Log in

No account? Create an account
entries friends calendar profile Elf Sternberg's Pendorwright Projects Previous Previous Next Next
Are you kidding me, part II - Elf M. Sternberg
Are you kidding me, part II
Thomas Ricks quotes a major in the military who asks:
At what point in time should journalists, bloggers, etc ... hold those who made wildly inaccurate predications on the lifting of the ban accountable? All the retired generals and officers (LTG Mixon, Merrill A. McPeak and Col. Dave Bedey for example) who predicted that soldiers would leave the military by the thousands, or John McCain and other politicians describing how it would affect us as a fighting force? At some point I feel that the public should be reminded of their predictions so the next time they make predictions that are way off the mark, fewer people will give them credence.
Ricks and the major about to discover something very important:

Being wildly off the mark is the best way to stay employed as a pundit. As this Hamilton College study from 2007 shows, being a terrible prognisticator was actually beneficial to your career if you were a conservative. On the other hand, if you were a liberal and your predictions turned out to be true, your career nonetheless languished, your advice was routinely ignored, and the Overton Window continued to be pulled further and further toward the right.

Predictions on economic and social issues that were most often wrong came from: Cal Thomas, Senator Lindsey Graham (R-SC), Senator Carl Levin (D-MI), Senator Joe Lieberman, Sam Donaldson of ABC, and George Will.

Tell me, have any of these people lost their jobs recently?

Nobody who was wrong about the "gays in the military" issue will ever see an iota of accountability for their shrill prognostications. If anything, they'll be feted by the right for standing up for... something or other.

Current Mood: annoyed annoyed

6 comments or Leave a comment
solarbird From: solarbird Date: October 14th, 2011 06:35 pm (UTC) (Link)
Yeah. Being correct buys you nothing.

This is another thing I think that came in wholesale from fundamentalist culture. (Again, More Shit I Was Talking About 20 Years Ago.) The fundamentalists don't give a rat's ass whether anything you say is empirically true. They only care if it's Christian Worldview Correct, as they see it. As long as you keep saying that, it doesn't matter what kind of fraud or hack you are. It really seriously does not, and within the timeline of my experience, never has.
ideaphile From: ideaphile Date: October 14th, 2011 08:05 pm (UTC) (Link)

Oh, come on

This is just as true in left-wing academia (but I repeat myself) and on the Daily Kos website as among right-wing commentators.

Building and defending grossly erroneous mental models of reality is a normal human behavior.

. png
ideaphile From: ideaphile Date: October 14th, 2011 08:30 pm (UTC) (Link)

The Hamilton College study

This study does not say what you think it says.

First, it was based almost entirely on people's predictions related to the 2008 election, so three of the people on the "ugly" list made it there on the basis of their support for the eventual loser of the election. The authors also noted of those rated "good" that "the predictors who made it to this list made a number of 'easy' predictions, that were 'obvious' and did not seem to be especially good tests of predictive capacity."

Second, the study is based on eventual outcomes, not on whether the predictions were sensible given the conditions at the time they were made. For example, predicting that Hillary Clinton would win the 2008 Democratic Party nomination was not at all unreasonable through much of the primary process, but under the terms of the study, it had to be rated as a complete failure.

Third, the study says NOTHING about the effect of predictive success on a pundit's career. Nothing at all, in spite of your very specific statement here.

So all in all, you didn't do very well here even in predicting the past.

. png
elfs From: elfs Date: October 14th, 2011 08:53 pm (UTC) (Link)

Re: The Hamilton College study

Agreed, the Hamilton study does not say that careers were affected by poor predictions. It only demonstrates that the predictions made by some pundits were systematically better than those made by others, and that on the grand scheme of things, the ability to accurately predict certain outcomes corresponds to an ideological position.

Nonetheless, I stand by my accusation that conservative, Very Serious People, have not been tragically hurt by poor predictions, or even poor policy decisions, whereas liberal DFH types have been systematically dismissed and derided, and continue to be so even when their predictions turn out true.

There's a reason why the phrase "Nobody could have predicted..." has become an ironic catchphrase of the left.
ideaphile From: ideaphile Date: October 14th, 2011 11:32 pm (UTC) (Link)

Re: The Hamilton College study

The study does not demonstrate that accuracy corresponds to ideological positions for the reason I gave: it was based on a specific election cycle in which the Republican candidate lost, so Republican-biased commentators were more likely to be wrong.

There's also another reason I didn't think to mention in my previous comment, which is that the sample size is so very small that the results are not statistically significant. (And indeed you chose to highlight the results from only a small fraction of the small sample.)

Furthermore, the sampling method (popularity) was very likely to bias the sampling toward pundits who take extreme positions (the minor theme of your post, with which I agree wholeheartedly from a position of legitimate authority, having been an analyst myself for many years), so it's possible there was a much larger class of prognosticators who make milder predictions that were more likely to be correct. For example, Republican-leaning pundits who were realistic about the poor chances of Republican candidates following eight years of George W. Bush were probably less likely to be invited onto Meet the Press, because they would not be seen as an effective balance to Democratic-leaning pundits.

Now, I don't know why you would expect some random class project by four college seniors at some random tiny liberal-arts college to have any useful value in this discussion, but the facts show that it doesn't. If you want to make a claim, that's fine, but you should not pretend that your position is supported by some great authority. In particular, your referencing this as a work as a "Hamilton College study" is grossly overblown, considering that no employee of the college is even listed as an advisor.

You wouldn't have accepted this kind of nonsense from George Will; you shouldn't try to fool us with it either.

. png
gromm From: gromm Date: October 15th, 2011 04:25 am (UTC) (Link)
I don't think it matters to their constituents that what they predict is accurate, only that what they predict is what they want to hear. The only time any of them actually hears anything about said predictions is when they're being made.

Unless their opposition wants to call them on that bit of bullshit, noone is likely to notice or even care. However, since they're currently crowing about repealing DADT, there's half a chance someone might speak up about it.
6 comments or Leave a comment