# Why the CNN Poll Cannot be Trusted

I know I blog for a mathematics website, but reading the post debate news this morning I feel the need to blog about the CNN debate poll. In the article, “Post-debate poll: Hillary Clinton takes round one,” the article notes, “Hillary Clinton was deemed the winner of Monday night’s debate by 62% of voters who tuned in to watch, while just 27% said they thought Donald Trump had the better night, according to a CNN/ORC Poll of voters who watched the debate.”

That seems pretty damning for Trump. However if you read down a bit you will note, “Although the survey suggested debate watchers were more apt to describe themselves as Democrats than the overall pool of voters, …” It is also noted in the video attached with this article, “that this polls skews Democratic.” The video states that 41% surveyed were Democratic and 26% surveyed were Republican.

This is where I take issue with this poll on a mathematical basis. Whatever your political leanings are, you can clearly see a Democratic bias in this poll. Yet many news stations have reported on it as being, “the most scientific.” Sadly most of the people typing these articles probably have very little idea as to what constitutes a scientific poll and how to run a statistical poll we can actually trust.

To do this CNN would need to actually sample a representative portion of the population for Republicans, Democrats, and Independents. Quickly looking online at Gallup data (which we can discuss another time as to whether you want to believe that or not) you see, “42% identify as independents, 29% as Democrats, 26% as Republicans,” in 2015.

Ideally if CNN wanted to do this right they would had similar representation of people being surveyed. They could easily set this up with a computer system to insure fairness.

Other polls claim that Trump won, such as many of the “snap polls,” like the one on the Time website. How are these you ask? Well even in their disclaimer you can see that this is not the best statistical method of determining who won the debate. Some people worry one person will vote multiple times, etc. I don’t see that as a worry as long as you set it up for one vote from one IP address. I took the poll this morning and couldn’t vote again.

What makes these polls questionable is that these polls are only polling people that go to that website and not necessarily the demographic the poll really wants to focus on. They could also be skewed if there is more than one registered voter per IP address and they can only vote once per IP address.

So it is clear both the CNN and the Time poll are not scientific. So which poll do you trust? Is there a poll out there you can trust? Unfortunately probably not, regardless of your political ideology.

So what is one to do? Well, I argue that you should think for yourself and vote on your beliefs regardless of what your friends, significant other, or what any poll tells you.