I recently ran a game theory analysis of following on Twitter, specifically focused on following with the intent of getting someone to follow in return; originally, I intended it for myself, but I think the findings are worth talking about (especially for those who try to use follows and ‘follow backs’ to build your audience on Twitter). The extended form analysis is down below:I get that this chart might be a little hard to read at first. Basically, Player 1’s choices are highlighted in light blue, and Player 2’s choices are highlighted in green. Off to the right side, you’ll see the results of their combined choices, phrased as simple equations. The V1 and V2 indicate the value of a follow from Player 1 and Player 2, respectively, and there are two equations for each result (one from Player 1’s perspective and one from Player 2’s).

To explain this a little better, let’s run through the game once. It starts when Player 2 follows player 1 (the ‘Original follow’ on the chart), and then Player 1 must decide whether to follow back or ignore this action. Let’s say they choose to follow back; Player 2 then has the option to continue following them, or to unfollow. If they choose to continue following Player 1, then they both obtain the top result (V2-V1 for Player 1, and V1-V2 for Player 2).

What this means is that Player 1 gets the value of Player 2’s follow, minus the value of their own follow (and vice versa for Player 2). These values are relative based on each player’s perspective; Player 1 might value their own follow highly and Player 2’s highly as well, whereas Player 2 might not value either.

This explains a lot of behavior when it comes to following and following back. For example, it’s well known that those who follow many others have a high rate of returning follows. Based on this analysis, that likely occurs because the more people someone is following, the less they value their own follow (perhaps because they rarely check their feed or it’s already cluttered anyway). Hence, why they follow back with more predictably.

It also explains why unfollowing someone whom you received a follow back from almost always results in a loss of that follower. The equation for that scenario is the second down on the chart (0-V1, V1), which shows that you’ve forced Player 1 into a losing situation; the only way they don’t lose the game is if V1 is zero, which is almost never true. That’s why you see some people who create throwaway accounts just to follow, and why they generally aren’t successful (since hardly anyone values what’s essentially an empty follow).

Scenario three (the ‘celebrity’ result), in which Player 1 doesn’t follow back and retains the original follow anyway, typically occurs if Player 2 followed Player 1 for some reason outside of the game (for example, they saw them on TV or heard them on the radio).

Finally, the bottom result—scenario four—is a failed game, in which neither player gains anything. This should logically be the most common result. Game theory explains it thusly: Player 1’s decision is the most operative, as Player 2’s second choice is the same regardless (continue following or unfollow). Therefore, Player 1 can look at the end result and determine their best play.

If they follow back, they’ll receive either V2-V1 or 0-V1. If they ignore, they’ll receive either V2 or 0. Therefore, the worst result of ignoring is a zero sum game (no one wins, but Player 1 can’t lose). This has to be compared to the best and worst result of following back. In order for Player 1 to choose the follow back option, V2-V1 must be greater than 0, and 0-V1 must be extremely low (preferably approaching 0).

In other words, someone you follow must value their own follow little and value yours highly, or else they won’t follow back. And yet, the more people you’re following the less value your follow seems to carry. Twitter users aren’t stupid; they can tell that someone following a hundred thousand other users isn’t likely to see their tweet.

It’s a paradox of sorts: the fastest, most reliable way to gain followers is to follow others, and yet the more you do this the less likely you are to receive follow backs. That’s why many authors seem to approach a ‘critical mass,’ where they no longer receive enough follow backs to grow at a reasonable rate (or they seek out other valueless followers, which is why they start to look like spam accounts).

Author:ValerieThomas

1. Hi Valerie. I have too many other things to look at with my book, but thanks for the thought

1. Sorry, I’m not sure I understand… If you’re signed up for email updates, you might have received an automated email from my blog, but if you don’t have time to read a new post that’s totally fine, haha.

1. I am not sure I understand either. I am new to this whole scenario

2. Super awesome analysis; I wonder how muting would be taken into account with this? My general tactic is to follow back and then immediately mute so that I retain the follower but can still participate on twitter without a cluttered feed.

This is surely ‘losing’ in some sense from their perspective as they have a ton of empty numbers but is also losing on my end as people following a whole ton of other users aren’t likely interacting with their feeds/seeing my tweets anyway.

There is some sense in which it’s not losing though because both of our numbers are increasing, if only nominally.

1. Thanks! Your point about muting is excellent; I believe there are a good deal of confounding factors that this analysis doesn’t take into account. What I find interesting about muting is that it could be considered as either a third choice in step one (follow and mute) or as a second action for Player 1 (after following Player 2 back). If it’s assumed to be a third option, then the results would be V2, V1-V2 or 0, V1 (since a mute effectively changes a follow to have no cost for the person muting, but Player 2 doesn’t know that they’ve been muted).

If it’s taken as a second action then all four results (the two from the original game and the two just mentioned above) are possible, which creates a sort of prisoner’s dilemma; do you trust the other player/user not to mute you (and therefore render their follow nearly worthless)? Of course, then we get into the nominal value added by another number for someone’s follower count, just as you mentioned. I hadn’t considered muting, but come to think of it it’s actually incredibly important to understanding the game as a whole.

3. This is so on point, I just have to comment just to say that. You articulated so many of the things that get under my skin as a reader…and a few of those bad habits that I have to actively weed out of my drafts during the editing process. Next time someone wants my advice on writing fiction, I’ll be sure to refer them to this post too.