'Most Accurate Pollster' of 2020 Releases Rare Poll, Adds Insult to Injury for Biden
On Thursday, President Joe Biden delivered a debate performance so cringe-worthy that one had to see it to believe it.
Then, on Saturday, the most reliable pollster from the 2020 presidential election gave Democrats another reason to cringe.
AtlasIntel, a low-volume pollster with a sterling record of accuracy late in the 2020 presidential election cycle, released a poll that showed former President Donald Trump with a 5.2 percent lead over Biden nationally — a result that, when applied to the Electoral College map, would result in a landslide victory for Trump in the 2024 election.
Trump earned 45.5 percent of the vote compared to Biden’s 40.3 percent. Independent candidate Robert F. Kennedy Jr. polled a distant third at 10.3 percent. No other candidate received support from even one percent of poll respondents.
📊🇺🇸 ATLAS/CNN BRASIL POLL – US ELECTION
Trump leads the race for the White House with 45.5% of voting intentions, while Joe Biden sits at 40.3%, according to the Atlas poll released on GPS CNN this Saturday.
Robert Kennedy Jr. is the preferred option for 10% of the electorate. pic.twitter.com/HNT5a6z81Z— AtlasIntel (@atlas_intel) June 29, 2024
The AtlasIntel poll attracted attention because the outlet delivered the most accurate polls of 2020.
In fact, on March 25, 2021, establishment pollster Nate Silver of ABC News’ FiveThirtyEight posted a chart to the social media platform X — then known as Twitter — showing how 25 polling outlets performed in the 2020 presidential election based on average error.
“The best-performing pollster was AtlasIntel!” Silver wrote in an accompanying tweet.
In the three weeks preceding the 2020 election, only six of those 25 outlets conducted fewer total polls than AtlasIntel’s 14.
Nonetheless, with only a 2.2 percent average error, AtlasIntel established itself as the most reliable pollster.
“Yeah, they incorrectly had Trump winning a few states, but they were close on the margins, and that’s the better metric,” Silver added.
So the most accurate pollster had Trump winning states he actually “lost”? Go figure.
Finally, here’s how the most prolific pollsters fared in the general election.
The best-performing pollster was AtlasIntel!
Second was Trafalgar!
Yeah, they incorrectly had Trump winning a few states, but they were close on the margins, and that’s the better metric. pic.twitter.com/yqfbjCequB
— Nate Silver (@NateSilver538) March 25, 2021
Furthermore, Trafalgar Group and Rasmussen Reports ranked second and third, respectively, on Silver’s list of most accurate pollsters.
Trafalgar has not conducted a general election poll since March. On June 20, however, Rasmussen released the results of a poll showing Trump with a 9-point lead over Biden head-to-head and a 10-point lead in a race that included Kennedy and other long-shot candidates.
In other words, the third-most accurate polling outlet showed Trump with an even bigger lead nationally.
The polling news of the weekend, however, belonged to AtlasIntel. And that pollster made a point of highlighting its new poll’s likely electoral consequences.
Sunday on X, AtlasIntel shared an eight-minute YouTube video from political analyst Lester Tellez of “On Point Politics.”
“The team at @onpointpolitics looked at the AtlasIntel figures released today and studied the implications for the electoral map. Unsurprisingly for a +5 advantage, the forecast looks very strong for Trump at this point,” AtlasIntel wrote.
The team at @onpointpolitics looked at the AtlasIntel figures released today and studied the implications for the electoral map. Unsurprisingly for a +5 advantage, the forecast looks very strong for Trump at this point. You can find the forecast here: https://t.co/OsIKLeDdFZ
— AtlasIntel (@atlas_intel) June 30, 2024
For those who follow polling, Tellez’s video could hardly have been more shocking.
First, Tellez presented a “demographic model breakdown” of AtlasIntel’s results. This involved analysis of the usual categories: age, education, sex, income and race.
After examining respondents’ answers to issue-related questions, which showed Trump holding double-digit leads on immigration, the economy and foreign policy, Tellez showed how AtlasIntel’s poll would translate to the Electoral College.
In sum, the hitherto presumptive swing states no longer qualify as swing states.
Tellez’s breakdown of AtlasIntel’s results gave Trump an enormous 13-point lead in Georgia and an even more eye-opening 14.9-point lead in North Carolina, which pushed the Tar Heel State close to the “safe” category for the soon-to-be Republican presidential nominee.
Likewise, the Democratic Party’s former “blue wall” has turned red. Tellez gave Trump a 10.3-point lead in Pennsylvania, a 10.2-point lead in Wisconsin and an 8.4-point lead in Michigan.
The Southwest has also turned heavily toward Trump. The former president leads by 10 points in Arizona and 9.2 points in Nevada.
In the 2020 election, Biden received electoral votes from all of those states except North Carolina.
Meanwhile, the only “actual” swing states are Minnesota, New Hampshire, Virginia and New Mexico.
Of those four swing states, the AtlasIntel results showed Biden winning only New Mexico.
All told, Trump prevailed in an Electoral College landslide, 342-196.
Americans have every reason to greet polls and even election results with a degree of skepticism. But when individual pollsters have a track record of success, their results deserve careful attention.
Results like these, of course, would amount to a political earthquake and a well-earned repudiation of the corrupt establishment.
Truth and Accuracy
We are committed to truth and accuracy in all of our journalism. Read our editorial standards.
