Morning Joe Sort of Apologizes to Nate Silver But Not Really

Joe Scarborough penned a spectacularly snotty op-ed in which he sorta not really apologizes for mocking Nate Silver’s election forecasts.

It’s the sort of apology issued by people who can never admit that they were wrong about anything. In other words: Republicans. And of course it attempts to revise history, because that’s what Republicans do.

Here’s what Scarborough claims in his “semi-apology”:

Since the president’s reelection, liberals have been cluttering my Twitter feed with demands that I apologize to Nate for dismissing his 74.8374629% prediction in October that Barack Obama would win. I have ignored those requests because as is usually the case for ideologues on Twitter, their rage is unfocused and based in ignorance. These critics conveniently forget that I consistently predicted an Obama win throughout the year and even said on Election Day that the president’s stubborn lead in swing state polls would doom Romney.

I won’t apologize to Mr. Silver for predicting an outcome that I had also been predicting for a year. But I do need to tell Nate I’m sorry for leaning in too hard and lumping him with pollsters whose methodology is as rigorous as the Simpsons’ strip mall physician, Dr. Nick. For those sins (and a multitude of others that I’m sure I don’t even know about), I am sorry.

But on October 29, Scarborough called Nate Silver and anyone who doesn’t believe in the almighty tossup, “jokes”:

“Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning. And you talk to the Romney people, it’s the same thing,” Scarborough said. “Both sides understand that it is close, and it could go either way. And anybody that thinks that this race is anything but a tossup right now is such an ideologue, they should be kept away from typewriters, computers, laptops and microphones for the next 10 days, because they’re jokes.”

And here’s what he said three days prior to that:

If the President is sitting at 47 percent in the polls, boy, I’d be shocked if he won. I’ve seen he’s in 48 in some of these polls. So that’s my gut. That said, he’s got an incredible ground game and a lot of early voters are coming out and those early voters that are coming out are voting for Barack Obama. So you have… if you’re a behavioralist — and I am too when it comes to these things — if you’re undecided about Barack Obama and you’ve known him for give years and followed him for five years and he’s been on the front page of your newspaper for five years and been on your TV set every night for five years and everybody’s been talking about him for five years and you still don’t know if you’re going to vote for Barack Obama, history would suggest you’re not going to swing in the last ten days.

Scarborough dismissed math in favor of his gut and got his ass handed to them by Nate Silver. And instead of outright apologizing, and figuring out a way to turn the egg on his face into a delicious omelette, he demonstrates that he still doesn’t get it.

Here’s the thing, Joe — Nate Silver was not “predicting an outcome.” He was calculating the probability that one outcome would come to pass as opposed to another.

It’s not magic — it’s math.

[cross-posted at ABLC]

Share On Facebook
Share On Twitter
Share On Google Plus
Share On Pinterest
Share On Reddit

63 replies
  1. 1
    Bulworth says:

    Latest Gallup has Mitt up by 3!

  2. 2
    jl says:

    Mitt was spotted pumping gas the other day and it made the news. Morning Joe pumps gas and it is ‘the news’. We are doomed.

  3. 3
    Gladys Cravitz says:

    wonder if he had a “gut” feeling that he would find a dead intern in his office one Monday morning back in the day?

  4. 4
    Violet says:

    @Bulworth: UNLIMITED CORPORATE CASH!

  5. 5
    rlrr says:

    @Bulworth:

    The Romentun is unstoppable! Obama is toast.

  6. 6
    Schlemizel says:

    probably can copy and paste from here

    http://www.balloon-juice.com/2.....poltroons/

    But Joke Scar bashing never gets old so lets have a go at it

  7. 7
    Richard says:

    One wonders if there will now be rush by Republicans to get math removed from school curriculums. After all, it, like evolution, offends their faith based sensibilities, doesn’t it?

  8. 8
    quannlace says:

    From Joe, to Peggy Noonan, to god-knows how many ‘pundits’ on Fox news, what was with all that ‘I have a feeling…” ? I guess that’s what you have to fall back on when you’re trying to avoid reality.

  9. 9
    Rafer Janders says:

    “Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning. And you talk to the Romney people, it’s the same thing,”

    Statistical probability, how the fuck does it work?

  10. 10
    ThatLeftTurnInABQ says:

    such an ideologue

    If Scar had any more projection you could mount theater seats with cupholders on his sorry ass and sell popcorn in the lobby.

  11. 11
    daverave says:

    This can only be good news for John McCain’s election chances, also, too.

  12. 12
    Napoleon says:

    Speaking of Silver over at TPM they have a clip of BHO pardening the turkeys during which he cracks a Nate Silver joke.

  13. 13
    Rafer Janders says:

    @quannlace:

    From Joe, to Peggy Noonan, to god-knows how many ‘pundits’ on Fox news, what was with all that ‘I have a feeling…” ?

    Having a feeling that tonight’s gonna be a good good night a feeling woohoo is perfectly valid if you’re one of the Black-Eyed Peas. But if you’re a paid (supposedly) professional pundit, not so much.

  14. 14
    ThatLeftTurnInABQ says:

    @quannlace:

    what was with all that “I have a feeling…”?

    “I have a feeling” = Don’t look now, but my hands are in my pants.

  15. 15
    Ruckus says:

    “Maths are hard”

  16. 16
    Chyron HR says:

    @Bulworth:

    I wish Dipstick would come back for a day. I think we all owe him our thanks–I mean, what if the GOP had supported someone who actually tried campaigning in swing states instead of just making up magic polls that showed him winning Oregon?

  17. 17
    Rafer Janders says:

    Here’s the thing, Joe—Nate Silver was not “predicting an outcome.” He was calculating the probability that one outcome would come to pass as opposed to another.

    Why, why, why do people not understand the difference between these two things?

    If I have a coin, and I note that it has a 50% chance of coming up heads if flipped, I have not predicted it wrong if it comes up tails instead. I’ve simply noted the relative probability of each of two possible events coming to pass.

  18. 18
    ding dong says:

    Joe dead intern is a dickk wadd.

  19. 19
    Ted & Hellen says:

    Why do you care enough what Joe Scarborough said about anything to write a post about it?

  20. 20
    jackmac says:

    And I’m 100 percent sure Morning Joe is an snotty a**hole.

  21. 21
    Liberty60 says:

    math, magic, its all the same to Republicans.

  22. 22
    NorthLeft12 says:

    The worst best part is that they still don’t get it.

    I would not be surprised if Joe/Peggy/George/etc. are trying to figure out who fed this information to Silver. That is exactly how they work, and they can’t imagine any other way. I would love to see these pundits actually get a real job and have to function in the real world. I think it would be entertaining to actually work with them. Until I had to pick up all their slack.

  23. 23
    trollhattan says:

    These guys (Scar being one of hundreds) continue to make Colbert seem like the smartest, most relevant guy in the room. His very first show introduced truthiness and the importance of speaking from the gut. Seven years later they’re still at it, and not being the least ironic in the process. Hannity lectures folks on food stamps to consider rice and beans.

    Sheesh.

  24. 24
    Studly Pantload, the emotionally unavailable unicorn says:

    @Ted & Hellen: Y’all’s new here aintcha? If it can be mocked, that’s what we’re here for. Especially if the mockee makes his living by being a clueless-but-smug prick on the teevee three hours everday.

  25. 25
    jl says:

    @Rafer Janders: You’re a party pooper.

    But also too, Silver predicted expected electoral votes (313), which implies a win, and expected popular vote share (50.8). But you are right, Silver never made an official prediction that Obama would be the winner.

  26. 26
    Svensker says:

    @ThatLeftTurnInABQ:

    That made my LOL.

  27. 27
    👽 Martin says:

    Aside from the fact that Joe is a dick and wrote a dickish article, I’m not going to pile onto the GOP here over the polling. Nate Silver rose out of the very same polling issues that the Dems suffered from in 2000 and 2004. He hopped up and gave everyone a credible analysis that cut through the known biases in the polls the public was getting. The GOP will take Nate and other aggregators more seriously in the future, and they’ll likely offer up their own as well. And the aggregators will keep the polling outfits a bit more honest by making it harder for them to grift their clients by providing results that they want to see.

    We’ve seen this all before on the left, and not that long ago. We should worry more about the day that the GOP stops making assumptions about the electorate in the polls and instead accepts the polling and then works to shape the electorate in their favor. That’s what 2008 brought for the Dems and moreso this year. Right now the GOP isn’t even playing that game. This will get harder once they show up.

  28. 28
    gVOR08 says:

    Scarborough made it obvious that he’s too innumerate to have any understanding of what Silver does. Joe thinks Silver does the same thing he does, then dresses it up in some math mumbo jumbo. He doesn’t understand that Silver does numbers first, then conclusion. Joe’s a conservative. He believes what he believes because he believes it. One of the things he believes is that other people all operate the same way.

  29. 29
    Anoniminous says:

    Innumeracy. It’s what for breakfast.

  30. 30
    Uncle Cosmo says:

    @Violet: UNLIBERATED COPROLITE CACHE, more like.

  31. 31
    Schlemizel says:

    @👽 Martin:

    You say that but can you point to some evidence? Not only was 2000 close Gore actually won so any Dem pointing to polls showing a Gore win would be hard to be labeled as wishful thinking. 2004 was close but the folks I was reading thought Boy Blunder was in the stronger position and our only hope was GOTV.

    I did not see anyone in the campaign that was stunned. Certainly John did not have $25,000 in fireworks ready to blow off Election night.

    There is a huge difference between hoping we can pull out a close one & the huge victory predictions for Willard teamed up with his total lack of preparation for the defeat

  32. 32
    Jay C says:

    It’s not magic—it’s math.

    But given the sorry-ass “predictions” and “unskewing” coming from starboard before the election (and the wailing and whining afterwards) – it’s hard to discern that they actually know the difference. Or care, for that matter….

  33. 33
    Jay in Oregon says:

    @quannlace:

    From Joe, to Peggy Noonan, to god-knows how many ‘pundits’ on Fox news, what was with all that ‘I have a feeling…” ? I guess that’s what you have to fall back on when you’re trying to avoid reality.

    That’s pretty much it. If you do things like refer to facts or evidence, those can be used to conclude that you are full of shit.

    They rely on their “gut”, which is just a hacktacular form of appealing to authority. “I’m an expert in these things, and so I can rely on my finely-tuned mind to discern the truth of things.”

    It’s been said several times since Nov. 6; if Nate Silver or Sam Wang were horribly wrong in their predictions, they’d be looking for work. But the Noonans and Scarboroughs and Bobos of the world can fuck up royally, day in and day out, and never pay the price.

  34. 34
    Nick says:

    I really think the genuine confusion – which is so profoundly stupid that I don’t think it even occurred to us liberals – is that pundits think “75-25 odds” is the same thing as “75% – 25% in the polls.” Or on the flip side, if Obama led 51-49 in every poll in every swing state for two months leading to the election, the Serious pundits would say he had a 51% chance of winning, rather than Silver’s “assuming the polls are independently and identically distributed-ish, Obama has a 99.995% chance of winning.”

    The other big misconception has been diagnosed multitudinously – the basic inability to understand that probability means you will probably get an outlier from the mean – but I really think the stupid might be more basic than we were thinking.

  35. 35
    Rafer Janders says:

    @Nick:

    I really think the genuine confusion – which is so profoundly stupid that I don’t think it even occurred to us liberals – is that pundits think “75-25 odds” is the same thing as “75% – 25% in the polls.” Or on the flip side, if Obama led 51-49 in every poll in every swing state for two months leading to the election, the Serious pundits would say he had a 51% chance of winning, rather than Silver’s “assuming the polls are independently and identically distributed-ish, Obama has a 99.995% chance of winning.”

    Well, yeah. That’s exactly what Scarborough is doing here when he writes that: “Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning.” He’s confusing two entirely different mathematical ideas.

  36. 36
    FlipYrWhig says:

    @Nick: I think you’re exactly right. They don’t get that if 55% of the public is for Doe and 45% for Schmoe, Doe would have an overwhelming chance at winning, not a 55% chance of winning.

  37. 37
    FlipYrWhig says:

    @Rafer Janders: it might help them if they thought of it as a 75% chance of getting 50% plus one of the vote. But even that many numbers frightens and confuses them.

  38. 38
    rob! says:

    Sully rips Joe Scar a well-deserved new one here: http://andrewsullivan.thedaily.....-joke.html

  39. 39
    ThatLeftTurnInABQ says:

    @Nick:

    I really think the genuine confusion – which is so profoundly stupid that I don’t think it even occurred to us liberals – is that pundits think “75-25 odds” is the same thing as “75% – 25% in the polls.”

    A good point.

    A possible contributing factor to this type of thinking may be that for most of the 20th Century the Dems and GOP were in the process of swapping their regional and cultural identities, and during that period huge landslides were possible: 1932, 1936, 1964, 1972, and blowouts just a little short of that mark were even more common.

    My guess is this happened because partisan identity back then was not so tightly nailed down (as it is today) either ideologically or regionally/culturally, so there were plenty of swing voters who truly could go one way or the other in massive numbers, depending on the details of each election. Which meant that back then a presidential candidate who was a dead-certain lock to win could expect to win by 15+ points and would be polling accordingly prior to the election.

    But we don’t live in that era any more. As a result of the GOP’s Southern Strategy running to its logical endpoint, our political environment is more static than in the mid-20th Cen. The last 4 Presidential elections have all featured similar looking maps, albeit growing a bit bluer over time. And that means that today a candidate like Obama in 2012 who was very, very likely to win does not enjoy the massive margin of victory that FDR, LBJ, or Nixon enjoyed, and polls accordingly.

    A lot of people haven’t figured this out yet, that in our era a 7% win is a landslide and a 4% win is a near-landslide, because the electorate is pretty well locked-in and doesn’t have that much give to it. And of course our pundits being what they are, they will be the last ones to understand what is going on.

  40. 40
    jayackroyd says:

    Sam Wang has commented on that “prediction” business. He’s said that you’ll inevitably be viewed as making a prediction of the result when you report very high odds is favor of an event occurring. Now, HE was quoting really high odds–in the 90s in the last weeks of the election, 100! the week of, using his preferred model — but I think it’s a little disingenuous to deny that a betting line is a kind of prediction.

    The real point is that the Church of the Savvy–inhabited by the rolodex wielding insidery media professionals–turns out to be the venue of a false god. There’s no reason to listen to Chuck Todd ever again. Sam published his source. He used a publicly available XML feed. Joe Scar can just go suck dead donkey dicks–that’ll work just as well as calling up his in the know pros. And THAT’s the most ridiculous thing that happened this Sunday, or whenever.

  41. 41
    👽 Martin says:

    @Schlemizel:

    You say that but can you point to some evidence?

    Yeah, all of Daily Kos. Dems of all stripes were disbelieving the polls – “Oh, they didn’t count cellphones”, “Oh, their internal polls say something different” and so on.

    The polls in aggregate weren’t wrong, but nobody could see them in aggregate. People picked the one that told them what they wanted to hear (with flamewars over which polling outfit was best and sample size and so on) and people made all manner of excuses over whether the polls were right or not. That’s entirely what the GOP did this cycle. Nate included all of the GOP polls in his analysis and it came out with what it came out with – so it’s not like the GOP polls were so far wrong that it threw off his model.

    But the left doesn’t argue over these things any more. We go to Nate or Sam Wang or whoever and get the aggregate and believe the aggregate and then question why polls aren’t aligning with the aggregate.

    Yeah, 2000 and 2004 were closer elections, but everyone still had their own polling biases. There was no credible aggregate to suggest that the ones showing Kerry leading or trailing were likely right or wrong. By comparison, Nate got 2008 dead right other than Indiana. He got everything in 2012 right other than one senate race. Both Nate and Sam had the most likely outcomes as Obama winning 332 or Obama winning 303 (losing Florida). The issue of polling reliability is effectively eliminated on the left since 2008. But the right hasn’t bought into it yet. But they will.

  42. 42
    gelfling545 says:

    @Ruckus: The 14 year old granddaughter is struggling with algebra & whining (a lot) about how hard it is. Finally her mother snapped at her “Well you had better learn it and learn it well or you might grow up to be a republican.”

  43. 43
    jayackroyd says:

    @ThatLeftTurnInABQ:

    This isn’t quite right, IMO. It’s not that they think 75% is a popular vote prediction. What they think is that a low single digit lead (51-49) is a probability assessment–that a tight race is coin toss.

    Their innumeracy lay in their not recognizing that the EV lead never changed hands based on state polling–that it would be close state by battleground state, but the outcome was not in doubt.

    But, really, of course what Joe is doing is trying frantically to validate the Cult of the Savvy–that he knows more than the polls do. Which, is, of course, both teh crazee, and what Fluffy puts into his bran flakes every morning.

  44. 44
    👽 Martin says:

    @Rafer Janders:

    Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning.” He’s confusing two entirely different mathematical ideas.

    He’s also doing something a bit different. The campaign was acting as though they had a 50.1% chance of winning. They may not have believed that, but everyone on both sides is equally terrified of getting complacent from overconfidence and discouraging turnout from an effort that seems hopeless. Regardless of what the campaign believes, they’re going to broadcast a 50.1% likelihood of winning message.

    Joe can’t discern the two because he’s inherently a politician and needs to believe propaganda as much as everyone else.

  45. 45
    jayackroyd says:

    @👽 Martin: Nicely said!

  46. 46
    Professor says:

    @jayackroyd: I would have preferred if Joe Scar sucked on ten salted dicks!

  47. 47
    Turgidson says:

    @ThatLeftTurnInABQ: And of course our pundits being what they are, they will be the last ones to understand what is going on.

    Of course. And they’ll make smug, self-satisfied and certain predictions about things they do not, and don’t want to, understand every step of the way. And they’ll be wrong the overwhelming majority of the time. And they’ll remain gainfully employed.

    I mean, jesus. Has Bobo been right about fucking anything he’s ventured an opinion about in the last decade? And he’s the thoughtful, reasonable, humble Burkean, modest above-the-partisanship guy.

    It’s just about meteor time.

  48. 48
    ThatLeftTurnInABQ says:

    @jayackroyd:

    Yes, I understand that the innumerate pundits can’t translate between polling levels and probable outcomes, but what I’m pointing out is that in addition to the former problem, up until pretty recently we didn’t have many national elections which were both close and static. That is something new, because today’s electorate is less up for grabs than was the case thru most of the previous century, and a lot of the folk wisdom re: how politics works is stale. The pundits are recycling folk tales about how the electorate behaves that made more sense back in the 1970s than they do now.

  49. 49
    J.D. Rhoades says:

    I have ignored those requests because as is usually the case for ideologues on Twitter, their rage is unfocused and based in ignorance.

    Whenever they can’t respond to what people are saying, they get all pissy about the way it’s said.

  50. 50
    Calouste says:

    @gVOR08:

    Scarborough made it obvious that he’s too innumerate

    He’s a journalist, so basically someone who needs a calculator every morning to work out how many shoes to put on.

  51. 51
    bemused says:

    Joe’s ego writes checks his body can’t cash. They bounce.

  52. 52
    jayackroyd says:

    @ThatLeftTurnInABQ: Thanks. That is a really good point. It’s the theme of the Southern Strategy–the populist racists opted for the racism over the populism, which has ossified the electorate. There are no swing voters. Close and static is the immediate future. What the republican poobahs are realizing in this post mortem, is that it won’t stay close. Static, yes. So Rubio!

  53. 53
    Citizen_X says:

    74.8374629%

    A clear symptom of innumeracy: utter cluelessness about significant figures.

    their rage is unfocused and based in ignorance

    We call that P-R-O-J-E-C-T-I-O-N.

  54. 54
    rikryah says:

    he’s an assclown.

    period

  55. 55
    Triassic Sands says:

    Being wrong is as Republican as bigotry. They are inseparable.

    It looks like Joe got both sides of the equation wrong — since his claim that both sides thought they had a 50.1% chance of winning is wrong in both cases. From all post-election reports, it appears that Romney thought he had a 100% chance of winning (thus the extreme shock at losing). Obama on the other hand always had to calculate the chances of a successful theft of the election by Rove and Co., so his confidence, while fairly high, probably couldn’t have been 100%.

  56. 56
    Patricia Kayden says:

    @Bulworth: I won’t even look at Gallup in 2016. They were nothing but wrong this time around.

  57. 57
    Rathskeller says:

    I would add a third factor to the bias and the innumeracy: he’s straight-up lying to himself now about what he really believed. Right here, in print, he writes about how he always thought Obama was going to win. We’re watching the narrative harden in the head of a MSM member. Everyone knew. There was no doubt. I never called Nate Silver a joke who should be kept away from computers and typewriters.

    This is how they live with themselves. They’re always right in their memory of it.

    Typewriters?! Right now, I couldn’t tell you where a typewriter physically was to save my life.

  58. 58
    ThresherK says:

    @jl: I can hear the reporter’s exchange with Mitt at the filling station now:

    “Governor, what are you doing with that nozzle?”

    “Gasoline stuff!”

  59. 59
    AA+ Bonds says:

    He’s one of plenty of journalists, Democratic and Republican, who in a just world would be jailed for war crimes because of their willing complicity in the invasion of Iraq.

    But what are you gonna do? Hanging is too good for those people.

  60. 60
    Ruckus says:

    @gelfling545:
    I tutored statistics in college and found the biggest thing people have a problem with is how to look at math or statistics. Most people, if they have a problem, is that they are trying to solve the problems(nuts and bolts) but have no understanding of the logic behind the problem. Most schools teach rote not theory and I think it leaves a lot of people scratching their heads trying to figure out why something is, when what they are tested about is rote memory. For example, in teaching machine programing I found out I had to teach what we were trying to accomplish before I could teach how to accomplish it. Some people need to understand the logic to apply it.

  61. 61
    RSA says:

    @Nick:

    I really think the genuine confusion – which is so profoundly stupid that I don’t think it even occurred to us liberals – is that pundits think “75-25 odds” is the same thing as “75% – 25% in the polls.”

    That’s a good thought, and kinda scary. I suspect it’s something simpler, though, that the average pundit (and for all I know the average person) thinks things are basically either certain or not, and any uncertain binary event has a probability of 0.5. Then they’ll add a fudge factor to match their intuition. I mean, “50.1%”? Why not 51% or 50.000001%?

    I know that people play cards and bet on sports games and so forth, but I guess a lot of them are just not applying what they know to elections.

  62. 62
    jayackroyd says:

    @Ruckus: The whole point of probability and statistics as a discipline is we suck at assessing data evidence.

  63. 63
    different-church-lady says:

    Joe, it was never Nate’s fault that you didn’t understand the difference between a probability that one candidate would win and that’s same candidate’s probable percentage of the vote. Sure you said you thought Obama would win, but that’s not the insult you hurled, so why you’re trying to use it to get off the hook I don’t know.

Comments are closed.