A new New York Times/Siena College poll released Sunday shows Vice President Kamala Harris and former President Donald Trump tied at 48 percent each in the national popular vote, confirming the trend toward Trump that 538.com identified last week.

That poll and a handful of other new ones have pushed Harris’s margin in its national average down from +1.7 to +1.5, below the margin the site’s model believes Harris needs to win a majority of Electoral College votes this year. Currently, the site’s model gives Trump a 54-in-100 chance of winning the election, and Harris a 45-in-100 chance.

However, the closeness of the race bears repeating (according to 538) what the site has said over the years (since 2008): A close race in the polls does not necessarily mean the outcome will be close. All seven swing states are still within a normal polling error that goes to the candidate who is currently “losing” in each of them. While the polls have projected a close race, the 538 model’s forecasts are based on a combination of polling and campaign “fundamentals,” such as economic conditions and party sentiment in the state.

The proliferation of presidential election polls may make them seem like a game of hyper-personalization, when what they should do is determine the political trends in each electoral district in the United States.

Polls in American politics are relentless, unnerving, and frustratingly contradictory. There are national polls (that is, in every state), polls in the seven swing states, polls from small counties that predict an entire election, and partisan polls designed to frustrate the other side.

But polls may not reflect reality, and sometimes they seek to shape it. And it’s not just what the polls say, or even how they’re compiled, that’s the bigger problem here, experts say, as the obsession with polling has come to define the way Americans view politics.

It is noticeable that the Democrats are suffering from panic because of the opinion polls, as the presidential race is unnervingly close for them, especially since Harris appeared at the end of last July to be rushing with a force that is difficult to stop.

Although polls show Harris and Trump still neck and neck, Democrats are pessimistic and Republicans are confident, eight days before Election Day.

But if you look at the Guardian's 10-day average of polls for Harris, for example, the Democratic nominee remains in position to maintain a one-point advantage over her Republican rival, 47% to 46%.

Polls in the seven swing states are also equally close and offer little clear evidence about who will reach the 270-vote threshold in the Electoral College, the magic number that puts the winner in the White House.

According to polling averages in the seven battleground states, Harris leads by 1 point in Michigan and by less than 1 percentage point in Pennsylvania, Georgia, Wisconsin and Nevada. Trump leads by 2 points in North Carolina and by 1 point in Arizona.

Taken at face value, the numbers are neither catastrophic for Harris nor a victory for Trump. If they match the outcome on November 5, Harris will win a majority of the electoral college votes.

The latest polling data from FiveThirtyEight, a polling organization, also showed that Trump’s recent surge in approval ratings may have peaked. The site’s latest odds for Trump, based on a range of national data, dropped from 53% on October 21 to 51% for Harris’s 49% by the evening of October 24. The Economist’s forecast also showed Trump’s odds dropping from a peak of 56% on Wednesday (10/23/24) to 53% the following day (Thursday, 10/24/24).