Writings and observations

Various numbers

One of our campaign season tenets is that polls should be
addressed with caution, the results of any single poll be given only limited weight. As we’ve seen periodically this year, the best take on what a poll says (and remember, at best it’s a snapshot, not a prediction) tends to come when a group of polls are bunched together, and their results looked at en masse. Even then, an outlier sometimes hits the mark closer.

So consider Idaho’s 1st district and the two recent polls there. Last week we noted the results of a Daily Kos Research 2000 poll conducted September 17-18, showing Republican Representative Bill Sali at 46% and Democratic challenger Walt Minnick at 35%. We noted it largely because there were other readings from Idaho too, providing context.

Today the Minnick campaign released its own Harstad Strategic poll, conducted September 9-11, giving Sali 38% and Minnick 43%. The results are different enough that you have a hard time seeing how both can be on target – unless, if they are, there’s the suggestion from the timing the pollsters were in the field that Sali has gained and Minnick diminished, rapidly. Not that any reason for that comes to mind. (We have, in fact, been hearing some growing chatter – anecdotal, of course – that Minnick’s chances have seemed to be improving. It is as noted anecdotal.)

Occam’s Razor suggests this: Polls are rough shots, and best taken as raw estimates, especially one by one. Maybe if we get a slew of polls in the Idaho 1st, a sharper focus will emerge.

Share on Facebook