While the political world is awaiting the first presidential debate to kick off on Thursday night, analysts say the debate’s actual effects on the race won’t be clear until days or even weeks later.

While there will be snap polls conducted during and immediately after the debate, the founder of Race to the Whitehouse, Logan Phillips, tells the Sun, “What you really want to look for is polls that are conducted after the debate happens.” 

“You’ll see the initial impact, and that will probably crescendo five, six, seven, eight days after the debate,” Mr. Phillips says. “If it’s sustained a week or two weeks later, then it might be a much more meaningful event that can change the course of the campaign.”

In 2020, the polls shifted after the first presidential debate, which was held on September 29 that year. 

On the day of the debate, the FiveThirtyEight average of polls had President Biden leading President Trump by seven points on average, 50 percent to 43 percent. Two weeks later, Mr. Biden’s lead had expanded to about 10 points, 52 percent to 42 percent.

There was less movement in the polls after the second debate, which was held on October 22, 2020. On average, Mr. Biden’s lead shrunk to just more than eight points on Election Day from just below 10 points nationally, though this had more to do with an increase of support for Trump than a decrease in support for Mr. Biden.

How much these shifts were directly attributable to the debates in 2020 is hard to say. Mr. Phillips notes, “This stuff doesn’t exist in a vacuum so it’s hard for us to tell how it affects people.” 

Mr. Phillips also said that polls of debate watchers that are released immediately after the debate are limited in scope and that they only measure the impact among some of the most politically engaged voters who view the debate and take the survey.

Yet he also noted that because the debate Thursday is happening months earlier than is typical, it suggests that “there’s less interference from other events.”

At the same time, Mr. Phillips says “It might make the debate a little less consequential.”

“It’s probably good news for Trump that it’s this early,” Mr. Phillips says. “He has a routine on these things — he doesn’t prepare for the first debate and then prepares way harder for the second one.”

Simultaneously, he says, “I think that Biden does want to change the nature of the race and enter September in a stronger position,” adding that “Biden is hoping to prove that he is stronger than he is.”

In this respect, Trump’s campaign has often focused on portraying Mr. Biden as old, incompetent, and senile, setting the stage to make it easy for Mr. Biden to overperform expectations. Mr. Phillips says that these expectations matter.

“Donald Trump’s campaign has been successful in painting Joe Biden in a negative way and making it look feeble,” Mr. Phillips says. “I also wonder if we are going to see the reverse of what normally happens,” referencing the trend that incumbents typically underperform in the first debate.

Trump’s campaign has already begun attempting to hedge against an overperformance by Mr. Biden, claiming repeatedly and without evidence that Mr. Biden will be on some sort of performance-enhancing drug for the debate.

A political scientist at John Jay College, Brian Arbour, is more agnostic about when the effects of the debate will come to roost in polling, saying, “The real honest answer is that I don’t know.”

Mr. Arbour presents two reasons why it’s hard to predict. The first is that  “effects of debates are really small, particularly in polling effects and usually by the end of the campaign they have evened out.”

In 2012, for instance, President Obama was widely seen to have had a bad performance in the first debate, though the effect had largely disappeared by Election Day.

“The polling nerd part of this is that it has to do with who is more likely to take a survey,” Mr. Arbour says. “It could be ephemeral even by the best standards.”

Mr. Arbour also suggested, like Mr. Phillips, that Thursday might see a bucking of the trend of incumbent presidents performing poorly in the first debate, though he has a different reason.

“The speculation of why presidents do poorly is because they get challenged in a way that is different from how their life has been in the last three years,” Mr. Arbour says. “Who challenges Trump internally? … He seems to be even more in a bubble.”

In terms of who stands to gain more from a strong debate performance, Mr. Arbour said that both candidates had good reasons to debate and to want to overperform expectations.

“In general I think Biden has more at stake and part of that is my general theory that I don’t know what arguments Trump makes to make people like him and want to vote for him,” Mr. Arbour says. “It seems to me that much of a variance is on the Biden side. He has both more to gain and to lose than does Trump.”

For Trump, it was in his interest to avoid being seen as trying to dodge a debate.

“In 1992, the Bush campaign seemed to be ducking a debate so the Clinton campaign sent people in chicken suits to Bush events and it became much more of a story than it should have been,” Mr. Arbour says.

The lead data scientist at Decision Desk HQ and a professor of data science at Washington University in St. Louis, Liberty Vittert, shared a similar perspective on what the debate could mean for the candidates, saying, “Biden has more to lose, in my opinion.”

In the Decision Desk HQ model, Trump has a 58 percent chance to win the general election, with 77 electoral college votes in play as toss-ups. Ms. Vittert noted that that is just slightly higher than the 56 percent chance the model gave Trump when it was launched a month ago.

“In that month we have had a Trump and Hunter Biden conviction, VP speculation, and some interesting fundraising, i.e., lots of news but little movement,” Ms. Vittert says.

Ms. Vittert added, “Depending on whether there are goofs or big hits on either side we can expect poll numbers to shift, but I can’t see them shifting more than a few percent personally.”

She concluded that those few percentage points could matter because “this race is going to be decided on a few (if not less) percent.”

The debate will be hosted by CNN on Thursday night, starting at 9 p.m. EST. It will be moderated by CNN’s Jake Tapper and Dana Bash.