I now put President Obama's chances at reelection at 55 percent, up 10 points from a year ago, and five points since the beginning of the summer.

There are "facts" and speculation that drive my odds-making. The facts are: 1. Mitt Romney is a weak, mistake-prone candidate, and the convention in Tampa was neither controlled nor bounce-inducing, as such events normally are. Obama has a better campaign, particularly on the data-mining and voter-contact side, so if the race is close, this operation will be decisive. 3. Romney is trying to win by maximizing white voters, which makes it hard to win some of the most important swing states and explains, in part, why his road to 270 electoral votes is much tougher than Obama's. Finally, Romney's message — Obama is a failure; I am a success — isn't working, at least not yet.

Given all of these factors, why aren't my odds of an Obama victory even higher? (Nate Silver, for example, puts them at 80 percent.) Well, here some speculation curbs my enthusiasm. I am worried that Obama has failed so far to answer this closing question: Why would he be more successful on the economy in a second term than he was in his first? I had hoped Obama would have begun his answer at his convention by framing his election as more of a referendum on specific steps he will take to boost job creation and address the debt. That didn't happen. My guess is that in order to close the sale, Obama will have to answer voter concerns not about "are you better off today than you were four years ago?" but rather, "will you be better off four years from now under my administration or his?"