“Wave” has been by far the most overused and abused term in this midterm cycle. Some analysts just use it to mean any strong result, usually benefiting Republicans. Some have benchmarks in mind, usually double-digit GOP gains in the House and GOP control of the Senate. But in terms of the latter, Charlie Cook has the most sensible way of looking at it:
If there is a wave, don’t look at either the Romney +14 states, or the three Republican-held seats for evidence. It is the open seat in Iowa, the race in Colorado, and the Hagan and Shaheen seats that will sound the alarm. If Republicans win three, or sweep all four, that’s a wave. However, counting normally Republican voters in Republican-tilting states voting Republican in a Republican-tilting year is not something that constitutes a wave. A wave is really when one side wins the lion’s share of the purple swing states or pulls off a bunch of upsets. That is a wave.
It’s become common to lump together the “red states” as being more or less alike. But Cook points out that six of the Senate seats that could flip from D to R tomorrow are in states Romney carried by 14 points or more in a presidential election with its significantly more favorable dynamics and turnout patterns. Winning those shouldn’t be considered a “Republican wave” if the term means anything at all other than spin.