Originally Posted by
Ahovv
Except all major online polls are voluntary, resulting in more skewing that you normally would see. They are posted publicly for people to vote on and can become distorted based on external efforts. Magazines polling their subscriber base would clearly have its own bias, further amplified by the audience having a full list beforehand.
The most accurate polls are more random in nature. For example, polls conducted by Pew Research do not post publicly and then ask an audience to vote. Sure, they are technically skewed because not everyone has a landline. I'm not suggesting they are flawless. But polls by reputable companies are meant to have as little external influence as possible and cover a wide spectrum, not limited to one particular subscriber.
To argue all polls are equally accurate is just nonsense. Honestly it's surprising anyone can believe that.
- - - Updated - - -
1) The drudge report would have its own bias even if people who normally don't view drudge report voted.
2) The outcome of a drudge report-only poll can be expected to match the screenshot I posted where Sanders was only around 8%.
3) It is virtually impossible for Sanders to come from 8% to 29% when the 8% figure was several hundred thousand votes in. So, on top of the Drudge Report initial bias, there was then obvious external influence putting Sanders far above what can be expected for the website.
The main point I'm trying to get across here is that nobody should reach the conclusion "Wow, 29% of regular Drudge Report viewers voted for Sanders!" On top of that, any non-random poll such as this one cannot be expected to match real-world results in any way. That's all.