Everyone take a deep breath. We’re all starving for tangible election results and now we have them. But just as earlier Republican wins in congressional special elections this year are no guarantee the party will have a good 2018, losses on Tuesday night don’t necessarily tell us a Democratic wave in the House has developed.

Democrats had to win the governorship in Virginia, and they did. After coming up short in every House special election this year in districts President Donald Trump carried last fall, Democrats didn’t have an excuse to lose a state that Hillary Clinton won by more than 5 points. And Ralph Northam responded with a resounding victory for Democrats. That being said, the win maintains the status quo considering Virginia already has a Democratic governor.

The House of Representatives was in play before the results in Virginia and the House is in play after Virginia. The president’s party has lost House seats in 18 of the last 20 midterm elections, and the average seat loss in those 18 elections was 33 seats. Democrats need to gain 24 next year for a majority. We knew this before Tuesday night and it’s still the case. Did Democratic chances improve dramatically? I think it remains to be seen.

Republicans win the federal race, while Democrats crush the state races. This trend has been going on for the last six months. On Tuesday, Republican John Curtis kept Utah’s 3rd District in GOP hands in a race where Democrats never lifted a finger. Democrats picked up a modern-era record number of seats in the Virginia House of Delegates, as well as gained state government trifectas in Washington and New Jersey. It’s not entirely surprising, considering Democrats have been taking over GOP seats in state legislatures in special elections all over the country since Trump took office. Republicans, meanwhile, have been holding their congressional seats, albeit ones that Trump carried in 2016.

Democrats are winning Democratic areas and Republicans are winning Republican areas. Democrats won two gubernatorial races on Tuesday in Clinton states. And even though they made major gains in the Virginia House, 17 Republicans held Clinton seats coming into Election Day, according to Nathaniel Rakich of Inside Elections. Democratic challengers in Virginia ran 2 points behind Clinton, on average, and Democratic incumbents ran 1 point ahead. Democrats took over a critical state Senate seat in Washington, which Clinton carried by 37 points in 2016. If the same trend held for the House of Representatives next year, Democrats would fall just short of a majority (considering there are 23 Republicans who hold Clinton districts). Democrats have shown they can boost turnout in Democratic areas, but still haven’t proved they can win in Trump country in a high-profile race.

Federal and state races are often about different issues. Democratic gains in the Virginia House of Delegates are important because of their impact on the next round of redistricting, but I’m a little leery of extrapolating results from non-federal races (particularly ones with about 20,000 total votes cast) to federal races where hundreds of thousands of votes will be cast and the issue set could be different. Republicans couldn’t rely on their regular bogeymen (or woman, in the case of Nancy Pelosi) to save them in Virginia because it wasn’t relevant to a state race. The Pelosi well might dry up over the next year, but the GOP has been able to draw from it time and again in order to motivate base voters. In addition, I’m not sure what kind of campaigns these Virginia Republicans ran, and in some cases, Democrats outspent the GOP side by a wide margin. Next year, GOP incumbents will have more resources to defend themselves. Of course, that doesn’t guarantee victory.

Republicans can’t rely on redistricting to save them in 2018. Republicans drew the current state legislative map in Virginia and are on the verge of losing control. When parties draw maps, they don’t draw safe seats; they draw safe-enough seats that will elect their candidates under normal circumstances. But if there is a larger trend or wave, more seats can become vulnerable because a party’s vote was spread thin to maximize gains.

Don’t rely too much on the national generic ballot and presidential job approval ratings. Those numbers can help describe the national political mood, but we don’t have national elections. The presidential race and Senate majority is fought on a state-by-state level, while the House is a district-by-district battle, and Trump is not uniformly liked or disliked around the country. He is doing better in the types of districts Democrats need to win for a majority. And even if you had a model that predicted the number of seats Republicans will lose, you still have to look at the district level to know precisely which seats will fall to the Democrats. For example, John Katko represents one of 23 Clinton districts held by Republicans. According to Tuesday’s results, he would lose next November. But Democrats don’t have a credible challenger against him.

It’s hard for candidates to have it both ways. Multiple Republican consultants referred to Gillespie as being in a political no man’s land. He was an establishment politician trying to message to Trump supporters without getting too close to the president. It obviously didn’t work. Would cozying up to Trump or disavowing Trump have caused Gillespie to win? Probably not. But the race might have been closer. GOP sources point out how Republican Ken Cuccinelli wasn’t interested in moderating his conservative message in the 2013 gubernatorial race and lost by less than 3 points.

Don’t lump all “pundits” together. Sure, the people on “Morning Joe” predicted a Gillespie victory, but all of the major handicapping outlets, including Inside Elections, The Cook Political Report and Larry Sabato’s Crystal Ball, had Northam with an advantage. And in general, the polling pointed to the correct winner while underestimating Northam’s strength considerably, except for Quinnipiac University, which showed Northam winning by 9 points.