I think we might be able to judge whether or not the presidential polls were accurate or not by taking a look at the National Football League (NFL).
Does that seem like an unusual tie-in? Permit me to explain.
Let’s take the next 51 NFL games. In how many of those 51 games do you think you could pick the actual winner? I’m talking about picking the winner without the aid of a point spread. They call that a straight-up bet. Don’t know what a point spread is? Don’t worry about it.
If the Los Angeles Rams were playing the San Antonio Spurs who do you think would win the game? Your choices are the Rams or the Spurs. After you looked at each one of the 51 games how many of those games do you think you could pick the winner, again, without the benefit of a point spread?
I wanted to get some professional input on this question. I went to the foremost professional sports better who subscribes to my newsletter. This fellow uses offshore accounts to make his wagers! If he’s not professional I don’t know who is.
I asked him how many winners of the next 51 NFL games he thought he could pick successfully again without using the point spread. He gave my question some thought. His answer was “34”. PB was telling me he could pick the winner of a random NFL game about two times out of three. That sounded reasonable to me.
Of course, some of those games might have a prohibitive favorite. Picking the winner in a game like that would be a little bit easier one would think. Let’s say the Pittsburgh Steelers who are currently undefeated after nine games were playing the New York Jets. The Jets have yet to win a single game in nine attempts. You would likely think the Steelers would win the game easily.
The Steelers-Jets analogy has some similarities to state presidential elections. It would be pretty easy to pick the presidential winner in the District of Columbia. Wyoming wouldn’t be that hard either right?
Now back to the NFL. If the Philadelphia Eagles, whose record is 3-5-1 (they are in first place in their conference by the way!), were playing the Cincinnati Bengals who sport a record of 2-6-1 predicting the winner might be more challenging. I am not an NFL fan. I stopped watching long ago after their third labor dispute canceled either some of the games or the entire season. However, for the life of me, I can’t figure out how a game ended in a tie in 2020.
So you are probably asking yourself, “Randy why did you pick 51 NFL games and not 50?” If you did ask that question I’m glad to see you are keeping up!
We just finished a presidential election where there were 51 electoral contests. Those contests included elections in our 50 states and the District of Columbia.
I’m going to keep this simple. I will share with you just two statistical tables. Those tables will show you how accurately the various presidential polls that I tracked were able to predict two things. First, the winner in each of these 51 contests. Second, the margin of victory in each of these 51 contests.
I will conclude that the polls were very accurate in picking the winner of each of our states and the District of Columbia. I will tell you that the polls were much less accurate in predicting the margin of victory.
I will leave it to the reader to decide whether it’s more important for a presidential poll to pick the winner or the margin of victory in a political election.
For those of you who are having major heartburn over my hypothetical NFL football game between the Los Angeles Rams and the San Antonio Spurs, I will offer this to console you. Not everyone is a sports fan. I am a sports fan. I just wanted to see if you were. Of course, everyone knows the San Antonio Spurs are not a football team but a basketball team. They compete very successfully in the National Basketball Association. I will thank those readers in advance for noticing the discrepancy and maintaining an open mind by reading this far.
Let’s take a look at table #1.
|Table #1 – 2020 Presidential Poll Predictions|
|2020 Projected winner||Electoral votes||Projected Trump||Projected Biden||Projected Poll Variance||Projected Winner|
|District of Columbia||3||3||77.0%||Biden|
This is a simple summary of which candidate was favored in the average of the polls I checked. By the way, I began tracking every poll listed by my “Pocket Polls” app in early August. I checked 163 different presidential polling organizations.
That’s right. 163 DIFFERENT polling organizations. I only used the latest predictions from each polling source. When a new poll came out I replaced the old results with the new results from that poll source.
The average of all of these polls state by state predicted an electoral college win by Biden over Trump, 350-188.
The average of these 163 polls predicted the presidential winner in 49 of 51 state electoral elections. Remember our expert sports gambling czar told me he might be expected to accurately predict the winner in only 34 of the next 51 NFL games. Getting 49 of the 51 state winners right is pretty friggin’ amazing in my book.
The polls missed picking the winner in Florida and North Carolina. The polls went with Biden in both cases and Trump won those two states. This resulted in a near-final electoral vote win by Biden of 306-232.
I can only conclude that the presidential polls, on average, did a great job of picking the winners. They correctly picked the winner in 49 of 51 contests. They can’t get much better than that. Now let’s take a look at table #2.
|Table #2 – 2020 Presidential Poll Actual Results vs. Predictions|
|2020 Actual Winner||Electoral votes||Actual Trump||Actual Biden||Actual Winner||Projected Poll Variance||Actual Winner’s % Margin||Variance from Projection||Variance In favor of…||Actual Trump||Actual Biden||% of Votes Counted|
|District of Columbia||3||0||3||Biden||77.0%||87.9%||10.9%||Biden||5.3%||93.2%||99%|
My favorite saying of the year if not the decade is that “People are not much interested in what they see, read or hear. They are most interested in how they feel about things.”
If you feel that the polls “always get it wrong” then you may not be very much swayed by the data I have presented. Perhaps you have your own data to support how you feel. That would be good. Personally, I don’t like to feel one way or the other without some data to back me up.
Columns #1-5 show the state, their electoral votes and which candidate won the election in that state.
Column #6 seems like an important column. It shows how much each candidate was expected to win each state according to an average of 163 different polls. Of course, not every one of those poll sources conducted a poll in every state. As an example, Trump was projected to win Alabama by 13.9%.
Column #7 tells you how things really turned out. Again by way of example Trump actually won in Alabama by 25.6%. By the way, these actual results numbers come from the Associated Press and CNN. Don’t like those sources? I can’t imagine any other source would have different numbers on the actual results which of course in public information.
The actual results are from states reporting 99% of their vote total…except for a handful of states where two weeks after the election they have reported 95-98% of their total votes.
Thank goodness that New York was not a battleground state. Two weeks after the election they have counted just 76% of their votes!
I think column #8 provides important information. This shows the variance between the poll predicted results and the actual results. As an example, Trump was expected to win Alabama by 13.9% of the votes and actually won by 25.6%. He did better than expected by 11.7%.
Did one candidate do better than expected compared to the very often? That’s where the data from column #9 comes in. Yes, one candidate did do better than the polls predicted! Trump did better than the polls predicted compared to Biden a somewhat incredible 44 times out of 51! I find that amazing.
Sometimes doing “better” didn’t matter much. Again as an example, the Massachusetts polls predicted a Biden win by 36.5%. Biden “only” won by 33.0% in the Bay State. That difference doesn’t do Trump much good. However, it is important to note that Trump did better than expected in 44 out of 51 contests.
I have heard political pundits explain the poll predictions in two ways. Some have said that republican voters might not have gone to the polls because Trump was expected to lose. Then I have heard other pundits say Biden supporters didn’t go to the polls as much because Biden was expected to win and their vote wasn’t needed.
I can only say this about that. I feel bad for folks who take every bit of “news” and declare that news was only shared to hurt their candidate. I hear that line of thinking a lot….which makes me dislike politics even more.
Finally, columns 11, 12 and 13 show the actual results by state and the percentage of the vote counted as of November 16, 2020. That’s 14 days after the election was held.
I hope you found what I have shared interesting. Do my conclusions that the polls were excellent in predicting the winner by state and that one candidate did much better than the polls predicted square with your perceptions?
I had no idea how much time recording data from 163 polling organizations was going to take. Some of these groups did polls in nearly every state multiple…..I mean multiple times during the campaign. I tracked literally thousands of pieces of data. I don’t plan to do that again!
One final reminder. Doing this study was about presidential polls and not politics. I don’t do politics in my newsletters. I can tell you this. It doesn’t really matter that you love one candidate and hate the other. You can rest assured there are about 50% of the people in America who feel exactly the opposite of how you’re thinking about things. Not everybody gets that point. Most conservatives hang out with conservatives and most liberals hang out with liberals and never the two shall meet.
San Clemente, California
Oh my gosh! I was having a bad day!
Why me? Why did they choose me to receive the 2020 presidential election results the day BEFORE the election?
Patted down, checked for weapons, put in the back seat of a Ohio State Patrol police car and taken away…
How to get a free iPhone 12!!!!
The Financial Plan of a Lifetime – 2020
South Korea….one of my most unusual last minute experiencesDo you have random thoughts?
Belarus….in the news!
When you go to bed at night do you pray for rampant inflation?
Psst. I know who’s going to win the 2020 presidential election
India – Part 3….read this and you’ll feel like YOU went to India!
India…..Part 2…..The First Half of the Adventure to the Most Unusual Country We Have Ever Visited
India – Part 1….the planning stage
The Aftermath of the Minneapolis Protests and Riots
Household budgeting….you know you need it