Let’s look back at last cycle, focusing on Senate incumbents who were washed out with the wave . . .

NC: Elizabeth Dole’s totals in the last few polls: 46, 45, 46, 43.

Elizabeth Dole’s final result: 44.2 percent.

OR: Gordon Smith’s totals in the last few polls: 40, 42, 42, 46.

Gordon Smith’s final result: 42.5 percent.

AK: Ted Stevens’s totals in the last few polls: 46, 45, 44. (I’m not counting a Research 2000 one, for obvious reasons.)

Ted Stevens’s final result: 46.5 percent.

Now let’s go back to 2006.

OH: Mike DeWine’s totals in the last few polls: 42, 44, 44, 42.

Mike DeWine’s final result: 43.3 percent.

PA: Rick Santorum’s totals in the last few polls: 39, 40, 43, 39.

Rick Santorum’s final result: 41 percent.

VA: George Allen’s totals in the last few polls: 49, 45, 44, 44.

George Allen’s final result: 49.3.

Here and there, you could find an incumbent finishing a little better than the polls suggested; Conrad Burns finished 3 percentage points higher than the RCP average; Jim Talent finished 1.2 percent higher.

This would appear to confirm a gut-level suspicion we have, that an incumbent in the mid-40s or lower swimming against a tide tends to finish right around that final poll level; the undecideds split between the challenger and third-party options.

With that in mind:

CA: Barbara Boxer’s totals in the last few polls: 46, 43, 48, 46.

NV: Harry Reid’s totals in the last few polls: 47, 46, 46, 47.

CO: Michael Bennet’s totals in the past few polls: 46, 45, 45, 45.

WA: Patty Murray’s totals in the last few polls: 50, 49, 48, 49.

Of course, there’s still 13 days for these poll numbers to change . . .