performance whereas the other systems do not," and then says he combines the computer ratings with "human rankings...in moderation" like preseason ranking, which he calls "a subjective estimate, but it nevertheless adds some value."
Just a few paragraphs later, he defends himself against possible charges that he might be favoring Michigan State because he grew up in East Lansing with the bold statement, "However, there is no subjective component to the FiveThirtyEight model."
I get his point - any possible hometown bias could not have entered into his ranking of Michigan State as #1 in the East - but that last statement can not only be considered overblown from a philosophy of science perspective, more importantly it directly contradicts his own previous statements in the article. It was surprising to see it so naked like that.
posted by mediareport at 10:13 AM on March 17, 2014
The article about polling in Crimea and Ukraine is interesting, too. This is an exciting launch. And Fox News doesn't actually use a fox, do they? It's a bit of a poke, I guess, but great that 538 is claiming one as its mascot.
One quibble with Silver's NCAA bracket-building piece. Early on Silver points out that one of the ratings systems he uses, the Moore ratings, "place a lot of emphasis on recent performance whereas the other systems do not," and then says he combines the computer ratings with "human rankings...in moderation" like preseason ranking, which he calls "a subjective estimate, but it nevertheless adds some value."
Just a few paragraphs later, he defends himself against possible charges that he might be favoring Michigan State because he grew up in East Lansing with the bold statement, "However, there is no subjective component to the FiveThirtyEight model."
I get his point - any possible hometown bias could not have entered into his ranking of Michigan State as #1 in the East - but that last statement can not only be considered overblown from a philosophy of science perspective, more importantly it directly contradicts his own previous statements in the article. It was surprising to see it so naked like that.
posted by mediareport at 10:13 AM on March 17, 2014
And Fox News doesn't actually use a fox, do they? It's a bit of a poke, I guess...
Another possible poke: Vox Media, the parent company of the other hot neojournalism startup of the moment, also uses an orange stylized foxhead logo.
posted by Iridic at 10:20 AM on March 17, 2014
Another possible poke: Vox Media, the parent company of the other hot neojournalism startup of the moment, also uses an orange stylized foxhead logo.
posted by Iridic at 10:20 AM on March 17, 2014
I knew those toilet seat covers were (mostly) a joke!
I just assumed their business model was based entirely on public bathroom desegregation manic-panic originating in the southern U.S, gradually making their way north and even into Canada because overly cautious facilities personnel spotted them in supply catalogs and figured they were A Thing.
posted by CynicalKnight at 10:23 AM on March 17, 2014
I just assumed their business model was based entirely on public bathroom desegregation manic-panic originating in the southern U.S, gradually making their way north and even into Canada because overly cautious facilities personnel spotted them in supply catalogs and figured they were A Thing.
posted by CynicalKnight at 10:23 AM on March 17, 2014
Just a few paragraphs later, he defends himself against possible charges that he might be favoring Michigan State because he grew up in East Lansing with the bold statement, "However, there is no subjective component to the FiveThirtyEight model."
I don't think this is actually a contradiction. The key question here is whose "subjectivity" is in play. While he acknowledges that the pre-season rankings are to some extent "subjective" the "subjectivity" in them is that of those organizations that compile them. When 538 builds its model, it simply plugs those pre-season ranking numbers in regardless of what 538's writers 'subjectively' think of them. Thus, from the perspective of 538 there is no element of "subjectivity" in the compiling of its rankings whatsoever--there is simply an algorithm that gets run with the various numbers plugged in as they find them.
Think of it like polling. A poll can be rigorously "objective" from the p.o.v. of the pollsters, even if the data it is collecting is entirely "subjective" from the p.o.v. of the people polled ("what is your opinion on X issue/person/etc."). 538 as "objectively" drawn on the "subjective" pre-season rankings created by a couple of different organizations.
posted by yoink at 10:24 AM on March 17, 2014 [1 favorite]
I don't think this is actually a contradiction. The key question here is whose "subjectivity" is in play. While he acknowledges that the pre-season rankings are to some extent "subjective" the "subjectivity" in them is that of those organizations that compile them. When 538 builds its model, it simply plugs those pre-season ranking numbers in regardless of what 538's writers 'subjectively' think of them. Thus, from the perspective of 538 there is no element of "subjectivity" in the compiling of its rankings whatsoever--there is simply an algorithm that gets run with the various numbers plugged in as they find them.
Think of it like polling. A poll can be rigorously "objective" from the p.o.v. of the pollsters, even if the data it is collecting is entirely "subjective" from the p.o.v. of the people polled ("what is your opinion on X issue/person/etc."). 538 as "objectively" drawn on the "subjective" pre-season rankings created by a couple of different organizations.
posted by yoink at 10:24 AM on March 17, 2014 [1 favorite]
We're live! But there's a 70.617854% chance you'll be able to see http://www.fivethirtyeight.com . The site is rolling out in stages. So sit tight!
posted by MCMikeNamara at 10:26 AM on March 17, 2014 [1 favorite]
posted by MCMikeNamara at 10:26 AM on March 17, 2014 [1 favorite]
Wonderful! Was looking exactly for something like this since... hum, since forever I guess.
posted by Riton at 10:38 AM on March 17, 2014
posted by Riton at 10:38 AM on March 17, 2014
Anyone else notice that the post about the site ("What the Fox Knows") was posted at 5:38 AM EDT?
posted by grubi at 10:39 AM on March 17, 2014 [9 favorites]
posted by grubi at 10:39 AM on March 17, 2014 [9 favorites]
This looks really incredible. I didn't know this was in the works, and I'm excited to make it a new regular internet stop.
posted by painquale at 10:40 AM on March 17, 2014
posted by painquale at 10:40 AM on March 17, 2014
The first thing I noticed is that unless I missed it--quite likely--was you have to subscribe to RSS feeds on each separate topic page instead of a full feed. (Yes, I'm a dinosaur.)
posted by immlass at 10:43 AM on March 17, 2014 [1 favorite]
posted by immlass at 10:43 AM on March 17, 2014 [1 favorite]
This is awesome. I'm looking forward to seeing him deconstruct the elections coming up (and Nate if you're watching please also start covering Canadian elections eh?)
Also is it just me or is Nate Silver all kinds of dreamy?
posted by feckless fecal fear mongering at 10:51 AM on March 17, 2014
Also is it just me or is Nate Silver all kinds of dreamy?
posted by feckless fecal fear mongering at 10:51 AM on March 17, 2014
Who else among MeFites read Signal And Noise: Why Some Predictions Fail and Others Don't?
I reviewed it -- very favorably -- for the Boston Globe.
Looking forward to the new site.
posted by escabeche at 10:51 AM on March 17, 2014 [1 favorite]
I reviewed it -- very favorably -- for the Boston Globe.
Looking forward to the new site.
posted by escabeche at 10:51 AM on March 17, 2014 [1 favorite]
That is good content and good clean presentation. He doesn't appear to have gotten his web coders from the same pool as Atlantic, New Republic, New York Times, or the like. Hooray for Nate!
posted by bukvich at 10:52 AM on March 17, 2014
posted by bukvich at 10:52 AM on March 17, 2014
DAGNABBIT I NEED THOSE BRACKETS SILVER ASHJASHDJHSAJDHAJSDJA
posted by Potomac Avenue at 10:53 AM on March 17, 2014
posted by Potomac Avenue at 10:53 AM on March 17, 2014
The first thing I noticed is that unless I missed it--quite likely--was you have to subscribe to RSS feeds on each separate topic page instead of a full feed. (Yes, I'm a dinosaur.)
You're ahead of me, I can't even figure out the RSS feeds on the topic pages. I really like the site so far though! I'm loving the Life and Science topics.
posted by yasaman at 11:06 AM on March 17, 2014 [1 favorite]
You're ahead of me, I can't even figure out the RSS feeds on the topic pages. I really like the site so far though! I'm loving the Life and Science topics.
posted by yasaman at 11:06 AM on March 17, 2014 [1 favorite]
The fox seems weird to me. Typical pundits are more foxy than Nate and his crew, who are exciting precisely because of the one big thing they know: that quantitative methods are widely applicable and underutilized.
posted by escabeche at 11:07 AM on March 17, 2014
posted by escabeche at 11:07 AM on March 17, 2014
I can't even figure out the RSS feeds on the topic pages...
Yeah, can someone who's good at internets crack the RSS for us? I know I could put it in my FB, but would rather put it in my feedly.
posted by OHSnap at 11:16 AM on March 17, 2014
Yeah, can someone who's good at internets crack the RSS for us? I know I could put it in my FB, but would rather put it in my feedly.
posted by OHSnap at 11:16 AM on March 17, 2014
immlass: "The first thing I noticed is that unless I missed it--quite likely--was you have to subscribe to RSS feeds on each separate topic page instead of a full feed. (Yes, I'm a dinosaur.)"
I *think* (might be wrong) that you can get almost everything through these two feeds (except maybe the "interactives"):
http://fivethirtyeight.com/features/feed/
http://fivethirtyeight.com/datalab/feed/
posted by crazy with stars at 11:17 AM on March 17, 2014 [3 favorites]
I *think* (might be wrong) that you can get almost everything through these two feeds (except maybe the "interactives"):
http://fivethirtyeight.com/features/feed/
http://fivethirtyeight.com/datalab/feed/
posted by crazy with stars at 11:17 AM on March 17, 2014 [3 favorites]
Oh, here's the interactives feed:
http://fivethirtyeight.com/interactives/feed/
posted by crazy with stars at 11:19 AM on March 17, 2014 [1 favorite]
http://fivethirtyeight.com/interactives/feed/
posted by crazy with stars at 11:19 AM on March 17, 2014 [1 favorite]
Hmm, they know one big thing, but they use it in many small places. Time for .... The Hedgefox!
posted by benito.strauss at 11:22 AM on March 17, 2014 [3 favorites]
posted by benito.strauss at 11:22 AM on March 17, 2014 [3 favorites]
So the secret to filling out a bracket is going chalk? Stupid statistics.
posted by Potomac Avenue at 11:37 AM on March 17, 2014 [1 favorite]
posted by Potomac Avenue at 11:37 AM on March 17, 2014 [1 favorite]
Ah, got it, you just add /feed to the topic pages. Clearly I got too used to the little RSS buttons.
posted by yasaman at 11:37 AM on March 17, 2014
posted by yasaman at 11:37 AM on March 17, 2014
Sometimes I participate in the bracket wagers at work even though I know nothing about sports. If I just fill out my bracket with Nate's predictions this year, a) is that cheating and b) will anyone at work notice?
posted by aka burlap at 12:29 PM on March 17, 2014
posted by aka burlap at 12:29 PM on March 17, 2014
Sometimes I participate in the bracket wagers at work even though I know nothing about sports. If I just fill out my bracket with Nate's predictions this year, a) is that cheating and b) will anyone at work notice?
It's publicly available knowledge and will be easy to find (given Nate's high profile in the prognostication biz, and given his connection to ESPN) for anyone remotely interested in leaning on professional opinion to construct their bracket. So I would say a) definitely not cheating and b) some people will certainly notice.
posted by yoink at 12:41 PM on March 17, 2014
It's publicly available knowledge and will be easy to find (given Nate's high profile in the prognostication biz, and given his connection to ESPN) for anyone remotely interested in leaning on professional opinion to construct their bracket. So I would say a) definitely not cheating and b) some people will certainly notice.
posted by yoink at 12:41 PM on March 17, 2014
Done and done, then!
posted by aka burlap at 12:54 PM on March 17, 2014
posted by aka burlap at 12:54 PM on March 17, 2014
Sad the Times couldn't find a place for someone like Silver doing data based reporting but seem perfectly happy with writers like Brooks and Friedman who pull facts out of their rears.
posted by octothorpe at 1:27 PM on March 17, 2014
posted by octothorpe at 1:27 PM on March 17, 2014
Sad the Times couldn't find a place for someone like Silver
The Times tried hard to keep Silver; they lost a bidding war for him. It would be hard to see how they could have given him the scope to set up a whole rival news/analysis site like this one, either. In the end it's probably good news that he got lured away from them.
posted by yoink at 1:30 PM on March 17, 2014 [2 favorites]
The Times tried hard to keep Silver; they lost a bidding war for him. It would be hard to see how they could have given him the scope to set up a whole rival news/analysis site like this one, either. In the end it's probably good news that he got lured away from them.
posted by yoink at 1:30 PM on March 17, 2014 [2 favorites]
Who else among MeFites read Signal And Noise: Why Some Predictions Fail and Others Don't?
Yeah, as someone in the modelling/prediction business, this is a great book relevant to both the lay reader and the professional. Highly recommended, and I'm thrilled that 538 is back from the grave of URLs I occasionally type in to see if there's new stuff. (And new Hark! A Vagrant!)
posted by Homeboy Trouble at 1:32 PM on March 17, 2014 [1 favorite]
Yeah, as someone in the modelling/prediction business, this is a great book relevant to both the lay reader and the professional. Highly recommended, and I'm thrilled that 538 is back from the grave of URLs I occasionally type in to see if there's new stuff. (And new Hark! A Vagrant!)
posted by Homeboy Trouble at 1:32 PM on March 17, 2014 [1 favorite]
seyirci: " (Who else among MeFites read Signal And Noise: Why Some Predictions Fail and Others Don't? I found it highly enjoyable---opinions?)"
Reading it now. Excellent book.
posted by zarq at 1:55 PM on March 17, 2014
Reading it now. Excellent book.
posted by zarq at 1:55 PM on March 17, 2014
As a poker playing guy who was once owned a company doing hard data and textual analysis to solve political and social-science-related problems who was trained in public policy... and has with a long-public disdain for opinion writers and pundits... I have to say I'm very excited about this new site and congratulate both Nate and ESPN for getting this project to this point.
But then I'm obviously as easy a sell on this as possible. Maybe the easiest sell on the planet LOL.
posted by mikel at 2:00 PM on March 17, 2014
But then I'm obviously as easy a sell on this as possible. Maybe the easiest sell on the planet LOL.
posted by mikel at 2:00 PM on March 17, 2014
from FiveThirtyEight: Fear not readers we have RSS feeds
posted by mcstayinskool at 2:12 PM on March 17, 2014 [1 favorite]
posted by mcstayinskool at 2:12 PM on March 17, 2014 [1 favorite]
Irrelevant to anybody but me, but on this same day, I received my 600-page full-color beautifully-bound volume of the entire "A Redtail's Dream" webcomic. It's a good day for foxes (and the morning after Cosmos on FOX explained evolution to the unwashed audience).
And I'm reading this on Firefox
posted by oneswellfoop at 2:15 PM on March 17, 2014 [1 favorite]
And I'm reading this on Firefox
posted by oneswellfoop at 2:15 PM on March 17, 2014 [1 favorite]
Thank you to all the folks who pointed me at the feeds! My RSS reader is happy now.
posted by immlass at 2:36 PM on March 17, 2014
posted by immlass at 2:36 PM on March 17, 2014
And okay, now that I'm reading the site in my feed reader, I'm not sure whether I'm more boggled by the color-coded books or by the fact that the Five Thirty-Eight staff doesn't seem to know the grammatical construction of the word "data" (which my teachers back in the dark ages incorrectly described as a collective noun) and that there's a difference between US and UK usage on which verb form to use with it.
posted by immlass at 3:23 PM on March 17, 2014
posted by immlass at 3:23 PM on March 17, 2014
the Five Thirty-Eight staff doesn't seem to know the grammatical construction of the word "data"
Sounds to me they just decided to go descriptivist rather than prescriptivist on a word that's slowly but steadily transitioning. Seems an appropriate strategy to take for a site devoted to observational statistical analysis.
posted by yoink at 3:34 PM on March 17, 2014
Sounds to me they just decided to go descriptivist rather than prescriptivist on a word that's slowly but steadily transitioning. Seems an appropriate strategy to take for a site devoted to observational statistical analysis.
posted by yoink at 3:34 PM on March 17, 2014
Seems an appropriate strategy to take for a site devoted to observational statistical analysis.
But a miss on some very low-hanging fruit for one devoted to using statistical observation in context. They're professional writers. You'd think that even if Silver wanted to poll to create a house style, he'd have a usage manual of some sort sitting around after working for the New York Times.
posted by immlass at 3:52 PM on March 17, 2014
But a miss on some very low-hanging fruit for one devoted to using statistical observation in context. They're professional writers. You'd think that even if Silver wanted to poll to create a house style, he'd have a usage manual of some sort sitting around after working for the New York Times.
posted by immlass at 3:52 PM on March 17, 2014
But a miss on some very low-hanging fruit for one devoted to using statistical observation in context.
If you read the whole piece it's clear that they did consult different style manuals and found that they differed on the issue. The NYT house style is notoriously rather conservative, so I wouldn't think should be allowed to rule on this issue for a new media venture like 538.
posted by yoink at 3:58 PM on March 17, 2014
If you read the whole piece it's clear that they did consult different style manuals and found that they differed on the issue. The NYT house style is notoriously rather conservative, so I wouldn't think should be allowed to rule on this issue for a new media venture like 538.
posted by yoink at 3:58 PM on March 17, 2014
Oh, and on some checking, even the NYT stylebook has given up on this issue.
posted by yoink at 4:02 PM on March 17, 2014 [2 favorites]
posted by yoink at 4:02 PM on March 17, 2014 [2 favorites]
I did indeed miss their consultation with style guides! But I don't particularly care which one Five Thirty-Eight uses as long as the house style is consistent. And I wouldn't pick the Times style guide either, because even without having read their style guide, I know their house style chaps my hide on a bunch of issues.
posted by immlass at 4:57 PM on March 17, 2014
posted by immlass at 4:57 PM on March 17, 2014
MonkeyToes: "The original aphorism, by the political scientist Ray Wolfinger, was just the opposite: The plural of anecdote is data.""
Holy cow. Holy cow! I love it.
posted by barnacles at 9:33 PM on March 17, 2014
Holy cow. Holy cow! I love it.
posted by barnacles at 9:33 PM on March 17, 2014
Very cool, thanks for posting. Only sports site in my twitter feed now. :)
posted by jeffburdges at 8:22 AM on March 18, 2014
posted by jeffburdges at 8:22 AM on March 18, 2014
It seems they did argue it out (and also conducted a poll for affirmation) and decided to go with 'data is'
posted by TwoWordReview at 12:04 PM on March 18, 2014
posted by TwoWordReview at 12:04 PM on March 18, 2014
"Sometimes I participate in the bracket wagers at work even though I know nothing about sports. If I just fill out my bracket with Nate's predictions this year, a) is that cheating and b) will anyone at work notice?"
Having just filled mine out, there are a couple of things to note: First off, his picks are almost all favorites (as they should be, given that it's probability), so they don't add that much value on top of just going chalk (a bit of jargon I honestly only learned today, so don't take me as an expert). Secondly, and more importantly, they don't make specific predictions past the first round. They can tell you that, hypothetically, Kansas has a 40 percent chance to go to the Sweet 16, but the actual odds will depend a lot on what team they face in the 32 round; Silver's odds just handicap the overall percentage — they don't tell you who to pick. Third, as he notes, they're almost certainly wrong in terms of outcomes. Like, I think he has MSU at 15 % to win the tourney. Sure, but that means an 85 % chance they don't. It's a little help, but maybe not as good a wager as looking at people who make direct picks (e.g. Dick Vitale). But then, even the best of those only called about 60 percent of the games right last year, so that's limited too.
posted by klangklangston at 5:39 PM on March 18, 2014 [2 favorites]
Having just filled mine out, there are a couple of things to note: First off, his picks are almost all favorites (as they should be, given that it's probability), so they don't add that much value on top of just going chalk (a bit of jargon I honestly only learned today, so don't take me as an expert). Secondly, and more importantly, they don't make specific predictions past the first round. They can tell you that, hypothetically, Kansas has a 40 percent chance to go to the Sweet 16, but the actual odds will depend a lot on what team they face in the 32 round; Silver's odds just handicap the overall percentage — they don't tell you who to pick. Third, as he notes, they're almost certainly wrong in terms of outcomes. Like, I think he has MSU at 15 % to win the tourney. Sure, but that means an 85 % chance they don't. It's a little help, but maybe not as good a wager as looking at people who make direct picks (e.g. Dick Vitale). But then, even the best of those only called about 60 percent of the games right last year, so that's limited too.
posted by klangklangston at 5:39 PM on March 18, 2014 [2 favorites]
Why Do We Expect So Much From Nate Silver?
posted by homunculus at 11:49 AM on March 19, 2014 [1 favorite]
posted by homunculus at 11:49 AM on March 19, 2014 [1 favorite]
Why Do We Expect So Much From Nate Silver?
Um...because of a fairly substantial track record?
posted by yoink at 11:56 AM on March 19, 2014
Um...because of a fairly substantial track record?
posted by yoink at 11:56 AM on March 19, 2014
going chalk (a bit of jargon I honestly only learned today, so don't take me as an expert).
For the record, I had heard the phrase before but still had to Google it to make sure I knew what it meant on Monday. I've noticed myself using four times since. This is why I am forgiving of people who I initially find insufferable because occasionally I am also the worst.
posted by MCMikeNamara at 11:57 AM on March 19, 2014 [1 favorite]
For the record, I had heard the phrase before but still had to Google it to make sure I knew what it meant on Monday. I've noticed myself using four times since. This is why I am forgiving of people who I initially find insufferable because occasionally I am also the worst.
posted by MCMikeNamara at 11:57 AM on March 19, 2014 [1 favorite]
I mostly enjoyed The Signal and the Noise, but thought Silver's section on climate change was a bit too much an attempt to placate both sides of the debate. I was hoping for better climate coverage from the new FiveThirtyEight site. Alas, no.
Roger Pielke, Jr. isn't exactly a climate skeptic, but he's kind of a contrarian slate-pitchy gadfly who seems to care more about needling the climate change movement than he does about the science. I think I'd rather have a straight-up skeptic there than someone who's just there to sling mud with questionable science.
posted by tonycpsu at 8:20 PM on March 19, 2014 [2 favorites]
Roger Pielke, Jr. isn't exactly a climate skeptic, but he's kind of a contrarian slate-pitchy gadfly who seems to care more about needling the climate change movement than he does about the science. I think I'd rather have a straight-up skeptic there than someone who's just there to sling mud with questionable science.
posted by tonycpsu at 8:20 PM on March 19, 2014 [2 favorites]
From tonycpsu's link, First Climate Article On Nate Silver’s Data Website Uses ‘Deeply Misleading’ Data, Climatologists Say:
But just as Pielke’s article has been written before, so too it has been criticized before. Dr. Kevin E. Trenberth, a distinguished senior climate scientist at the National Center for Atmospheric Research, has criticized Pielke’s data for its simplistic nature. Simply showing that an increase in damage has corresponded to an increase in wealth ignores the fact that communities are now more prepared than ever for extreme storms, Trenberth wrote at the time. Trenberth says data on increased disaster preparation measures should, to some degree, cancel out Pielke’s findings.
“This is the same old wrong Roger,” Trenberth said by e-mail. “He is demonstrably wrong and misleads.”
Mann agrees that the data analysis is too simplistic. “Pielke, in this case, continues to use an extremely controversial ‘normalization’ procedure when analyzing these data...That procedure assumes that damages increase with population but it completely ignores technological innovations (sturdier buildings, hurricane-resistant structures, better weather forecasting, etc.) that have served to reduce societal vulnerability, thus likely masking some of the aggravating impacts of climate change.”
...Some have pointed out that Pielke’s own study does not support the claim that he makes from it — that the most damaging extremes won’t be detectable in weather statistics for many decades. The 3-year-old research involves only rare, land-falling tropical cyclones, and looks only at the damage data from those cyclones. The study also explicitly ignores future rising sea-levels from climate change, which research from the National Oceanic and Atmospheric Administration has shown highly contributes to damaging storm surges.
Dr. John Abraham, a thermal science professor at the University of St. Thomas famous for his formation of the Climate Science Rapid Response Team, criticized Pielke’s assessment of the IPCC’s report. “You should know that we have already detected significant increases in Atlantic hurricane intensity, in extreme heat waves, large precipitation events and regional droughts,” Abraham wrote in an e-mail to Nate Silver expressing his disdain for the article, forwarded to Climate Progress.
“It’s ludicrous to say that extremes have not increased, and they have certainly increased in ways that are completely consistent with expectations based on atmospheric physics and climate model projections in response to increasing greenhouse gases,” Jennifer Francis, a research professor at Rutgers University’s Institute of Marine and Coastal Science who specializes in the connection between climate change and extreme weather, said...
“This post is surprisingly sloppy,” Abraham said. “I wouldn’t accept this kind of writing from my own students, even undergraduates.”
Great, just what we need. Another editor publishing sloppy work in the name of contrarianism. I look forward to Silver's response.
posted by mediareport at 7:48 AM on March 20, 2014
But just as Pielke’s article has been written before, so too it has been criticized before. Dr. Kevin E. Trenberth, a distinguished senior climate scientist at the National Center for Atmospheric Research, has criticized Pielke’s data for its simplistic nature. Simply showing that an increase in damage has corresponded to an increase in wealth ignores the fact that communities are now more prepared than ever for extreme storms, Trenberth wrote at the time. Trenberth says data on increased disaster preparation measures should, to some degree, cancel out Pielke’s findings.
“This is the same old wrong Roger,” Trenberth said by e-mail. “He is demonstrably wrong and misleads.”
Mann agrees that the data analysis is too simplistic. “Pielke, in this case, continues to use an extremely controversial ‘normalization’ procedure when analyzing these data...That procedure assumes that damages increase with population but it completely ignores technological innovations (sturdier buildings, hurricane-resistant structures, better weather forecasting, etc.) that have served to reduce societal vulnerability, thus likely masking some of the aggravating impacts of climate change.”
...Some have pointed out that Pielke’s own study does not support the claim that he makes from it — that the most damaging extremes won’t be detectable in weather statistics for many decades. The 3-year-old research involves only rare, land-falling tropical cyclones, and looks only at the damage data from those cyclones. The study also explicitly ignores future rising sea-levels from climate change, which research from the National Oceanic and Atmospheric Administration has shown highly contributes to damaging storm surges.
Dr. John Abraham, a thermal science professor at the University of St. Thomas famous for his formation of the Climate Science Rapid Response Team, criticized Pielke’s assessment of the IPCC’s report. “You should know that we have already detected significant increases in Atlantic hurricane intensity, in extreme heat waves, large precipitation events and regional droughts,” Abraham wrote in an e-mail to Nate Silver expressing his disdain for the article, forwarded to Climate Progress.
“It’s ludicrous to say that extremes have not increased, and they have certainly increased in ways that are completely consistent with expectations based on atmospheric physics and climate model projections in response to increasing greenhouse gases,” Jennifer Francis, a research professor at Rutgers University’s Institute of Marine and Coastal Science who specializes in the connection between climate change and extreme weather, said...
“This post is surprisingly sloppy,” Abraham said. “I wouldn’t accept this kind of writing from my own students, even undergraduates.”
Great, just what we need. Another editor publishing sloppy work in the name of contrarianism. I look forward to Silver's response.
posted by mediareport at 7:48 AM on March 20, 2014
Um...because of a fairly substantial track record?
Well, a track record that lots of other folks, including Sam Wang at Princeton, routinely match.
I don't think this is actually a contradiction. [in Silver's NCAA bracket piece]
I think it was a very clear contradiction, and you're making a distinction without much of a difference - one Silver himself doesn't make in the piece. Again, leaving aside the larger issues with a statement like "our statistical model has no subjective component" from an epistemology of science perspective (issues I've never see Silver deal with directly, although I haven't read his book), Silver is very clear:
The model, which is in its fourth year, is principally based on a composite of five computer power ratings...the Moore ratings place a lot of emphasis on recent performance whereas the other systems do not...These computer ratings are combined with a couple of human rankings...Preseason rankings [are] a subjective estimate, but it nevertheless adds some value, based on our research...However, there is no subjective component to the FiveThirtyEight model.
I get that you like him a lot, yoink, but it's ok to think Silver slipped a bit and showed just a tiny bit too much hubris there. He's human.
posted by mediareport at 7:59 AM on March 20, 2014
Well, a track record that lots of other folks, including Sam Wang at Princeton, routinely match.
I don't think this is actually a contradiction. [in Silver's NCAA bracket piece]
I think it was a very clear contradiction, and you're making a distinction without much of a difference - one Silver himself doesn't make in the piece. Again, leaving aside the larger issues with a statement like "our statistical model has no subjective component" from an epistemology of science perspective (issues I've never see Silver deal with directly, although I haven't read his book), Silver is very clear:
The model, which is in its fourth year, is principally based on a composite of five computer power ratings...the Moore ratings place a lot of emphasis on recent performance whereas the other systems do not...These computer ratings are combined with a couple of human rankings...Preseason rankings [are] a subjective estimate, but it nevertheless adds some value, based on our research...However, there is no subjective component to the FiveThirtyEight model.
I get that you like him a lot, yoink, but it's ok to think Silver slipped a bit and showed just a tiny bit too much hubris there. He's human.
posted by mediareport at 7:59 AM on March 20, 2014
Um...because of a fairly substantial track record?
It has to be more than that, because the things he made his name predicting are just not very hard to predict (though of course the public might not be aware of that). His gifts aren't so much in being a wackily advanced modeler -- most (but far from all) of what he does is first-year graduate student stuff. What he's good at are finding things that the public is interested in learning about, and explaining what he's doing in a clear, nonpatronizing but accessible way.
Both of which are things where political science fell flat on its face. We're bad at explaining ourselves to laymen because we don't do it very often, and are grade-A dorks, and we basically ignored the whole issue of near-term election prediction because, well, it just isn't very interesting or challenging to us. Even though obviously there's tremendous demand for it.
posted by ROU_Xenophobe at 9:05 AM on March 20, 2014 [1 favorite]
It has to be more than that, because the things he made his name predicting are just not very hard to predict (though of course the public might not be aware of that). His gifts aren't so much in being a wackily advanced modeler -- most (but far from all) of what he does is first-year graduate student stuff. What he's good at are finding things that the public is interested in learning about, and explaining what he's doing in a clear, nonpatronizing but accessible way.
Both of which are things where political science fell flat on its face. We're bad at explaining ourselves to laymen because we don't do it very often, and are grade-A dorks, and we basically ignored the whole issue of near-term election prediction because, well, it just isn't very interesting or challenging to us. Even though obviously there's tremendous demand for it.
posted by ROU_Xenophobe at 9:05 AM on March 20, 2014 [1 favorite]
That's not to say that he's not more knowledgeable than a first-year graduate student. Other things he does make his methodological depth plain. But the models he used in his election predictions in 2008 and 2012 are (except in a couple of ways) pretty simple... which is a very sensible approach for a popular-press person to take.
posted by ROU_Xenophobe at 9:08 AM on March 20, 2014
posted by ROU_Xenophobe at 9:08 AM on March 20, 2014
ROU_Xenophobe has it. Nate Silver isn't famous because he's an amazing statistician -- lots of people have the statistical skills he has, as he has often pointed out. He's famous because he can write, which is a skill very few people have.
posted by escabeche at 10:02 AM on March 20, 2014
posted by escabeche at 10:02 AM on March 20, 2014
Well, a track record that lots of other folks, including Sam Wang at Princeton, routinely match.
I'm not talking about his track record of election predictions. I'm talking about his track record of producing very compelling journalism based on statistical analysis--you know, the thing that this new site is doing.
I think it was a very clear contradiction, and you're making a distinction without much of a difference
Then you're not thinking hard enough. It is perfectly possible to produce "non-subjective" analyses using "subjective" inputs--this is simply a fact; every scientific poll ever conducted is an example. Silver doesn't have to "make that distinction" in his piece anymore than someone writing a story about the manufacture of a new plane has to explain how powered flight is possible at all.
posted by yoink at 10:12 AM on March 20, 2014 [2 favorites]
I'm not talking about his track record of election predictions. I'm talking about his track record of producing very compelling journalism based on statistical analysis--you know, the thing that this new site is doing.
I think it was a very clear contradiction, and you're making a distinction without much of a difference
Then you're not thinking hard enough. It is perfectly possible to produce "non-subjective" analyses using "subjective" inputs--this is simply a fact; every scientific poll ever conducted is an example. Silver doesn't have to "make that distinction" in his piece anymore than someone writing a story about the manufacture of a new plane has to explain how powered flight is possible at all.
posted by yoink at 10:12 AM on March 20, 2014 [2 favorites]
Then you're not thinking hard enough.
Well, that was pleasant. Luckily, the issue here isn't whether or not it's possible to "produce 'non-subjective' analyses using 'subjective' inputs." In order to believe it was, you'd have to argue that "The model" at the beginning of Silver's NCAA piece which is described as having subjective components is something different from "The 538 model" he says later in the same piece is completely free of subjective components.
I don't find that plausible in a piece aimed at laypeople. More plausible is that in the rush to get the site up and running, Silver made a few errors, and allowed some soft thinking of his own (or of his hires) to appear. Again, it's a quibble, since he does acknowledge that his model indeed contains multiple subjective components, and I like to think he would have agreed to a rewording if an editor above him had pointed out the possible confusion for lay readers.
More importantly, I'm not alone in finding that Silver and his hires have been publishing some rushed and sloppy work in the initial push. This piece from The Week talks about the idiocy of Silver's claim to Total Non-Subjectivity, with lots of good links to pieces debunking particularly ridiculous examples of Silver's first-round offerings:
Nate Silver's FiveThirtyEight and the dangers of being ideologically neutral
What went wrong? One major problem has to do with ideology. In an attempt to focus solely on objective analysis, Silver is ignoring one of the hardest-won journalistic lessons of the last decade — there is no such thing as ideology-free journalism...
FiveThirtyEight's science coverage stinks of sublimated ideology. The opening science pieces were pretty bad, but far more telling was Silver's hiring of climate troll Roger Pielke Jr.
He covers that a bit, but be sure to read the bizarre original piece debunked in the "pretty bad" link (it was 538's lead science story at launch). "Gut feeling" is a key element of the model in use there, and remember, this was a Silver hire's first foray on a site that insists it's committed to "non-subjective" science journalism. Is it a deliberate troll? A tongue-in-cheek thing that fell flat? Who the hell knows. It sure doesn't read like one. Anyway, back to The Week:
Here's where we find Silver's ideological commitment, I think: contrarianism. Note how he hates opinion columnists in part because they aren't original. This quite awful piece is absolutely dripping with the stuff. (Read a thorough debunking here.)
...Sigh.
I had been under the impression that media journalist Jay Rosen had successfully pushed the pitfalls of "ideology-free" journalism into the mainstream, but apparently we've got to learn it all over again. If Nate Silver wants to patch up his stumbling enterprise, and not just make an extremely expensive Freakonomics knockoff, that's where I would start.
posted by mediareport at 8:22 AM on March 23, 2014 [6 favorites]
Well, that was pleasant. Luckily, the issue here isn't whether or not it's possible to "produce 'non-subjective' analyses using 'subjective' inputs." In order to believe it was, you'd have to argue that "The model" at the beginning of Silver's NCAA piece which is described as having subjective components is something different from "The 538 model" he says later in the same piece is completely free of subjective components.
I don't find that plausible in a piece aimed at laypeople. More plausible is that in the rush to get the site up and running, Silver made a few errors, and allowed some soft thinking of his own (or of his hires) to appear. Again, it's a quibble, since he does acknowledge that his model indeed contains multiple subjective components, and I like to think he would have agreed to a rewording if an editor above him had pointed out the possible confusion for lay readers.
More importantly, I'm not alone in finding that Silver and his hires have been publishing some rushed and sloppy work in the initial push. This piece from The Week talks about the idiocy of Silver's claim to Total Non-Subjectivity, with lots of good links to pieces debunking particularly ridiculous examples of Silver's first-round offerings:
Nate Silver's FiveThirtyEight and the dangers of being ideologically neutral
What went wrong? One major problem has to do with ideology. In an attempt to focus solely on objective analysis, Silver is ignoring one of the hardest-won journalistic lessons of the last decade — there is no such thing as ideology-free journalism...
FiveThirtyEight's science coverage stinks of sublimated ideology. The opening science pieces were pretty bad, but far more telling was Silver's hiring of climate troll Roger Pielke Jr.
He covers that a bit, but be sure to read the bizarre original piece debunked in the "pretty bad" link (it was 538's lead science story at launch). "Gut feeling" is a key element of the model in use there, and remember, this was a Silver hire's first foray on a site that insists it's committed to "non-subjective" science journalism. Is it a deliberate troll? A tongue-in-cheek thing that fell flat? Who the hell knows. It sure doesn't read like one. Anyway, back to The Week:
Here's where we find Silver's ideological commitment, I think: contrarianism. Note how he hates opinion columnists in part because they aren't original. This quite awful piece is absolutely dripping with the stuff. (Read a thorough debunking here.)
...Sigh.
I had been under the impression that media journalist Jay Rosen had successfully pushed the pitfalls of "ideology-free" journalism into the mainstream, but apparently we've got to learn it all over again. If Nate Silver wants to patch up his stumbling enterprise, and not just make an extremely expensive Freakonomics knockoff, that's where I would start.
posted by mediareport at 8:22 AM on March 23, 2014 [6 favorites]
The Nieman Journalism lab has a good roundup of links to criticism of Silver's launch offerings, including this interesting New Republic piece, "The Emptiness of Data Journalism." Its intitial defense of punditry isn't completely convincing (I think Silver's criticisms of the idiocy of opinion columnists/haircuts tend to be very accurate; as I said above, I'm excited for Silver and want very much for his new venture to succeed), but this bit at the end rings sharp:
The intellectual predispositions that Silver ridicules as “priors” are nothing more than beliefs. What is so sinister about beliefs? He should be a little more wary of scorning them, even in degraded form...I do not expect Silver to relinquish his positivism—a prior if ever there was one—[just] because I find it tedious.
Silver proclaimed in the interview that “we’re not trying to do advocacy here. We’re trying to just do analysis. We’re not trying to sway public opinion on anything except trying to make them more numerate.” His distinction between analysis and advocacy is a little innocent. (Like the insistence of the man who went from the Times to ESPN that he is an “outsider.”) Is numeracy really what American public discourse most urgently lacks? And why would one boast of having no interest in the great disputations about injustice and inequality? Neutrality is an evasion of responsibility, unless everything is like sports. Like Ezra Klein, whom he admires, Nate Silver had made a success out of an escape into diffidence. What is it about conviction that frightens these people?
posted by mediareport at 8:51 AM on March 23, 2014 [2 favorites]
The intellectual predispositions that Silver ridicules as “priors” are nothing more than beliefs. What is so sinister about beliefs? He should be a little more wary of scorning them, even in degraded form...I do not expect Silver to relinquish his positivism—a prior if ever there was one—[just] because I find it tedious.
Silver proclaimed in the interview that “we’re not trying to do advocacy here. We’re trying to just do analysis. We’re not trying to sway public opinion on anything except trying to make them more numerate.” His distinction between analysis and advocacy is a little innocent. (Like the insistence of the man who went from the Times to ESPN that he is an “outsider.”) Is numeracy really what American public discourse most urgently lacks? And why would one boast of having no interest in the great disputations about injustice and inequality? Neutrality is an evasion of responsibility, unless everything is like sports. Like Ezra Klein, whom he admires, Nate Silver had made a success out of an escape into diffidence. What is it about conviction that frightens these people?
posted by mediareport at 8:51 AM on March 23, 2014 [2 favorites]
Nate's repsonded to criticisms of Pielke's article and said that he wants to publish a response from an expert in the subject who disagrees with it:
Tabitha Powledge has posted an article on PLoS Blog Network criticizing the new 538's science coverage. She ends with this, though:
Reception to the article ran about 80 percent negative in the comments section and on social media. A reaction like that compels us to think carefully about the piece and our editorial process.
Tabitha Powledge has posted an article on PLoS Blog Network criticizing the new 538's science coverage. She ends with this, though:
I can tell you from experience that if it looks as if the folks at 538 are making it up as they go along, that’s because they are. It’s what you do at a startup.posted by nangar at 10:32 AM on March 29, 2014 [1 favorite]
I expect they’ll get better. Maybe they’ll even figure out how to do data journalism, which would be a boon for everybody.
FiveThirtyEight’s Venezuela Problem: When it comes to Venezuela, shoddy data work and simplistic reasoning are too often embraced.
posted by tonycpsu at 6:42 PM on April 9, 2014
posted by tonycpsu at 6:42 PM on April 9, 2014
« Older Waves from the Big Bang | Dear Future Mom... Newer »
This thread has been archived and is closed to new comments
ftw.
posted by chavenet at 9:19 AM on March 17, 2014 [17 favorites]