The most surprising — and least surprising — thing about the 2018 midterms is that we were aware of almost this exact outcome at the start of the cycle.
The big picture: The incumbents who lost were mostly predictable because they were the most vulnerable all long: House Republican incumbents in districts Hillary Clinton won in 2016 and red-state Democratic senators in Trump country.
By the numbers: After the 2018 midterm elections, 20 House Republican incumbents lost their seats to a Democrat. Of those, 10 were in Clinton-won districts, while 9 were incumbents who Republicans had previously said were in competitive, coin-flip races.
And two were genuine surprises — Mia Love in Utah, who appeared headed for a loss, and Dan Donovan in New York, who has officially been defeated.
- President Trump won the state of Utah by nearly 20 percentage points in 2016, and zero Republicans I talked to throughout the cycle ever mentioned Love's name when discussing which members were in danger.
- Donovan was New York City's last House Republican in the state's 11th district, which Trump won by 10 percentage points in 2016 and which Cook Political Report rated as "lean Republican." Even Democrats were texting me on election night in disbelief of his loss.
- Three red-state Democratic senators lost their re-election bid, and all voted against Brett Kavanaugh for the Supreme Court. Nevada Sen. Dean Heller was always considered the most vulnerable Republican senator. He had an on-again-off-again relationship with the president, and he was the only GOP senator defending a seat in a state that Clinton won.
- Both Republican governors who lost were in for a reckoning as their states' politics and demographics have shifted, and they've both faced serious challengers in the past.
The bottom line: Trump focused most of his time on Republicans running for Senate and governor, leaving many House GOP incumbents out to dry in 2018. And national events (like the Supreme Court fight) were too big to overcome for Democrats in red states.
Go deeper: