Recent political history shows that the course of a fall election is almost always set by Memorial Day. RealClearPolitics Senior Elections Analyst Sean Trende recently noted that “election outcomes are more-or-less baked in” by the end of the second quarter of an election year. Not even the financial crash of 2008 made a significant dent in that year’s outcome, which Trende says was largely expected in May of that year. One probably needs to go back to the Cuban missile crisis of 1962 to find an event that might have significantly helped the party in power on the eve of a midterm vote.

Political waves also take predictable courses, and the final outcome is almost always worse for the losing party than analysts predicted six months out. In May 2010, the Rothenberg Political Report (now Inside Elections with Nathan L. Gonzales) projected a big Republican year, with GOP House gains of two to three dozen seats. Their final pre-election forecast predicted gains between 55 and 65 seats. The GOP ultimately picked up 63 from Democrats.

In 2018’s Democratic wave, Cook’s June 1 House ratings listed seven GOP-held seats as leaning or likely Democrats and another 23 as toss-ups. Its final pre-election analysis had 18 GOP-held seats in Democratic territory and another 29 as toss-ups. Democrats gained 40 seats and retook House control.