Interesting note in the google doc where this post gets its ballot numbers from: For these reasons, I do not think the absentee ballot stats are particularly informative at this early stage as to what will happen in Ohio, much like I do not think the wildly pro-Democratic Iowa statistics are informative as to what will happen in that state. I suppose if you want to do a better apples to apples comparison for Ohio, you might run the 2008 numbers for the same point in time prior to the 2008 election. That might give us some sense of the overall level of early voting to be expected in the state. But I would again caution about drawing inferences about "party" as I have seen some do with your numbers. I do not think these numbers tell us the Ohio polling is incorrect. We need more information before we can draw that conclusion. Also I'm kinda confused about the whole over-sampling of dems kerfuffle. From the looks of it, the complaint is that when polling a random sample of likely voters 48% said they were D and 42% said they were R?If you are not aware, Ohio does not truly have party registration, it is simply a record of the last party primary a voter participated in. For a number of reasons then, a comparison of 2008 and 2012 "party" is not particularly informative. Another important change is that all registered voters have been mailed an absentee ballot request form - something I applaud SoS Husted for doing since the larger Democratic counties were the ones that tended to do this in 2010. As a result of that change from 2008, I expect mail ballot requests to increase. Finally, Ohio considered in-person early voting to be a form of mail balloting in their reported statistics.The numbers are going to change considerably once in-person early voting starts, particularly closer to Election Day when more people (particularly Democrats) tend to vote in-person early.