Americans went to the polls on Tuesday, Nov. 3, to decide control of Congress in a major electoral test for President Donald Trump, with Republicans defending the House and Senate and Democrats seeking to check his presidency. The contest will determine Washington’s balance of power, but the answer may not come quickly: initial results are expected on Tuesday, while final results could take days to emerge.

One numerical indication in the closing campaign concerned the House. An Oct. 25 projection by the Battleground Tracker model placed Democrats on course to win 229 seats, enough to take control. Its estimates ranged from 215 Democratic seats, which would leave Republicans with a narrow majority, to 237. Those figures framed a contest in which Democrats had a projected route to power but Republicans still had a path to retaining the chamber. They were estimates, not votes counted or seats secured.

Both parties approached election day with control of the House and Senate at stake. Republicans were fighting to preserve their position in Congress, while Democrats hoped for a broad electoral advance. For Trump, the voting therefore carried significance beyond any single congressional contest: Republican retention would keep both chambers in his party’s hands, while a Democratic takeover of either chamber would change that balance. The test was whether voters would sustain the existing congressional balance or choose to change it.

Read more: Aspen Neuroscience raises $141 million in Nasdaq IPO

The House projection’s range was as important to understanding the race as its central estimate. At 215 seats for Democrats, the outcome would be continued Republican control, albeit with a narrow majority. At the projected 229, Democrats would take over. The upper estimate of 237 indicated a stronger possible Democratic result. Because the range included outcomes favouring either party’s control, the projection did not settle which side would prevail when ballots were counted.

That uncertainty also limits what can be read into the Democratic advantage in the central estimate. A forecast pointing towards a change in control is different from a result confirming one. The Oct. 25 figures offered a guide to the possible distribution of House seats before election day, rather than a guarantee of the eventual balance. They gave Democrats grounds to anticipate a takeover while leaving room for Republicans to argue that holding the chamber remained achievable.

Trump urged his supporters to vote as the campaign drew to a close. In a campaign update dated Nov. 2, he was quoted saying “we win in a landslide” if his supporters cast ballots. His message presented a large victory as conditional on participation, not as an outcome already secured. That appeal was not a House-seat forecast; the model’s estimates encompassed both a Democratic takeover and a narrow Republican majority rather than a single settled result.

Read more: U.S. strikes Iranian targets in Strait of Hormuz as Tehran asserts self-defense right

Democrats, meanwhile, were seeking enough seats to gain congressional control and put a check on Trump. Their hopes for a broad victory would have to be tested against the results of the battleground races that determine the House and Senate. The distinction matters: an ambition to make substantial gains is a campaign objective, whereas control depends on the seats actually won. Neither the Democrats’ hopes nor Trump’s prediction could substitute for the decisions being made by voters.

The Senate contest was part of the same struggle over Washington’s direction, but the House projection could not answer who would control it. The 229-seat estimate and the range from 215 to 237 applied to Democratic representation in the House. Those numbers therefore needed to be read as a measure of that chamber’s possible outcome, not as a forecast for Congress as a whole. Republicans were defending both chambers, and Democrats’ effort to change the balance extended to both.

Attention will turn from campaign appeals to counting as polls begin to close at 23:00 GMT. The first results are expected on Tuesday, though the prospect of final results taking days means the opening returns may not provide a complete answer. The immediate question will be how the battleground contests develop as results arrive. The broader question—whether Republicans retain the House and Senate or Democrats gain control of either or both—will depend on the outcomes those results ultimately establish.

For now, the election remains a choice being made rather than a verdict already delivered. The pre-election House figures pointed towards a Democratic takeover but explicitly allowed for Republican survival; Trump’s appeal and Democrats’ hopes reflected competing expectations about what voters would do. As counting begins, the measure of Trump’s electoral test will shift from projections and appeals to the seats each party wins. Until the results establish control, neither side’s preferred outcome can be treated as settled.