electoral politics gone mad

[jfb]

Slate's Election Scorecard, today:

Several of the states we've been watching closely—let's call them Tier 1--are pretty consistently going to one candidate or the other: Nevada, New Mexico, and Iowa to Bush; Pennsylvania, Michigan, New Hampshire, and now Wisconsin to Kerry. Of these, only Wisconsin and Iowa show signs of vulnerability. Florida, Ohio, and Minnesota have not settled into such a pattern. In those three states—let's call them Tier 2--polls are far from agreement. Our tentative theory is that these are the states in which poll results are most influenced by variations among likely-voter screens. To put it another way, these are the states in which turnout will most certainly decide who wins. If the Tier 1 states are distributed as current data suggest, then the election reduces to a simple equation. Whichever candidate takes two of the Tier 2 states is the next president.

Later in the day:

You can make a good case that Minnesota is falling into Tier 1 on Kerry's side. That would leave Bush with the job of winning both Florida and Ohio--or packing his bags for Crawford.

Of course, polls are always wrong, subjective analysis is wronger still, and thinking the game is over is a sure way to lose it. But for entertainment value you can't beat this kind of scope-tightening.