Since the 1920s we've had one president (HW Bush) who fits the bill under normal circumstances.
The following were all oddities:
* Nixon -> Ford
* JFK -> Johnson
* FDR -> Truman
It doesn't have to be statistically significant to be a pattern. In fact, one of the few presidents to have a proper "successor" was Roosevelt with Taft and that didn't turn out so well...
Reagan to GW Bush was the anomaly really. Even with Obamas popularity its not surprising there was a party changeover - unfortunately the candidate from the other side was trump.
He had the largest landslide victory of any two term president.
He basically won every state. And that goodwill carried over to George HW Bush's campaign.
> The following were all oddities:
> Nixon -> Ford
> ...
Ford wasn't even elected to the presidency, NOR the vice presidency-- he was appointed VP after Agnew resigned. And then lost to Carter in '76.
It's certainly controversial if used as a defense for why a party 'couldn't win,' which is how I read the original comment. It's not like there is some magical system that conspired to elect the opposite party for Clinton, Bush Jr., Obama, etc. but didn't act for Reagan, Coolidge, etc. The fact of the matter is that poor presidencies (or presidencies that the public did not correctly identify as poor) played a large part in re-election of the party, perhaps more so than the running candidates. And on average, we have had a lot of presidents who leave much to be desired, so it flip flops as a result.
If it's not being used as a defense, then it's just an incorrect point, because as you said, there is nothing statistically significant either way, which is basically what the original comment asserts by declaring it impossible.