In case you've been exposed today to all the "news" about a tightening presidential contest (sometimes referred to by conservative gabbers as a "dramatically tightening" contest), you need to spend some time at FiveThirtyEight.com. As Nate Silver has carefully explained, the poll most often cited for this proposition is a Gallup Tracking poll that utilizes a "historical" turnout model--i.e., one that is heavily based on 2004. This poll shows McCain within two points of Obama, at 49%-47%, which might be alarming to Democrats if somehow the turnout patterns happened to closely resemble those of four years ago, which hardly anyone credible expects.

But a second Gallup tracking poll, based on the "current intentions" of voters, shows Obama up by six percent. Gallup started releasing this second tracking poll recently for the precise reason that experts were criticizing its "traditional" model as potentially misleading.

If the "traditional" Gallup model is indeed skewed significantly towards a "redder" electorate than is real, then it wouldn't be that surprising that McCain's base-pleasing debate performance on Wednesday night might bump his numbers there a bit.

In any event, we'll need better evidence than a half-self-repudiated Gallup Tracking poll to conclude that the race is in fact "tightening" to any significant extent.