Not true.
No one saw Cantors defeat coming.
The lesson is you never know what may happen.
Because pretty much no reputable polling firms were looking at this race. Had they been, we would likely have known it was close and Cantor may lose.
The lesson is you never know what may happen.
True, you don't know what might happen far in advance and conventional wisdom is often laughably wrong - but the closer you get to a primary election it is usually possible to identify contests that will be at least competitive, and in general elections polling is quite accurate.
Few people saw Cantor's defeat coming because they simply weren't looking. And you can't count "internal polling" because what is released is hopelessly biased.