If you were hoping that the final polls before Election Day would provide a clearer picture of the presidential contest, you were hoping in vain.

Pretty much every theory about what will happen on Tuesday has received new supporting polling results in recent polls – and also some new evidence that casts doubt on it.

Do late deciders support Donald Trump or Kamala Harris? Are the polls missing Trump voters again – or have they overcorrected to the point where they significantly underestimate the Democrats? Is Harris better on the Rust Belt or Sun Belt battlefields?

Recent polls have not provided a consistent answer to any of these questions, and uncertainty about the outcome appears to be greater than ever.

What Democrats are hoping is that late deciders will prevail for Harris (perhaps spurred by Trump's controversial rally at Madison Square Garden last week). Some new government polls suggest that may be true. The New York Times reported that its final state polls showed that “among the 8 percent of voters who said they had recently decided to vote,” Harris was ahead “55 percent to 44 percent.”

And yet, recent national polls have actually moved Trump's way, and the averages now show Harris' national lead has fallen to 1 point or less – her lowest level in months. If there were a national trend in Harris' favor, we would expect it to show up in recent national polls, but that is not the case.

Final state polling averages, meanwhile, show a race that is essentially deadlocked, with candidates trailing by 1 point or less in nearly every swing state. But Nate Silver has argued that there are clear signs of widespread “herd behavior” among pollsters—that pollsters, like sheep, adjust their results to conform to an expected close result.

Statistical principles suggest that if the race is truly tied, most polls should be close to a draw. But there should also be a lot of differences, as some polls show clear advantages for both candidates, and we're getting very few of those this year. “The odds are 1 in 9.5 trillion, while at least that many polls show such a narrow margin,” Silver wrote.

But does herd politics hurt one candidate more than the other? Both parties have reason to hope that the polls will lack support for their side. The reason for Republicans is that pollsters underestimated support for Trump in 2016 and 2020. That could happen again: New York Times chief poll analyst Nate Cohn wrote Sunday that in the Times' most recent polls, “white Democrats were 16 percent more likely to respond than white Republicans,” which “raises the possibility that that the polls could be underestimating Trump again.”

Others suspect that pollsters have overcorrected to the point where they are now overestimating Trump's support. Highly respected Iowa pollster J. Ann Selzer shocked the political world on Saturday with a poll that showed Harris gaining three points in her state, despite widespread belief that she was a sure-fire Trump candidate. state acts. Some assume that Selzer noticed a shift toward Democrats that other pollsters missed, but others suspect that her poll is just an outlier that doesn't match the results.

The details of the swing state map are also highly uncertain. The averages show Harris with a very slight lead in Michigan and Wisconsin, while Pennsylvania is essentially tied. They also generally show a slight Trump lead in Georgia, North Carolina and Nevada — and a clearer Trump lead in Arizona. This suggests that Harris' best path to victory is to hold onto the Rust Belt.

But the New York Times/Siena College poll released Sunday upset much of that map, showing Harris ahead in Georgia, North Carolina and Nevada, as well as Wisconsin — while Pennsylvania and Michigan are tied. (Arizona still went to Trump pretty heavily.) Even if the tied states went to Trump, those margins would be enough for a narrow Harris Electoral College victory.

How much value should we put in the Selzer survey?

The general advice from election experts when interpreting a barrage of polls is to “stick with the averages.” Looking at individual polls can be interesting, but any single poll can be an outlier, and it's tempting for followers to cherry-pick and see what you want to see.

As of Monday morning, the New York Times averages show:

If those results are in on Election Day, whoever wins Pennsylvania will decide the election. And yet we should not assume that these will be the results either. Final state poll averages often differ from results by several points. And given that so many of these final averages show a difference of 1 point or less, the only reasonable conclusion here is: It's really close.

Now, some numerologists tend to qualify the “stick to the averages” advice by saying that there may be a few pollsters who stand out from the rest and deserve at least a little special attention. This elite class includes Selzer's Iowa polls and the New York Times/Siena College state polls.

Both have earned respect in past election cycles for their lack of herd loyalty — for producing results that the national polling average missed.

In 2016, Selzer's most recent poll showed Trump in 7th place in Iowa, while other polls showed a closer contest. Trump won the state by more than 9 points. In 2020, most pollsters again showed a close contest, but Selzer found Trump at 7 and he won by 8.

So Selzer is a pollster who did that not She underestimated Trump – she accurately estimated her state's support for Trump in the last two cycles. And their latest poll shockingly shows Harris beating him by 3 points.

Theories that explain this are circulating in the political world. Has Selzer noticed something unique happening in Iowa — perhaps a backlash against Republicans' strict anti-abortion laws in the state? Even bigger is the theory that she could be one of the few to see a nationwide shift toward Democrats, one that other pollsters refuse to believe.

Then again, no one is perfect, and even the best pollsters sometimes get it wrong due to chance, so maybe she's just wrong!

The New York Times/Siena College poll also has a reputation for avoiding herding, but recent swing state polls are more mixed for Harris.

Interestingly, the Times' polls show Harris by 2 in Wisconsin (the tipping point state of 2020), by 3 in Nevada (a state where some early voting analysts believed Republicans looked strong), and by 3 in North Carolina (a state Trump won) by 2 in his previous two runs). Trump held four rallies in North Carolina in the final days of the campaign, which some interpreted as a sign that his team was worried about his prospects in the state.

But the Times polls also show Michigan and Pennsylvania tied, perhaps a sign that Harris can't rely on the Rust Belt after all. They also show a 1 point difference in Georgia, which we probably shouldn't put too much emphasis on.

Overall the picture is clear as mud. The polls don't tell us who will win. We could be in for an extremely close competition. Or one of the two candidates could beat their poll numbers by a few points and win quite significantly. The only way to find out is to count the votes.