Absurdly low value information

That is about all I can say about the decision of a major media consortium's decision to do entrance polling for the Iowa caucus. Political Wire noted this Politico story:

a consortium of the major TV networks and The Associated Press will conduct an entrance poll to measure how people say they will vote.

Those results will be broadcast long before the official vote is announced and, in some cases, before the voting is finished...

The entrance poll has a greater chance of reflecting the official results on the Republican side than on the Democratic side.

That’s because Republican caucus voting is pretty straightforward in Iowa: A voter goes into the caucus site, listens to some speeches and then writes a name on a ballot. At the end of the evening, whichever candidate gets the most votes wins...

On the Democratic side, however, entrance polling is what we columnists like to describe as “fraught with peril.”

That’s because Democratic voting in the Iowa caucus is not straightforward. There is an “alignment” where voters go to different spots in the room to indicate whom they wish to vote for, and, if a candidate does not get 15 percent of those present, there is “realignment,” with those voters going to other candidates.

If the Democratic Iowa race is as tight as it seems so that the plurality winner and the expectations winner as well as the delegate count winner(s) will be highly dependent on both second choice realigned voters, and the vagaries of organization, local speechifying, and idiosyncratic social networking effects, the headline information released only an hour or two before good information starts coming in has very little value.

I can understand wanting to entrance poll to gain information for the cross tabs, and for analytical purposes but the headline numbers as an end in and of themselves don't sound like they have a lot of value.