: The story’s over…Rhetorica update…
It’s good to be back from a short break. I woke up early this morning feeling enthused about getting back to the ol’ blog. Hmmmmm…what does that mean?
Classes at Park University start next week. My blogging schedule will change slightly. I’ll post the new times in the left-hand sidebar later this week. If you’d like to take a look at my classes, click here.
I won’t make any decision on the return of Radio Rhetorica for a couple of weeks. Stay tuned.
While I was relaxing, Lying in Ponds posted its year-end results of partisanship among columnists.
Hunting the Muse offered an interesting challenge to my contentions in this post. I would ask this question: How is electability determined and who gets to make that determination and why? I offer my essay on the press-politics of primary campaigns as a possible answer. In any case, I must agree with the author that understanding electability is important.
I like how you use the Mayer predictive model to ask the press to focus more on policy.
I must quibble with what the Mayer model means, though. To me, it is an analysis of past performance that datamines to find correlations. The most you can really say is: “When data was in this pattern in the past, the following happened.” You can’t really say that it will therefore happen again, though. I suppose you could say that it is statistically more likely, but you can’t say that it is *as* statisically likely to happen as it has performed in the past.
This happens a lot in the stock market. People comb through the last x years of data and find something that the top performers all had in common, that the losers didn’t. Then they use this strategy to pick stocks in the future. Invariably, they find that the performance of these stocks is not as good as it was in the past when they did the analysis. They still often perform well, but the point is that you can’t really call it a reliable “predicitve” indicator.
This is because as time passes, the behavior of the raw data changes. In the stock market there are new economic realities. In the voting population, new tastes and desires emerge.
Also, we have a pretty small amount of historical presidents. It’s a small sample size.
Finally, I have an even better predictive model than Mayer’s. The nominee is always, *always* the candidate that leads in the final national poll before Iowa that has not dallied with a woman named Donna Rice.
(… actually, I don’t know who was second to Hart in that national poll. But I’m joking anyway.)
Curt…have you read Mayer’s essay from Political Science & Politics?
My presentation of his model is necessarily sketchy because I’m using it in a particular way that he did not intend. I think you should consider his presentation. You may, however, come away with the same criticisms.
No I have not, and thanks for the pointer.