Daniel W. Drezner is a professor of international politics at the Fletcher School of Law and Diplomacy at Tufts University and a regular contributor to PostEverything.

A troll doll named Morrissa. (Nathaniel Grann/The Washington Post)

Longtime readers of The Washington Post are probably aware that there are some nasty people on the Internet. The 2016 presidential campaign has brought out some even nastier elements on both ends of the ideological spectrum in the United States. Click here or here if you’re interested in the fratricidal flame wars occurring on the left; click here or here if you’re interested in the emergence of the alt-right as online acolytes of one Donald J. Trump. The point is, there are a lot of ideologically extreme trolls dancing around on the interwebs right now.

As someone who has written about trolls on occasion, these kinds of groups fascinate me. Are there high-quality and low-quality trolls? Do trolls change their mind (click here for one interesting case)? Mostly, as someone who has read John Stuart Mill’s “On Liberty,” I was curious if his 19th century assertion about the intellectual vitality of extreme sects remained true in this century:

There are many reasons, doubtless, why doctrines which are the badge of a sect retain more of their vitality than those common to all recognized sects, and why more pains are taken by teachers to keep their meaning alive; but one reason certainly is, that the peculiar doctrines are more questioned, and have to be oftener defended against open gainsayers. Both teachers and learners go to sleep at their post, as soon as there is no enemy in the field.

Over the past 10 days, and rather unintentionally, I have been at the center of a natural experiment on Twitter to test Mill’s proposition on both the alt right and far left. A little more than a week ago, this tweet prompted some anger on the left parts of the ideological spectrum:

And then, Monday, this tweet sent the alt-right into a few paroxysms of fury:

So this is an interesting opportunity to compare and contrast. How are far-left trolls different from alt-right trolls? How are they similar?

Like any good social scientist, it’s worth pointing out a few differences between the two experiences. Most of the left trolls came around as a result of responses by either Alternet’s Adam Johnson or the Intercept’s Glenn Greenwald; there was no similar high-profile tweet from the alt-right crowd. Furthermore, my contretemps with the left was about a macro-historical question; my interactions with the right were about … well, I’ll get to that in a moment. But those biases in the inputs probably had an effect on the quality of the trolling.

In both cases, I was accused of many unsavory qualities and spent much of the next 24 hours parrying angry tweets directed at me. In both instances, the experience of being in the eye of a troll storm felt pretty similar. The far left objected to any form of “tone policing”; the alt right felt similarly.

After both experiences, I can draw the following tentative and thoroughly inductive findings:

1) Although anti-Semitism came from both sides, it was the principal rhetorical gambit of the alt-right. A few members of the anti-American left went right to Israel to explain my position, but it was a decided minority. On the alt right, it took at most one or two interactions to get to tweets about sending me to the ovens. Truthfully, most of the alt-righters didn’t have much more of a point than that. This could substantiate the argument made by some alt-right proponents that their anti-Semitism is intended as a provocation rather than as a genuine belief. Or it could mean that the alt-right is just populated by a lot of crude anti-Semites.

2) The alt-righters really like their anonymity. Almost every angry leftist who engaged with me did so with a Twitter account that contained their real name. Almost all of the alt-righters did not use their real name; the one who did use a real name subsequently deleted all of their tweets directed at me. It’s almost as if they’re afraid of what would happen if their tweets under their real identities were revealed or something.

3) The far-left really likes to get in the last word. Have you ever known someone who, no matter what, absolutely needs to have the last word? Not because they have a substantive point to make, but just because? That was how my far-left trolls engaged on Twitter. With some of them I would just tweet back something anodyne to see if they would respond. And they always did.

4) Both the far-left and the alt-right dislike the United States intensely, but for very different reasons. For the far-left, U.S. hegemony is the source of most bad things in the world and none of the good things, full stop. Any data that substantiates the notion that the world has improved while the United States has been a superpower is dismissed as ginned up by a pay-for-play economist or something. For the alt-right, it’s the decreasing fraction of Americans who are white Christian men.

5) This has to be a worse experience for women. To be honest, most of the trolling from both sides was pretty puerile (although the interactions on the left did produce this Ryan Cooper column that I do not agree with but is worth a look). I could devote most of my energy to watching a baseball game or cooking dinner and parry off the vitriol with the occasional break. Then again, I had the Twitter equivalent of fresh legs. I’m fortunate that I’m not the target of this kind of trolling on a regular basis. Unlike, you know, an entire gender.

So, my conclusions: A) never tweet; B) despite his best efforts, Trump is not making racism great again.