The Gansler-Ivey poll results are catnip for people like me who follow campaigns but also a good example of why outlets that try to estimate the current shape of election campaigns (e.g. 538, pollster.com) do not include them in their analyses.

The press release includes some interesting numbers. I was less interested in the top lines than in the report of Doug Gansler’s favorability ratings. If opinions of the AG have indeed improved since the spate of very bad press earlier this year, that would certainly be good news for the Gansler-Ivey campaign.

However, the press release was more telling for what it did not include than what it did. There is no information about the questions that were asked. One poster on Seventh State’s Facebook page claims that the questions were primed to elicit negative responses about Lt. Gov. Anthony Brown.

(Update: The Gansler campaign informs me this was not the case and that the questions about candidate ratings and horse race numbers were asked prior to the questions regarding the exchanges in any case.)

While the poll reported Gansler’s favorability ratings, it did not do the same for either Brown or Mizeur. Additionally, there are no demographic breakdowns. I’d be especially interested to know the gender, racial, and religious composition of the survey, as well as the results for these demographic groups.

This information would make it possible to answer several questions. For example, does the share of women estimated in the electorate correspond to past gubernatorial elections? Women routinely makeup a disproportionate share of Democratic primary voters in Maryland but do they in the polling sample? How strong does the poll state support is for candidates among groups whose support they might hope to consolidate?

So, while fun to read, I’ll be looking forward to the next poll reported by an outlet not associated with one of the campaigns.

Note: I’m supporting Gansler but I try to call it like I see it as is evident here.