Last night (October 3rd), Mitt Romney earned one of the most convincing victories in a presidential debate I've ever seen. And, as lackluster as Barack Obama was, 'earned' is the right word. After several disastrous weeks, Romney's polished performance has put life back into his campaign. But it is likely to prove more a stemming than a turning of the tide. For now, at least.

As an observer, the fun in these debates lies both in watching the dynamic between the two candidates and then trying to parse the instapundit reactions offered afterwards by TV and online media. Reading some of the American verdicts today, you might be led to believe that Romney is President elect - the job's done; Obama has no hope. Of course, that's exactly what many were saying about his rival only the day before.

The problem here largely rests in mathematics. The 'raw' overnight poll data scores highly in Romney's favour on just about every count. So it should. But an important caveat is that these numbers tell only a small part of the story.

Most obviously, Romney's victory will get the media to take his candidacy seriously again - assuming that his gaffe handlers now have their muzzles and coshes permanently at the ready. More importantly, though, is what influence the debate will have, if any, on those who didn't actually watch it. Better newspaper and TV coverage of the candidate will help, but only to a degree.

The popular image has the entire US voting population glued to the debates, particularly the undecideds. Most are not. They work to avoid these gabfests as eagerly as we Brits dive for the kettle on hearing the words "party political broadcast".

And this, funnily enough, is why the ranks of pundits proclaiming on the night rarely include actual pollsters. They would immediately dampen the arguments, the quality TV studio confrontations, by calmly observing that what matters is how the numbers shift over the next few days, not the immediate verdict on the gladiatorial contest.

Romney and his team know that getting a two-thirds thumbs up from the live viewers will count for little if it does not now translate into a significant, extended rise in voting intentions and popularity. Obama and his will be looking at the same medium-term numbers to see if they've dodged an arguably deserved bullet.

Polls that matter are about trends, indeed so ultimately are elections. For us watchers, the first debate was therefore good stuff. It keeps the contest going - and while Romney could never have won the election at it, he could indeed have lost it as a poor performance nestled atop his earlier missteps.

So, it's still about accumulation. Now though Romney has to build on last night's surprise comeback.

Edited: 04 October 2012 at 12:10 PM by Paul Dempsey