Following the two party conventions and a week of almost unprecedented gaffes by Donald Trump, the fundamentals of the 2016 presidential race are fairly clear: #Hillary Clinton is ahead, probably by a lot.

The Real Clear Politics polling average as of Sunday lists 11 #polls, and Clinton leads all 11 by anywhere from 1 percent to 15 percent, and by an average of 7 percent. Following several months of Clinton leading most polls, and Trump getting a brief bump around the time of his convention, Clinton is now at what is arguably her strongest point in the race.

But of course, some Trump supporters won’t accept these numbers at face value. It must be media bias. Or craven pollsters who fear Trump and are out to destroy him. And there’s always the time-honored trick of finding the most favorable-to-Trump outlier poll, declaring it a “shock poll,” and proclaiming the race closer than it actually is. But some have gone even further to “prove” the polls can’t be trusted.

Unskewed, again

Some have argued that the polls can’t be trusted because pollsters want Hillary to win, which betrays a fundamental misunderstanding of how polling works- pollsters earn success and professional standing by being right, not by favoring certain candidates. But then there’s the argument that Trump is only losing because the “sampling” is off- that the polls are over counting Democrats or undercounting Republicans.

That was the reasoning behind UnskewedPolls.com, an effort in the waning days of the 2012 election that took as its credo that if you don’t like how the numbers look for your candidate, just make up new ones. You’d think the original #unskewed Polls effort being totally proven wrong would discourage this practice, but apparently not. A site called Longroom.com is doing just that.

Twitter follows can’t vote

But there’s another argument against trusting polls that’s even more ridiculous: It’s that Trump is winning because he has more Twitter followers and Facebook likes than his opponent.

This was pushed by an obscure conservative blog called The Truth Division and later picked up by the more high-profile Gateway Pundit: According to Truth Division- “There’s no doubt about it — the liberal media’s polls are rigged from the bottom up. How do we know this? Simple — social media. Trump’s following on the major social media networks absolutely blow Clinton out of the water.” What follows is stats showing that Trump has nearly twice as many Facebook likes and 2 million more Twitter followers than his opponent- therefore, the polls are unreliable.

Where to start with this nonsense? Twitter followers and Facebook likes aren’t votes, nor are they scientific measures of voters’ intentions (which polls are.) Trump has 10.6 million Twitter followers compared to Clinton’s 8.1 million?

One, elections aren’t determined by Twitter follower counts. Two, there’s no way of knowing whether those followers are actual voters, and not people who are underaged, in foreign countries, bots or not otherwise registered. Three, following someone on Twitter doesn’t mean you like them- a lot of people who oppose Trump follow his tweets anyway. And four, those follower counts are dwarfed by how many people actually vote. Nearly 66 million people voted for President Obama in the 2012 election, and over 60 million for Mitt Romney. The majority of people in America don’t use Twitter.

Trump focuses on 140, Clinton focuses on 270

And finally, considering that Hillary Clinton joined Twitter four years later than Trump and has tweeted much less often (7,000 tweets versus neither 33,000) it’s impressive that her follower count is so close to his. It may have something to do with Trump’s campaign being staffed by people who were last relevant in pre-internet days.

In conclusion, that’s not how these things work. Presidential elections are decided by voting and the electoral college- not by how many Twitter follows the candidates earn. If it were the latter, then Kim Kardashian would be president.