How often have you heard this year’s presidential debates described as “critical,” “all-important” or even “decisive”? About as frequently as you have in every election year, I’d bet.

We like believing debates are central determinants of electoral outcomes. They give campaigns a focus, journalists a hook around which to wrap their coverage and citizens the illusion that we render verdicts based on intellectual trial by combat. Peering into a candidate’s character, as revealed in such trials, seems wiser than deciding based on 30-second spots and scripted speeches.

Vivid memories of debate lines drilled into our consciousness by endless repetition only heighten our level of attention. Who can forget Ronald Reagan’s “are you better off” question, or Mike Dukakis’s too-calm response to a question about his wife being raped? We retell these tales so often it becomes hard to imagine they weren’t dispositive.

It is a convenient mythology for all concerned, but the myth bears little relationship to the reality.

Presumably, a candidate who drooled into the camera for 20 minutes would find himself disqualified, but that hasn’t happened, and it’s not likely to. If debates were decisive, though, John Kerry would be president.

In fact, debates have probably never decided the outcome of a presidential election, and even when they have been important, it is because of some post-facto interpretation injected into the public psyche, not the result of what debate watchers thought they saw or heard in real time.

One study estimated the average change in candidate support caused by the 13 debates between 1988 and 2004 was just 1 percentage point.

Gallup identifies 1960 and 2000 as the only two instances in which debates “may have had a substantive impact on election outcomes,” and even those are, as we shall see, debatable.

Gallup’s brief list leaves out some memorable moments.

Before Dukakis’s alleged error in responding to the question about his wife, George H.W. Bush was averaging 54 percent of the vote. He came out of the exercise averaging 55 percent, before garnering 53 percent on Election Day.

The 1980 debate featured Reagan’s famous put-down of Carter and the “are you better off” closing question.

The impact? Uncertain. Reagan’s tracking showed him maintaining a five-point margin before and after the debate, which widened as the focus shifted back to American hostages in Iran.

Carter’s poll indicated the race was even before the debate, and tied again within four days afterward. Neither pollster’s data suggest the debate determined the outcome.

While 1960 is the iconic example of a debate that changed history, the real story is quite murky. Kennedy had been gaining headed into the first debate, and was behind Nixon by just one point. After all the debates, Kennedy led by four points in a race he eventually won by less than one — well within the margin of error of the pre-debate poll and certainly closer to that result than to Gallup’s post-debate reading.

Even the clearest case of impact is not that clear. In 2000, Al Gore led George W. Bush before the first debate, which a narrow plurality scored as a Gore win. Polls immediately after showed Gore expanding his margin. However, within a few days Gore began falling as Republicans branded him a “serial exaggerator,” using his debate statements, among others, as fodder for the attack. Gore continued to slip through the third debate, from which the vice president emerged four points behind Bush. Nevertheless, Gore won the popular vote by about one point. Determining the specific impact of the debates on that race is no mean feat.

So enjoy the debates, learn from them, but don’t expect them to change the outcome.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the majority leaders of both the House and Senate.