The divide in American politics is so stark that analysts are beginning to predict something that seldom happens: One party could make big gains in the House while the other adds seats in the Senate.

Not since 1970 has a midterm election provided such a split verdict, and only two other presidential elections, in 1996 and 1972, have demonstrated such division in congressional elections.

Now, particularly after the contentious Senate confirmation of Supreme Court Justice Brett M. Kavanaugh, Democratic energy is up in the suburban districts that will determine the House majority — just as Republicans claim conservative voters have been jolted awake in rural states that will determine the Senate majority.

Democrats could gain more than the 23 seats needed to take the House majority, which would normally be considered a “wave election” for Democrats up and down the ballot.

But Senate Republicans have gone from clinging to the narrowest of margins, a 51-to-49 majority, to believing they will gain seats, possibly three, the sort of result that would normally mean the national GOP had a very good night.

Charlie Cook, the independent analyst and founder of the Cook Political Report, called the Kavanaugh nomination process “a color enhancement event.” It positioned Democrats to perform even better in the metropolitan areas and Republicans to make gains in the exurban-rural regions.

“It made the reds redder and the blues bluer,” Cook wrote.

Midterm elections are supposed to be a verdict on the party holding the presidency, and almost always the tide breaks against that party, particularly in House and Senate races, and especially with an unpopular president.

Think of 2006, George W. Bush’s sixth year as president, when Democrats picked up more than 30 House seats and six Senate seats to win the majority in both chambers. Or in 2010, when Republicans picked up 63 House seats and six Senate seats in Barack Obama’s first midterm election.

The tides went the same direction, as the out-of-power voters raced to the polls and the small-but-decisive independents broke against the sitting president in similar fashion across the nation.

President Trump’s first midterm election seemed to be heading that way, as Democrats won special elections in deeply conservative places such as Alabama and southwest Pennsylvania and had a strong showing in Virginia’s statewide elections last November.

Democrats are now certain to make big gains in the House — Republicans have essentially surrendered about a dozen of their seats already — and the only questions remaining will be if they net the 23 seats needed for the majority and, if so, how big can they grow that margin.

But now, chances are growing for Republican gains in the Senate. The underpinning of such a split verdict can be found in a new report from a study by Third Way, the left-leaning think tank that analyzed voter data in 13 Senate races that will determine the majority.

Twelve of those states favored Trump in 2016 and nine of those seats are held by Democrats. To win the majority, Democrats need to successfully defend all nine of their seats and take two of the four GOP seats, and the gravity of that hurdle comes through in Third Way’s crunching of data from Catalyst, a liberal group with access to voter files.

From West Virginia to Missouri to Florida and beyond, Republicans start with an edge in terms of the voters most likely to show up at the polls in a midterm.

Across all 13 states, 40 percent of the midterm voters are likely to be “base Republicans,” regular midterm voters who essentially never split their ticket. Democrats could expect 27 percent of their “base” voters to head to the polls, with the rest being potential ticket splitters.

The numbers grow more stark when viewed in certain states.

Take West Virginia, where Trump won the state by 42 percentage points in 2016, the largest margin in any state that Democrats are defending in the midterms.

Just 15 percent of probable voters in West Virginia will be “base Democrats,” the type most likely to support Sen. Joe Manchin III (D) in his reelection bid.

More than half the state’s remaining voters have shown some proclivity for supporting Democrats, but their demographic makeup has become the party’s most challenging subset: 98 percent are white, just 1 in 5 hold a college degree and just 16 percent are under the age of 40, according to the Third Way study.

Manchin, a former governor, remains a slight favorite in the race because he is doing all that he can to run on local issues and keeping national politics out of the discussion despite attacks from Trump and others that he is a loyal Democrat. “They just try to identify you, what tribe do you belong to, and I tell them I belong to the American tribe and West Virginia is the best branch we’ve got,” Manchin said in an interview a few days before he broke with Democrats to vote for Kavanaugh.

The outlook in Texas and Tennessee is more bleak for Democrats. In Texas, 49 percent of probable voters are “base Republicans,” twice as many as the probable “base Democrats” that are expected to show up at the polls.

That’s why Sen. Ted Cruz (R-Tex.) has made almost no effort to appeal to centrist voters, believing that he can just regain support from conservative voters angry about his non-endorsement of Trump at the 2016 Republican National Convention.

In Tennessee, 60 percent of likely voters are “base Republicans,” explaining why Rep. Marsha Blackburn (R-Tenn.) has taken an approach similar to Cruz.

Democrats have adopted the opposite strategy with their nominees in those states. Former Tennessee governor Phil Bredesen is running as a centrist Democrat, who supported Kavanaugh and hopes to win a larger-than-usual share of Republican votes.

Rep. Beto O’Rourke (D-Tex.) has galvanized liberals and is trying to expand the pool of voters and bring in a larger-than-usual share of younger voters and minorities.

If Bredesen prevails, centrist Democrats will claim that was the right approach, while liberal activists will claim the high ground if O’Rourke pulls off the upset.

If they both lose, and by wide margins, it probably means they never had much of a chance — and that a historic split verdict is likely in the midterms.