Texas Governor Rick Perry is finally admitting that he may wind up in the presidential race after all. And as silly as it seems, Rudy Giuliani getting another round of presidential buzz. So perhaps it’s time to think about the shape of the fall campaign.

Let’s start by running through all the things that don’t matter — even though they are talked about endlessly by political reporters. First: Don’t bother looking at head-to-head matchups between Barack Obama and various Republican contenders. That’s the message from an excellent graph from political scientist Chris Wlezien, posted over at the Monkey Cage. It turns out that such polls this early in the process have basically no predictive value at all. As time goes on, that changes gradually. But this far out, nothing.

Second, ignore the endless speculation about who will run for the Republican nomination. I wouldn’t pay a lot of attention to the out-party nomination process. I wouldn’t go so far as to say it makes no difference at all who challenges the president, but the effects are probably very small. Perceived ideological extremism in a candidate might cost a few points, and other effects are certainly possible, but mostly what drives presidential election voting is a combination of partisanship and assessments of how the president is doing — assessments that are most heavily influenced by economic factors, although other things can come into play.

Meanwhile, state-level head-to-head polling is going to be especially meaningless. I recommend totally ignoring those until after the party conventions next year. No matter what state level polls say now, what happens in the states really just generally reflect national events and trends. It’s only in the last few weeks that state polls really start to matter — and that’s only if the race is so close that a slight shift in a state’s electorate, and in the electoral college overall, can make the difference.

So: ignore all those things. The only two things to watch, this far out, are: Indicators about the state of the economy, especially how it will likely look next year; and Obama’s approval ratings, which incorporate both economic factors and anything else that matters to people. When it comes to predicting who’s going to win in the end, nothing else really counts.