• Ohio (20), having moved out of Kerry's column a week ago, now slips farther away from him -- it's now assigned to Bush. This is a big blow. Ohio's been pretty volatile, moving from DEM to "in play" and back again, but when both the Mason-Dixon Plain Dealer poll (5/25 K41-B47-N3 moe2.6) and the 30-day Rasmussen survey (5/31 K44-B46 moe4) showed Bush ahead, I moved it back into "in play". Zogby's figures (K46.3-B49.1-N1.2 moe2.1 -- B+2.8 inside the moe), pushed it to Bush. It will now take several consecutive Kerry leads to move it back into "in play".



• There is a counterweight to Kerry's loss of Ohio: Florida (27), which has been in my "in play" column for quite a while, moved into Kerry's category. While not quite as active as Ohio, Florida's been fairly volatile as well, but it never totally slipped away from the Democrats. With no intervening polls in the state that I know of, the current Zogby results (K49.5-B47.9-N0.6 moe2.9 -- K+1.6 inside the margin) simply confirm the previous Zogby numbers, to which they are very similar (K49-B47.6-N1 moe3.4 -- K+1.4 inside)



• Iowa (7) also moved from in play to Kerry. The previous Zogby poll had Bush ahead (K44.9-B50.1-N0.8 moe4 -- B+5.2 outside the margin of error), which caused me to move the state into "in play" for the first time, but since then two other polls showed Kerry with a small advantage: Research 2000 (K48-B43 or K46-B42-N3 moe4) and Survey USA (5/26 K48-B45 moe3.6). Today's Zogby numbers (K48.9-B47.6-N1.2 moe3.7 -- K+1.3 inside) confirm that, and put the state into Kerry's list, if something less than firmly so.



• Nevada (5), on the other hand, moved from GOP to in play. The previous Zogby interactive numbers (K47.3-B43.5-N2.8 moe4.3 -- K+3.8 inside) and Chris Bowers' estimate of a Kerry +2.2 advantage in the state were enough to convince me that Nevada was not necessarily Bush's any longer, and the current Zogby figures (K47.3-B43.5-N2.8 moe4.3 -- K+3.8 inside), very similar to the old ones, if slightly under, shifted it out into "in play".



• West Virginia (5), which had been assumed to be strong Bush territory, got put on my suspect list with the results of a (possibly leaked, possibly private) Mason-Dixon poll (K47-B41 moe4), and the current Zogby numbers (K46.6-B43.5-N2.2 moe4.4 -- K+3.1 inside the margin of error) were sufficient to knock it out of Bush's box for the moment.
  



• Also as a result of the Zogby numbers, I've put Missouri (11) on my watch list - one more Bush lead and I'll be moving it back into the GOP column.



• On the other hand, I'll be watching for confirmation of a Kerry trend in West Virginia, and will probably move Nevada to Kerry with another lead there. Other than that, all my suspect/watch states were settled.

These changes put my current prediction at:

DEM: FL-27, IA-7, ME-4, MI-17, MN-10, NH-4, OR-7, PA-21, WA-11, WA-10
(118+168=286)

GOP: AZ-10, AR-6, CO-9, LA-9, OH-20, TN-11, VA-13
(78+148=226)

IN PLAY: MO-11, NM-5 , NV-5, WV-5 (26)

DEM 286 (270 needed to win) / GOP 226 / PLAY 26

Obviously, it's a bit of a paradox that Kerry's slippage in the Zogby results should end up with my increasing my prediction for his electoral vote total, but that's simply due to my not transferring states whenever there's a contrary poll result, instead waiting for some confirmation before making a change. These numbers were sufficient to confirm some moves that had been pending, and most of them (with the exception of Ohio) went in Kerry's favor. If Kerry's Zogby slippage is reflected in other poll results in the coming weeks, then some of those states will also slip away from him, as Ohio has.

But, in the end, it doesn't much matter how much Kerry wins a state by, as long as he wins the state.

Update: I expanded the list of changes in order to better explain why I made the moves I did.

Update: The "current prediction" link should be fixed.