In the final hours of this midterm cycle, we asked a group of top practitioners what they’ll be watching for on election night 2018 to get a feel for how the final results could swing.

Forget predictions, we wanted to gauge what bellwether district or statewide returns could foretell whether this will be a Democratic wave year, a triumphant GOP defense of the Hill, or a split decision.

Here’s what they told us:

Patrick Ruffini, partner and co-founder of Echelon Insights:

History is littered with examples of past bellwethers that are suddenly no longer bellwethers because the ground shifted. So, I’m interested in broader shifts and will be using modeling throughout the course of the night to figure out which kinds of districts across the entire battleground shifted the most off the 2016 baseline.

I’m particularly interested in two areas with 2020 implications: Do Democratic gains in Clinton suburbs extend out to suburban seats Trump carried narrowly? And exactly how well will Democrats do in the former “blue wall” and rust belt states of Pennsylvania, Michigan, Ohio, Wisconsin, Minnesota, and Iowa? Big Democratic wins there could provide a template for 2020.

Kelly Grace Gibson, partner at Hamburger Gibson Creative:

Pennsylvania will be a really interesting sign — how much does [Gov. Tom] Wolf win by? Do we take back the state House/state Senate?

And obviously in Georgia: where does the Kemp-Abrams race land? Do the Dems pick up GA-07? Do we pick up GA-06? Florida will be huge. If Andrew Gillum wins the governor’s race it is likely that [Bill] Nelson wins the Senate race.

Stefan Hankin, president of Lincoln Park Strategies:

We are going to be watching VA-10 to see how far behind [Rep. Barbara] Comstock is running — also VA-2 and VA-5 for indications on what the engagement level is in not just suburban districts, but more rural districts.

If Comstock is well behind and Democrats are close or ahead in VA-02 and VA-05 it’s likely to be a really good night. If Comstock is close, and 2 and 5 are seeming out of reach early, then it will be looking more like Dems coming up short.

Florida and Ohio are also of big interest. Most polls are showing Gillum and Nelson up, but on Trendency we have these neck and neck. Florida is always interesting, but you have two things going on that is making research a little tricky right now. The growing population in the Orlando area of Puerto Ricans and displaced residents in the panhandle due to the Hurricane. Our guess is both these groups are being under-represented in surveys.

In Ohio three weeks ago I would have said Richard Cordray is going to lose. In the last two weeks, he has closed the gap. Still, [we] put his odds slightly below 50/50, but they have gotten better.

Tracy Dietz, CEO of DonorBureau:

If Beto [O’Rourke] pulls it out that will be a huge slap in the face for Republicans. Even with all the chatter about turning Texas blue, it’s still a heavily red state. If Beto wins or even keeps that race really close, it’s a bad sign for the right. I don’t think he has a shot, by the way, but him winning speaks volumes.

If [Claire] McCaskill pulls out Missouri, that isn’t a devastating outcome for the GOP. That race has always been close.

Florida is probably the state to watch. Trump won Florida by 1.2 points. If Nelson wins or loses by a wide margin, I think it’s a good sign of what’s about to happen. If he wins, maybe Trumpisms are wearing on people. If he loses, people are still supporting Trump.

Gillum vs. [Ron] DeSantis has been a crazy race. Florida hasn’t had a Democratic governor in 20 years, so if Gillum wins that’s bad for the GOP.

Chris Talbot, president of Talbot Digital:

NH-01 has flipped every cycle for a decade, going red in the midterms each time. But look for Chris Pappas to keep the seat blue and carry on the great work of retiring Granite State legend Carol Shea-Porter.

Brendan Steinhauser, co-founder of Steinhauser Strategies:

Will Hurd TX-23 will be a bellwether. It’s been decided by 2 points or less over last two cycles. Will has won twice, but Democratic energy is real. If Will wins, I think the GOP could run the table in Texas. It’s a majority Hispanic district.

Cheryl Hori, founder of Pacific Campaign House:

My eyes are on Sri Kulkarni running for Congress in TX-22. This race has a lot of implications not only for Democrats’ hopeful blue wave, but also for the importance of outreach to the Asian American and Pacific Islander (AAPI) community and other diverse pockets of voters.

Kulkarni is running in Fort Bend, one of the most racially diverse and fastest growing congressional districts in the country. It’s also gone to Republican Pete Olson who beat out his Democratic opponents by 19 percent in 2016 and 35 percent in 2014.

It would be a misstep to ignore that TX-22 is a true melting pot with nearly 20 percent Asian Americans, 13 percent African Americans, and 25 percent Hispanic/Latino Americans. And if Kulkarni wins, it’ll be a clear signal that the era of white, male, Republican politicians dominating Texas elections is coming to an end.

Outside of the blue wave, Kulkarni’s team along with groups like The AAPI Collective and APIA Vote are pulling out all the stops for an AAPI wave in this district and across the country. The campaign is phone banking in 15 different languages and it doesn’t hurt that the man himself speaks 6 different languages. The AAPI community is the fastest growing voting block in the country and in this district — winning their vote will make or break the outcome of the election.

Over the course of this cycle, the Cook Report took TX-22 from safe Republican to likely Republican to lean Republican. If the son of an Indian immigrant has a chance of unseating a five-term Republican in Texas, Republicans across the country should take note: you’re on notice.

Casey Phillips, co-founder of RedPrint Strategy:

I’ll be watching the whiskey bottle on my desk get emptier and emptier, [in addition to] the whole state of Pennsylvania plus NY-24 and NC-13.

Chris Turner, CEO of Stampede Consulting:

Republicans are voting early in huge numbers. In a place like Florida, for instance, we’re four times better off in early voting than we were in 2016. The two Trump rallies have really amped things up there. Trump’s been a superhero for us on turnout.

The other thing I am tracking that affects me and our company directly is the level of enthusiasm of our door-to-door workforce, which in our case is made up of military veterans, their families, and conservative activists.

We saw per-worker production numbers jump 100 percent during the Kavanaugh hearings and those numbers have held through today. Justice Kavanaugh and the injustice that was done to him by the Democrats continues to serve as the driving force for Republicans across the country.

Voters who [are] not confused about their choices and who feel as though they are righting a wrong or turning back an injustice tend to storm the polls. Republicans are benefiting from both right now.

Emily Gittleman, digital director at 50+1 Strategies:

I’ve been looking at early voting in CA-10 and CA-21. Technically they’re races we’re working on, but only on turnout generally. The early numbers were fascinating and hopefully are a really good indicator of turnout since those districts have really high numbers of unlikely voters.

Beyond that, I’ll be looking at East Coast House races, since they’ll get their results before we will.