That's the broader pattern in the past four elections, in fact. The vice-presidential debates -- the second of the four total in each cycle -- don't seem to have done a whole lot to affect the chances of the ticket on the whole.
In 2000, there was a spike for the Republicans after the vice-presidential debate between Dick Cheney and Joe Lieberman, but that's because the presidential debate happened a few days prior. Otherwise, the post-debate polling averages have stayed fairly flat.
Because the vice-presidential debates have been held between the presidential debates, it would be hard to separate out the effect of those debates distinctly anyway. So let's take another extreme case.
The most famous moment in debate history came in 1988, during the vice-presidential debate between Lloyd Bentsen and Dan Quayle. Bentsen's dismissal of Quayle -- a moment which will be mentioned before every vice-presidential debate until the end of the Republic* -- was probably the most effective put-down in any American presidential debate. And yet that debate, held on Oct. 5, doesn't seem to have done the Democrat at the top of the ticket, Michael Dukakis, much good.
One reason the stakes for the vice-presidential candidates is lower is that, with the exception of that Palin-Joe Biden debate, fewer people are watching. With the exception of eight years ago, the VP debate is consistently the least-watched of the cycle. People ... don't really care?
For the current VP nominees, Sen. Tim Kaine (D-Va.) and Indiana Gov. Mike Pence (R), that's good news. There's not much either can really do to tank things for the person at the top of the ticket. You can go out there, pull a Quayle and still win the election. And if, through some tragedy, you become president before the term is up, you can at least be confident in knowing that you did your best to present yourself to the American people. It's not your fault that not many of them tuned in.
* Here is the Google search interest for Lloyd Bentsen over time. The spikes? Vice-presidential debates.