Here's their chart:
The Gallup numbers suggest that only twice in the past 52 years -- 1960 and 2000 -- did the candidate who went into the first debate behind wind up ahead after the last debate.
The 1960 debates -- the first ever to be televised -- are widely credited with turning the race to John Kennedy thanks to the handsome and healthy image he cut as compared to the jowly (and sweaty) Richard Nixon.
In 2000, George W. Bush's solid performances -- not to mention Al Gore's sighing -- seemed to make it acceptable for people still undecided about the then Texas governor to cast a vote for him.
In 2008, which the Gallup data doesn't cover, Obama's steady performance likely firmed up the growing sentiment that he was up to the job of being president. But, the debates had far less to do with the ultimate outcome of the race than the near-collapse of the financial industry in mid-September followed by John McCain's insistence that the "fundamentals of the economy are strong".
(As for the 1980 and 1992 elections, the good people at Gallup note: "The 1980 and 1992 debates may have influenced voter support for the third-party candidates running in those elections; however, they do not appear to have altered the structure of the races for the two major-party candidates.")
With those two notable exceptions, the candidate who entered the debates ahead in Gallup tracking left the debates ahead. Why?
Because, typically, the parameters of the race are very well established before the debates begin. It's hard for either candidate to do anything surprising enough to move the needle in a major way in the space of a 90-minute or two hour debate.
And, the candidates put so much time and practice into these debates that the chances of a big slip-up are significantly reduced. Both Obama and Romney will be rehearsed to within an inch of their lives when the lights go on next Wednesday night and, given that, it's hard to imagine either man scoring a decisive blow -- or blows -- in order to be declared the clear winner.
All that said, we still think that next Wednesday's debate does matter -- particularly for Romney. What polling struggles to track is how the debates can shift or affirm the media narrative of the race -- particularly because there are so many people watching.
Romney is almost certain to enter next week's set-to as the underdog, and his best -- and maybe only -- way to change that storyline is to show strong in the debate with Obama. A mediocre or poor showing will reinforce the idea that Romney is a struggling candidate, a judgement from which he may not be able to recover.