The battleground in the fight for control of the House is starting to come into focus with 99 days to go until the November election. It’s not exactly the battleground that analysts expected.

It’s not dominated by well-educated, suburban districts that voted for Hillary Clinton. Instead, the battleground is broad, and it includes a long list of working-class and rural districts that voted for Donald Trump in 2016.

The broader battleground is a positive development for Democrats. It’s a reflection of how much the Republican structural advantage in the House has eroded over the last year. What remains of it isn’t helping the Republicans as much as analysts assumed it would, at least not yet.

The broader battleground has also opened up a gap between two common ways of thinking about the midterms. National polls and historical voting patterns suggest that Democrats are only slight favorites to take the House, while early polls of individual districts, special election results and the ratings of expert prognosticators suggest that Democrats are in a stronger position.

To this point, we have mainly seen polls of the generic congressional ballot, which asks voters whether they intend to vote for a Democrat or Republican for Congress. Democrats have generally led on this ballot by 6 to 8 percentage points over the last few months, which is around what analysts believe Democrats need to have an even shot of retaking the chamber.

The Republicans still have structural advantages — gerrymandering; the tendency for Democrats to waste votes in urban areas; incumbency — but in some cases these have weakened.

A flurry of Republican retirements has led to 42 open seats, many of them the sort of well-entrenched incumbents in competitive districts whose retirements are the most valuable for Democrats. The Democrats have succeeded in recruiting well-funded and strong candidates in many of the battlegrounds, which has tended to lessen the advantage of incumbency even in the districts where Republicans are running for re-election. A court decision in Pennsylvania has eliminated the party’s gerrymander there.

On paper, the Republicans still have a big geographic advantage: There are only nine Republican-held districts that voted more favorably for Democrats in the last two presidential elections than the rest of the country did. But that advantage doesn’t seem to be helping the Republicans as much as it has in past cycles, when congressional election results were increasingly correlated with presidential results.

Instead, Democrats appear highly competitive in many conservative districts. Already, there are polls showing Democrats ahead in Kentucky’s 6th District, West Virginia’s 3rd, North Carolina’s 9th, New York’s 22nd and Montana’s at-large district. Trump won each by at least 10 points.

One possibility is that Democrats are unexpectedly putting conservative districts into play because the overall national political environment is more favorable to Democrats than the generic ballot polls imply. Another possibility is that a district’s presidential vote choice will play a smaller than expected role in determining how a district will vote for the House.

Indeed, there aren’t many polls showing Democrats excelling in the well-educated districts where Clinton won. Polls sponsored by Democratic groups have shown Republicans leading in Illinois’ 6th, Pennsylvania’s 1st, Washington’s 8th and California’s 39th. Even in the well-educated districts where Democrats lead in recent polls, like Virginia’s 10th or California’s 48th and 49th, the polls show Democrats merely running even with Clinton.

Individual House polls are never much better than a rough indicator, and this is an early stage of the race. But the overall pattern is fairly clear, and a similar pattern shows up in the special election results of the last year. Democrats have run far ahead of Clinton in white working-class areas that backed the president by a wide margin, including Conor Lamb’s victory in Pennsylvania’s 18th District in March. They haven’t run so far ahead of Clinton in the areas where she excelled, like Georgia’s 6th or Northern Virginia.

Recruitment could be part of the reason. The Democrats succeeded in luring many of their longtime top recruiting targets into the race in a lot of white working-class districts. They haven’t typically done a great job finding top-tier candidates in districts that supported Clinton, in part because there are fewer elected Democratic officials to recruit in the traditionally Republican areas where she often excelled.

All of this could change by November, but as we near the start of the general election season, the initial battleground map extends well beyond the districts she won in 2016.

The most vulnerable Republican-held districts are only somewhat better educated and somewhat more suburban than the country as a whole. They are broadly representative of non-urban America. They backed Trump for president. About 31 percent of residents have a college degree, slightly more than the national average (counting all those 18 and over).

The sheer number of competitive districts is important in its own right. On paper it would be enough to make the Democrats fairly clear favorites, if one assumes Democrats would do as well in each category as the party out of power has done in recent wave elections. The Cook Political Report currently rates 60 Republican-held districts as either “lean Republican” or better for Democrats. That’s the sort of number that provides ample opportunities for Democrats to find the 23-seat gain they need.

The number of competitive Republican-held seats is far greater than it was at this time in 2006, when Democrats had a more favorable political environment. Analysts then were struggling to identify how Democrats were going to cobble together the mere 15 seats they needed for a majority. In the end, they gained 31.

In 2006 and 2010, a lot of seemingly noncompetitive seats came on the board late in the cycle. There could be fewer late surprises this time around: The abundant Democratic fundraising so far has made a lot of otherwise safe-seeming seats appear obviously competitive earlier in the cycle.

If there’s an upside for Republicans, it’s that the fight for control will often be fought in Republican-friendly districts where the president won in 2016. Republicans can reasonably hope to gain in some of these districts once the campaigns get going and pull voters back into their traditional camps.

In other words, Republicans could hope to avert a big Democratic win by trying to make their geographic advantage work as well for them as it has in recent elections. That would tend to lock the Democrats into the disadvantageous playing field implied by recent presidential election results.

To do it, Republicans might try to play up the hot-button issues that defined Trump’s coalition, like immigration and trade. Republican Ed Gillespie used a similar strategy in the Virginia governor’s race last year and wound up polarizing the electorate along the lines of the 2016 election.

That wound up being a bad trade for the Republicans in Virginia, a particularly well-educated state that voted for Clinton by 5 points. It could be a more useful option this fall, since the 60 most vulnerable Republican seats in total voted for Trump by 3 points.

With 99 days to go, whether Republicans can succeed with a strategy like this will probably determine whether we’ll have a fairly close fight for the House, as a lot of fundamentals-based analysis initially has suggested, or a clearer Democratic advantage.

© 2018 The New York Times Company