Over the last weekend the final grades for 2024 were updated on our polling grade page linked above. The last section that was missing were pollsters who polled more than one state, but not necessarily just the 7 battleground states. I include the letter grade for each, and the average skew toward Trump or Harris. These are REAL grades, based on actual mathematical performance, not like some who give high grades to pollsters who were way off.





33 responses to “Polling Grades Update”
First
LikeLike
DW,
So in your knowledgeable, investigative opinion, which three pollsters are the best in accuracy?
LikeLike
Nice work, DW. Thanks.
Chicon
LikeLike
Paladin Polling was the best but it shut down.
LikeLike
LikeLike
Walt, I will do better than three…here are my top 15:
(1) AtlasIntel – the current gold standard.
(2) OnMessage-R – solid both on the average miss, and the skew.
(3) RMG Research – Run by Scott Rasmussen.
(4) Patriot – Lesser known but were really good on the 7 battlegrounds.
(5) Democracy Institute – Laughed at by RRH Elections and not taken seriously due to GOP skew…but results were pretty good.
(6) Rasmussen – Tops on the battleground state skew at only 0.29 to the blue.
(7) Trafalgar – A little off in 2024, in the direction of a blue skew.
(8) Insider Advantage – Solid and consistent.
(9) Quantus Insights – Newer pollster still working out some kinks.
(10) Big Data Poll – Rich Baris only polled three states for some reason.
(11) North Star Opinion Research – They should have polled more.
(12) Co/Efficient – On the money, but only polled two states.
(13) Civiqs-D – The most accurate Dem firm in 2024.
(14) Embold Reserach – only polled two
(15) Emerson College – B- grade not bad considering they polled 21 states
LikeLiked by 1 person
“Paladin Polling was the best but it shut down.”
Paladin will rise again.
LikeLike
test
LikeLike
I guess IncaPolls is “hors concours”
LikeLike
I had a dream not too many years ago as to how I started a marketing company named RMS (Research Marketing Strategies). We were located in Lexington KY.
I started out doing work for government entities, educational institutions and companies (large and medium). We did work for states who were considering changing their laws to allow recreational marijuana, deciding the best way to increase the enrollment in their colleges, etc., etc The business was successful. But I decided to add an experimental branch to do polling for candidates. It was a novel polling plan we came up with.
We used AI to locate 500 people in 5 key states with key races. We offered them $20 cash plus hundreds of dollars of restaurant and business product discounts/coupons we would send them for restaurants and products that THEY chose from a list of possibles. In return, the people would agree to sort of be our “employees” who would cooperate with us truthfully on our polling and questionnaires we would send them. We would end up with only 150 to 200 people in each state who would be working with us.
Communication to and from those in our poll group was all by email–so no mailing costs. The restaurant and business coupons and discounts were already being offered by those businesses and we were just going through and sourcing and sending them to our 150-200 people for their agreeing to work with us over a several months period. And sending them to the people who we knew would want those specific coupons/discount offers.
In the 5 states we started with (that all had very competitive races), we were off by less than 1% of the actual election results–but the other big polling companies were off by 4 or 5 or more %. Suddenly, we were deluged with work from candidates on state and national level.
That division of our business became the number one income producer.
LikeLike
jason yupanqui
October 16, 2025 at 9:05 am
Predictit has SCOTUS striking down Trump’s tariffs 60-40 chance.
I have mixed feelings about this.
I adamantly oppose the tariffs and would love to see them sh-tcanned, but at the same time I think commercial/trade policies belong to the executive branch.
Ideally the US should be the beacon/advocate for free enterprise, free markets, free trade in the world without SCOTUS intervention.
Bitterlaw
October 16, 2025 at 9:58 am
I saw that Trump may attend the SCOTUS oral argument. As an attorney, I recommend that he not do it. Both can be true.”
I agree. I think the justices might feel like Trump is trying to intimidate them and it could backfire.
LikeLike
In my dream, the people we targeted were computer generated ones that fit all the key geographic and other demographics of that state….
LikeLike
DW,
Thanks for the list and comments regarding them.
LikeLike
Interesting that Atlas Intel is the gold standard, they are not even a US company.
They are a Brazilian company headquartered in São Paulo.
Their polling has been good not only in the US, but for various European and South American elections too.
LikeLike
Maybe Trump believes the best way to prove that SCOTUS is not a rubber stamp of his policies is to show up and glare at the Justices.
LikeLike
Recent article in my local paper.
“MARTINSBURG [WV] — On Monday, a group of people gathered beside Williamsport Pike in the 5700 block for the rededication ceremony of the Stonewall Jackson monument.
The monument was first placed by the United Daughters of the Confederacy (UDC) West Virginia Division Berkeley County Chapter 264 in July 1929. When U.S. Route 11 was widened, the monument was relocated.
Just a few feet away from the monument, General “Stonewall” Jackson was sitting under an oak tree giving command during battle. As federal troops fired on them, a cannon ball ripped through the air above Jackson, cutting off a limb from the tree. However, Jackson walked away unharmed and unfazed.”
LikeLike
And what AtlasIntel demonstrates is that most of the rest of them COULD be accurate if they wanted to be so. But too many pollsters cannot resist the urge to put the thumb on the scale because they desperately want one side to win.
LikeLike
Got the dishwasher started, a load of laundry started.
Working on other things on the to do chores list the wife left before she drove off to do other things…
**************************************************
I will let you all in on a secret–BUT YOU HAVE TO PROMISE NOT TO SHARE IT!
>>>>>>>>>>>>>>>>>>>>>>>>>>>
Got a call the other day from the local chamber of commerce. They told me that my wife was nominated by a number people to receive their Citizen of the Year award at their annual banquet next month. The lady told me my wife was indeed selected to receive this award. I had to keep it a secret and somehow find a way to get her there on that night. it is a surprise!
*******************
Behind every good woman is a good supportive man. 🙂
NOW DON’T LET THE CAT OUT OF THE BAG
LikeLiked by 1 person
Will there also be Rommel and Tojo statues?
LikeLike
Congratulations to your wife. She has time to be a great citizen because you do everything at home.
LikeLiked by 1 person
Great job on the polling grades data and that VA gov comparison, DW.
LikeLike
Thanks Ffa2025. One of the things that fascinates me about the psychology of the polling industry and the way people react to it, is the fact that its considered normal and perfectly acceptable for any pollster to skew blue by 2 to 4 points. Even pollsters that go further than that are considered standard and revered.
But if ANY pollsters skew even just a tiny bit to the red, its the end of the world, there is no forgiveness or recovery, and they ever after are considered a disreputable pollster to be laughed at and ignored by all.
If the polling industry was fair and balanced, we would expect just as many to skew red as blue. But click the link, scroll down, and see for yourself.
LikeLiked by 1 person
Just to be clear, the letter grades are based on the average miss, not the average skew. Now for pollsters who polled just ONE state, those two metrics are the same. But if, for instance, you have a pollster who polled two states, say PA and WI. Lets say in PA, they said Trump would win by 12 (he won by 2), meaning they skewed red by a whopping 10. Then lets say in WI, they said Harris by 9 (trump won by 1), meaning they skewed blue, also by a whopping 10. The average skew would be right on the money at 0.00, but the letter grade is based on the average OFF the final results meaning the pollster was off by 10, and deserved the F grade.
This is why Quinnipiac got an F, but their skew wasn’t as bad as others.
LikeLike
Parity now as to who is at fault re gubment shutdown.
https://www.politico.com/news/2025/10/16/shutdown-blame-trump-congress-poll-00611275
LikeLike
I don’t give any polling any credit as to who is to blame as these polls are so easy to rig to get the results you want. Just front load key questions, and then ask who is to blame, and you get what you want out of it.
LikeLike
Thanks DW on the polling info. As I understand, Atlas Intel was the most accurate polling in 2020 both for national and the swing states.
About a month ago Atlas Intel released a poll on Trump’s approval using ‘adults’…..came in at 45% approval but mentioned that Trump’s approval rose with Black vote.
PS-nervously looking forward to tonight’s game (Blue Jays/Mariners). M’s gotta win with Louis Castillo taking the mound. Most of those Blue Jay fans come from Vancouver and some take the Black Ball ferry from Victoria. Translation…..baseball is huge in Canada especially the ones in British Columbia.
LikeLike
“AOC says “rivers were on fire” because of corporations like Deloitte “pouring chemicals” into waterways.
Deloitte is an accounting, consulting, and tax services firm.”
LikeLike
Are there actually any Canadian players on the Canadian team?
LikeLike
“AOC says “rivers were on fire”
You must not have gotten the memo.
It no longer matters what Democrats actually say. They can say they want the state GOP house leader killed with two bullets to the head, saving the lives of Pol Pot and Hitler, and his children dead. Virginia Democratic Party says, that’s A-Okay. So does Abigail Spanberger, Tim Kaine and Mark Warner.
LikeLiked by 1 person
HHR Putinistas janzam, Tina and Chicon hardest hit.
“A combination of falling oil prices and Ukraine’s increasingly technological advantages in conducting long-range strikes on Russia’s oil industry has become the greatest threat to the continued rule of Russian President Vladimir Putin.
The former KGB Lt. Col. “has his back against the wall. He is losing the economic war faster than he is gaining any military advantage in Ukraine,” writes long-time foreign policy analyst Ambrose Evans-Pritchard in Friday’s edition of the London Daily Telegraph.
“Putin is losing the war, so prepare for escalation”, reads the title of his most recent assessment. The summer offensive that Moscow was sure would result in a dramatic change in the situation on the battlefield did not achieve the Kremlin’s outcome. Putin’s military suffered some 800 casualties a day and ended up with little on the ground to show for it.
Russia also failed in its primary strategic objective: to break Ukraine’s fortress belt—a chain of well-defended cities and fortifications in the Donetsk region of the Donbas. The attempt to turn the tide of the war in Moscow’s direction fell well short of expectations.”
LikeLike
CNN launching new subscription service.
I think its based on the theory that people who wouldn’t watch CNN for free would watch it if they had to pay for it.
Pure genius.
LikeLike
Trump is going to meet with Putin in Hungary. After the Ukraine/Russia war ends, the Warmongers will be in desperate need of a war to watch while eating Cheetos.
Chicon
LikeLike
New Thread
LikeLike