Absolutely correct!
The shame is that we in the New Media have not countered as good as we should have.
Read the news and the net. The election is already decided.
It’s the end for Romney (unless something drastic happens)!
Hillary is on the ropes and Obama is the certain nominee!
Fred Thompson never had a chance and this proves it!
Howard Dean is not only going to New Hampshire, he’s going to South Carolina and Oklahoma and Arizona and North Dakota and New Mexico, and he’s going to California and Texas and New York … And he’s going to South Dakota and Oregon and Washington and Michigan, and then he’s going to Washington, D.C., to take back the White House! Yeaararh!!!
Wait, what was that last one?
One of the smallest states in the nation has just held one of the most unrepresentative contests in this election cycle and the talking heads in the old media (and the self appointed experts in the blogosphere) are making definitive pronouncements about who is in and who is out.
BS.
Ron Paul is still within the uncollapsed Quantum uncertainty state of this election, for goodness sake!
(do a Google search on Schrödinger’s Cat if you don’t know what I’m talking about)
No one, I repeat, NO ONE, can tell you that anyone is a sure thing or that anyone is definitely finished. Anyone who does, even if they preface it as a “let’s assume” exercise, should be immediately shot and thrown on the trash heap of punditry as a self aggrandizing maroon (hat tip to Bugs).
Having made such a sweeping declaration I am now required to offer an alternative. Well, guess what? I’ve got one! So here it is:
I propose that all credible “pundits” be required to attach a probability scale to their dinosaur media articles or blog posts and if they don’t do this those pontifications should be dismissed from serious consideration, at least in regards to this election.
This type of scoring is a concept that is well established in the IT field (information technology, or “computers” for the great unwashed). I’ve been an IT professional, specifically PCs, for 31 years, literally from day one of factory built PCs.
The Gartner Group, one of the big IT consulting firms, has one of the best methodologies and their explanation of this is probably the easiest way to get the idea across. I don’t propose that we adopt Gartner’s methodology, but it’s a good example of the technique.
What I propose is that a two step statement be required for every article or post making predictions or pronouncements in order for it to be taken seriously. The first step is the author’s estimate of their being right. The second step is their estimate of the probability of the prediction holding up over time.
Let me provide a practical example based on my reading of the reporting on the results from the Iowa Caucus.
If you read the dinosaur media the results for the Republican presidential nomination coming out of Iowa are:
- Huckabee
- Romney
- McCain
- Paul
- Giuliani
- Thompson
- Hunter
- Tancredo
- Cheney (I threw this one in for grins… yeah, that’s it)
Under my plan they would have to list it this way:
- Huckabee: 80%, plus or minus 80%
- Romney: 30%, plus 70% or minus -40%
- McCain: 70%, plus 1% or minus 70%
- Paul: 2%, plus or minus 1% (that’s my snarky comment…. move along, nothing to see here)
- Giuliani: 2%, plus 99% or minus 1%
- Thompson: 1%, plus 99% or minus 0%
- Hunter: 0%, plus 10% or minus 0%
- Tancredo: -5%, plus 10% or minus -5%
- Cheney: -1,000%, plus 1050% minus -995%
OK, that’s enough, you either get the idea or you don’t.
If you allow for the probabilities you come out with this as the odds on result for the Republican nomination:
- Romney: likelihood of winning 100% to 70%
- Thompson: likelihood of winning 100% to 50% (my prejudiced view)
- Giuliani: likelihood of winning 100% to 50%
- Huckabee: likelihood of winning 80% to 0%
- McCain: likelihood of winning 71% to 0%
- Cheney: likelihood of winning 50% to-5%
- Hunter: likelihood of winning 10% to 0%
- Tancredo: likelihood of winning 5% to -5%
- Paul: likelihood of winning 1.5% max
From now until the election I’ll try to add a comment anywhere I find a pundits statement that boils down to “what the truth is” in this election on who is going to win. I’m not suggesting that anyone should take my “score” as the “truth.” I’m suggesting that if we can come up with this sort of short hand for scoring posts and articles on the “horse race” that is screwing up this election then readers can judge the quality of a post based on a consensus of scores, particularly on scores from those they’ve learned to trust.
I hope that others will get in the same habit of scoring posts in the MSM and on the net in some consistent fashion. I guess this is “the poor man’s DIGG.” It’s like whenever I see a thumbs up from Roger Ebert I know I will likely hate the film and vice versa.
I'm pretty sure I got my math wrong in some of the above, but the actual scores aren't the point. The Dean scream reference made more sense before Howard repeated his scream and Obama repeated almost exactly the same litany of states, but I hope you get the idea.
On a related subject, I’m working on a post about my ratings for the various talking heads and their statements. To give you an idea, I’m giving Rush Limbaugh a score of 80 (I get an 81, best of the best), Ann Coulter a score of 60 (not reduced for being wrong, but on being ”counter productive” in my opinion by providing ammunition to our enemies when she is right but not “careful” in her wording), a score of 10 to Charlie Gibson and negative several million to both Keith Olberman and Chris Matthews.
But that’s just the opinion of one thinking human being. Others may differ.