Community management

Community management

27/07/2021
indiedev, psychology

Eventually I want to make an MMO. I’ve already sort of described it here but that idea will probably change significantly over the years. What won’t change is the fact that successfully running an MMO is both a massive technical and social undertaking.

I’m constantly working on improving the technical side of things by just programming a lot, but the social part of it is harder to practice because you actively need a community of people to manage, and until now I didn’t have such a thing. With SNKRX’s release though I got the opportunity to “manage” a “community” of tens of thousands of people for 1 month or so, which was a really good amount of experience to gain.

I put the words “manage” and “community” in quotes because the community of a single player game isn’t really comparable to the community of an MMO, in that it doesn’t really need to be that actively managed. I don’t really have to think about things like handing out bans, or worrying about cheaters/hackers, or word filters in the global chat, or anything like that. But still, experience is experience.

As things stand I have no way to tell when I’ll get an opportunity like this again, as I don’t know if people will play SNKRX again in great numbers after I release more updates, and I also have no way to tell if future games I release will enjoy similar levels of popularity. Although I suspect that they eventually will because my skills will keep increasing, but that might take many many years.

So because of this unknown I thought it would be useful to write this post for future reference. Mainly, I want to get a snapshot of my thoughts regarding community management now so that the me from 5-10 years in the future can have a really solid sense of what this experience was like, where my head was at, where I was right/wrong and why, and so on.

Also, this will be a fairly unfiltered post. I still prefer making these kinds of posts public because I always like when people share their honest thoughts on things, but I’d prefer if it went unnoticed. If you stumble upon this post and want to share it with people, please do so carefully and avoid larger scale social media (twitter, reddit, etc). Thanks!

Power #

One of the main ideas that seems obviously correct to me now about game community management is that you want to avoid leaking power. I already partly went over this here, but I thought it’d be nice to expand on it more.

This idea comes from Curtis Yarvin, who applies it primarily to politics, but because it’s an insight about collective human behavior it can also easily be applied to game communitiies. He explains this concept in multiple posts, but the one that does it best is probably this one. I won’t go over his explanation in detail here since the post is pretty self-contained, so you should read it if you want to understand what I’m talking about.

His analysis of how power leaks seems fundamentally correct to me. It’s one of those things where once I read it it just made so many things retroactively make sense and I had that sort of eureka moment. And the application of this to other situations, like with games, also follows naturally.

If you’re a game developer with a popular game you can think of yourself as a power source. The power you have is that you can change the video game. Everyone wants this power because everyone secretly wants to be a game developer. People are really drawn to you with attempts to have their voice heard and have their will implemented in the game, it’s just a natural thing that happens without people even really noticing.

What this means is that if you’re a developer who leaks this power by too casually implementing people’s will, you’re essentially polluting the lake. You’re letting people wear the ring, and when someone wears the ring, the ring changes them. Power changes people because the only thing power wants is more power.

So your game’s players, when they get a taste of this power, will want more of it. Every conversation people have about the game will now be focused on you, because you’ve essentially trained people, through your power leakage, that if they make suggestions you will listen to them.

And what this means is that, just like in science where ideas aren’t selected for truth anymore, your game’s player’s feedback will also not be selected for truth. The feedback will be selected for its ability to justify the usage of power, which roughly translates to feedback that convinces you to implement it.

This is a crucial point that is very important to understand. The type of feedback that will dominate in a community that has been corrupted by power is fundamentally different from the feedback of a pure community. Feedback made to convince you to implement it may often times be similar to pure feedback, but it comes from a different place, and that place is generally worse.

Some developers seem to intuitively understand this and try to avoid it:

But others…

Path of Exile #

In my previous post I mentioned reddit and how this idea of power leakage can affect a game’s subreddit. A few people who read that section immediately connected it to Path of Exile:

However, when I wrote it I actually wasn’t thinking of PoE, I was thinking of Valve, interestingly enough. One of the things I noticed about Artifact and Underlords is that Valve seemed to listen way too much to reddit and that they sort of fumbled without direction a little too much. Like there was no too strong vision behind the games and they sort of implemented anything that seemed good and was upvoted enough.

This isn’t necessarily a bad strategy if you’re trying to make the game survive in the first place, but it communicates a lot of weakness and might paradoxically drive people out of the game more. Not even to mention all the power leakage problems I already went over.

In any case, what I wrote in that post definitely applies more to Path of Exile, even if I didn’t consciously mean it when writing it. Now, I’m going to say a lot of things about PoE and GGG, but before that I’d like to make clear that along with Valve they’re probably my favorite developers.

I think the ability Chris has to be honest about his game and to communicate well with his players is really inspiring. You get the real sense that they’re just indie developers who really care about their game and are just trying to make things work, which they are. In fact, part of me writing these very honest posts about my game is inspired by how Chris does it.

Similarly, even though I’m going to also talk about PoE’s subreddit a lot, I think what they’ve managed to achieve with it has always been really good. To the degree that my previous game succeeded, it did so because I was allowed to post about it on that subreddit, and Chris and Mark personally intervened so that my post would stay up after the mods removed it for a while.

So you know, nothing that I’m going to say here comes from a place of ill will or bad intention. It just comes from this very detached analytical perspective of trying to understand what’s actually going on.

With all that said, one of the main things that motivated me to write this post is Path of Exile’s latest league, Expedition. A sort of unusual thing happened with it, which is that GGG decided to not listen to a big part of their community and to change the game in a way that is more in line with their vision of what it should be. You can see the announcement stream here:

Needless to say, the reaction by the community has been extremely mixed and there’s sort of a mini civil war happening, with some people siding with GGG while others thinking they’re ruining the game. Regardless of which side is ultimately right, one of the interesting things to watch has been reddit’s reaction. In case you didn’t know, GGG has always been very active in the subreddit. They’re constantly making posts in it, constantly appearing in comments, responding to players, etc. And they’ve been doing this for like almost 10 years at this point.

This means that, if my idea of power leakage is correct, you’d expect the subreddit to have been thoroughly corrupted by power and for most of its discussion to be directed towards GGG rather than towards the game. And that’s roughly what happens. As an example, let’s look at the top threads of the past week:

And here we can do some basic analysis and just classify each thread based on what it’s about:

#Focus
1meta thread about reddit
2GGG thread
3direct request to GGG
4comic referencing streamer’s negative reactions to the patch
5direct request to GGG
6GGG focused, sort of a small guilt trip on the community by Chris on a stream
7feedback, talking about the bad state of the game, makes direct mention to Chris
8feedback
9GGG focused due to a (falsely) perceived balancing mistake on their end
10GGG thread
11directly addressed to GGG
12streamer thread
13GGG focused
14game focused
15GGG focused through a streamer

Out of the 15 threads, 1 of them (#14) is focused on the game, while 10 of them are directly focused on GGG. 2 others are focused on streamers, 1 is focused on reddit itself, and another (#8) is the one that’s closest to pure feedback.

Thread #8 is the one that’s most interesting to me because it perfectly exemplifies how easy it is to confuse corrupted feedback with pure feedback. Remember that I mentioned that corrupted feedback often times looks like pure feedback, but comes from a worse place? Thread #8’s title says:

All of the Expedition splinters should have been AUTOPICKUP, the same as metamorph organs. PERIOD.

If this title dropped the “PERIOD.” part, it would have been worded perfectly. But “PERIOD” implies something. Period, or else…? There’s always this hint of blackmailing going on whenever people addicted to power aren’t getting their way, and so you see this behavior on whatever topics GGG has decided to stand their ground on and generally not change, despite continued efforts by the community to get them to change it, such as having to click lots of things, trade, and so on.

The nuance here is that corrupted feedback always comes from a place of trying to justify the usage of power (convince developers to change the game), whereas pure feedback comes from a more neutral and detached perspective. And more importantly, like all feedback, corrupted feedback can be either positive or negative. The positive form will take the form of praising power (praising the developer), but the negative form will generally take the form of blackmailing, because that’s all you can do when you’re addicted to but denied power: you can cope with some kind of bargain, like “if you don’t change this I’ll stop playing!”, which PoE’s subreddit also gets a lot of:

As a developer you can’t look at this and blame reddit. What’s happening here is what would happen anywhere, because everyone involved is a human being, and human beings respond similarly to incentives. PoE’s population has been thoroughly corrupted by power for years, and this is the result. A population that is completely addicted to getting their way, being enabled by developers who are reaping the benefits of doing so.

None of this is nefarous or wrong on either side of the equation, it’s just what naturally happens when people interact with one another. And PoE is just the best example I can use right now, but I’ve seen this exact same dynamic play out with countless games before.

We can look at a couple of other examples that aren’t taken to this extreme. For instance, Dota 2’s subreddit:

Here there’s only 1 thread directly aimed at Valve due to the fact that they take a much less visible role in the community. Valve isn’t completely neutral though, they will often do quick patches based on reddit threads and they also run multiple tournaments for the game so they have somewhat of an actively role still. But in general their interaction is much more limited and so you get way better results in terms of the level of corruption in their community.

There could be an argument here that I’m doing a sort of unfair comparison by comparing a moment where PoE is unpopular and Dota is neutral. And so if I were looking at Dota’s subreddit right after Valve implemented an unpopular change, it would have the same level of threads directed at Valve as there were directed at PoE. And this argument would be roughly correct, but there are some nuances.

First, the difference between both subreddits remains even when PoE’s is neutral. What happens when PoE is not going through any particular problem, like at the end of a league, is that still, most threads are focused on GGG in one way or another, generally with requests for improvements to the game. And when the game is doing really well you also get tons of threads praising GGG and talking about how they’re amazing. The point is that there are rarely moments where the focus isn’t GGG, and I think the idea of power leakage cleanly explains why this happens.

Second, one thing that’s noticeable when Valve disappoints Dota players is that more often than not the response is one of acceptance, rather than one of denial/anger/bargain. You can look at threads like this for an example:

And both the thread creator and the comments are much more lighthearted and accepting of the situation (the thread title is sarcastic). This is because Valve never promised an update when people thought one would come, so as one of the commenters mentioned, even the bargaining stage of the disappointment takes a different form, not one of blackmailing because these players intuitively understand that they have no power at all, but one of pleading, as in “please give us any information at all”.

For a subreddit that is completely 100% ignored by the developers, we can look at PUBG’s:

And here we have the perfect subreddit, it’s literally all about the game, and it’s all visual too. One thing I don’t like about reddit is how some moderators really don’t want their subreddits to be oriented towards images/gifs/videos and want things to be text only. However, video games are mostly visual and to me it makes no sense that a subreddit about a game would be mostly text.

In any case, when comparing all 3 subreddits in terms of active users and also comparing the number of people playing the game, PoE’s is clearly ahead of the other 2, and PUBG is clearly last. So there are clear benefits to managing your game’s subreddit like GGG has done so far. It is very obvious that GGG being as active as they are with reddit, but also with streamers, has massively contributed to the success of their game. So it’s easy for me to point at flaws without having appreciation for the positives of it. But I feel like the positives are obvious enough that they’re not really worth going over too much, and also most of them have hidden drawbacks that are hard to notice unless you’re looking at the problem from an odd/different angle.

Promises #

For instance, in the first video I posted in this thread, the CS:GO developer mentions that one of the reasons Valve doesn’t like promising things to players is because it allows them to be more flexible and responsive to player demands. This makes sense and is logical, but there’s a deeper reason why promises shouldn’t be made, and I only came to realize this in the process of writing this post.

One of the things a promise does is that it lets people plan around you. In PoE’s GDC talk one of things Chris mentions is that the game really started becoming more popular once they switched to a predictable release schedule:

And as Chris mentions, there are a lot of benefits to this. But one drawback is that making people plan around you further makes things about you. This is pretty obvious, right? Dota players aren’t as disappointed at Valve when there’s no patch because there’s no promise by Valve that every 4 months there will be a patch. Consequently, this makes things less about Valve, which corrupts people less as Valve isn’t leaking as much power, which makes their response to disappointment less aggressive and dramatic.

So a promise is fundamentally another form of power leakage, which is unexpected and something that I didn’t think about when starting writing this article but that totally makes sense. As a developer when you make a promise you’re essentially forming a contract with people about your game, and if you fail people will hold you accountable. But how will they do it exactly? That’s right, they can’t actually hold you accountable because they have no power.

The only power they have, which you’ve now given them, is to be upset that the patch was bad or that it wasn’t delivered in time. And how great of a power is that? There’s nothing they can actually do, so all you’re doing is further training them to ask you for changes and further frustrating them when those changes don’t happen.

So this is a power leak, but it’s sort of a fake power leak. With the normal one you leak power because you implement people’s suggestions and that lets them feel powerful, but with this one you promise them something, which is the ability for them to hold you accountable via the promise that the patch will come every 3 months and that it’s going to be good, but then if you fail them the promise is broken, which leads to even more frustration because they actually can’t do anything about it.

Another drawback is that when there is a promise, the patch coming in the first place is taken as a given, and so it’s taken for granted. When there’s no promise and the patch comes it’s always a pleasant surprise and you feel like a kid who got a nice gift from your grandma or something, but more importantly, if the patch doesn’t come then it’s not as devastating, both because players haven’t learned to expect it, but also because they’re not being given fake power, and so they aren’t as frustrated.

Finally, the main drawback of promises is that they’re essentially a cope from the developer. This is a point that the CS:GO video touches on lightly when he says that when promises are made people think about the future of the game rather than how it is now. And I’m definitely guilty of this with SNKRX.

Once I decided that I would update the game often I both told people what I would do (update it every week) and then I did that. And once people noticed that, they started playing the game more because the promise did its job, which is to add reliability and structure to an otherwise unpredictable situation.

But this is a cope because it’s a much stronger position, but also a riskier one, to do the updates without actually promising any future ones. It’s stronger because you’re not leaking power and whoever decides to play your game is playing it because of its state now, and not because it’s going to be better in the future, but it’s riskier because with this structure not existing, people who would have otherwise played the game might decide not to.

For instance, Northernlion started playing SNKRX right after I said that I would update it every week. I don’t know if both things are related, but if they are then this would have been a massive loss to the game if I hadn’t made the promise in the first place. So it clearly works, and the opposite clearly means catastrophic failure.

But this is the kind of risky thing where, you know, now that I don’t have to worry about financials, it can either fail catastrophically or succeed massively. The catastrophic failure means that the game doesn’t get that many more players and dies, but I’m personally not affected. But the massive success? That means the game is fundamentally good, because people aren’t playing it due to what will come in the future, but it also means the community is uncorrupted by power, and such communities are generally going to be much better at keeping the game alive in the long run than corrupted ones.

There’s also another more subtle problem with the structure that promises add, which is that because they’re a cope, they’re essentially a form of fake success in a way. Ideally you want people to keep playing the game because it is fun and addicting, and not because of the promises, or because of streamers, or any other set of things that are not directly related to the game’s quality.

Essentially, if the game’s success is built on those structures it means that the problems the game has haven’t been solved, they’ve just been postponed. For most developers this is exactly what they want because they’re running companies and employing people, and so giving the game life beyond what it naturally would have is always desirable. But for me? I’m not bound to anyone like that, I have no responsibilities to anyone but myself, and so I can take the way riskier path of having to actually solve the game’s retention at the most fundamental level possible and risk the game dying rather than having to keep doing updates at a predictable schedule and coping while worrying about player numbers.

I focused somewhat on GGG and Valve for this discussion, and I think they’re good examples of this dichotomy. Valve definitely takes the riskier approach despite being a massive company, and it shows in their failures with Underlords and Artifact but also in their massive successes. GGG doesn’t, and that has led them to a lot of stability and success in a very hard section of the market, but I think that their decisions have leaked way too much power and I personally haven’t seen a community that has been corrupted like this ever be purified by the developers, mostly because no one ever reaches the conclusions I’ve reached in this article in regards to the correct ways to act to not leak power, but also because even if they do reach the right conclusions, it’s very hard to mass reprogram people like that. What generally happens when this is attempted is that people just leave with a bad taste in their mouths and never come back.

When I look at this I’m also looking at it from the perspective of how I’m going to handle my future MMO. In an MMO it would be absolutely disastrous if most discussions were centered on me or on people working with me, instead of the game itself or the game’s community. An MMO is fundamentally a community experience and the discussions of this experience should be focused on the community, and to the degree that they’re focused on anything else then I’ve failed at what I set out to do.

But more importantly, like Chris says, I want my MMO to be a game that will be played forever. And so I have to make sure that the community is a community that will last forever, because an MMO is as much the game as it is its community. And communities corrupted by power have a very predictable lifetime, and it’s definitely not forever.

This analysis so far mostly focused on power because that’s what I’ve been thinking about the most lately, but there are other traits to be focused on when you’re trying to engineer a long lasting community. For instance, one unexpected one is that communities that are “edgy” last longer. The simple reason for this is that edgier people are lower in politeness, which is one dimension of the trait agreeableness, and people lower in politeness tend to be contrarians.

Contrarian communities are essentially self-propelling and self-renewing communities because any behavior that becomes the norm will be rebelled against, which will renew the community and prevent it from becoming stagnant. This is a very important property that is extremely visible on sites like 4chan, for instance, that no one talks about because I guess no one is paying attention. But it’s as clear as day to me that this is true.

This of course has limits, because at some point the edginess and contrarianness makes it so that the community is too undesirable for newcomers, but like everything it’s a balance. You can go for a community that will last a long time, or you can go for a community that appeals to everyone. Trying to do both will generally not work, and we can see this as even the most popular games ever eventually die out, even though the developers keep pushing updates of higher and higher quality.

You could look at this and explain it as just “well I guess after 10 years of playing League of Legends people finally got tired”, but it’s really an excuse. After 15 years of using 4chan people didn’t get tired, so 4chan must be doing something right whereas all these other communities are doing something wrong. In any case, this was sort of a tangent and I could go on about it forever, so lets get back on track.

Influencers #

The same power leakage dynamic applies to streamers and youtubers, although with them it’s a slightly more complicated situation. Do I want to go over influencers in more detail? It’s a longer explanation with a few more conditionals, but I guess it’s worth going over the short version of it.

Status #

Essentially, I consider that there are two groups of influencers. In my head I call them “normal guys” and “actual influencers”. Normal guys are dudes just like you and me who happen to be streamers or youtubers. They are people who act like a normal human being, probably mostly keep to themselves, and are basically just doing their best and entertaining their audience. I guess all streamers fit this category to varying degrees, but it’s very easy to identify people who are mostly in it rather than not.

I would consider people like MOONMOON, forsen, LIRIK and Mathil1 to be a few examples of people in this group that I happen to watch. Basically as a developer this group is fine, perfect even. You can expect them to act like a reasonable and intelligent person would act in their situation, which is that if they like your game and it works for their stream they will play it, and if they don’t like it or it doesn’t work then they won’t. Very simple, very easy, not really a mystery at all.

The other group though, “actual influencers”, are people who I would consider to be similar to the redditors in PoE’s subreddit. They have been corrupted by power and they’re addicted to it. But this isn’t as bad as it seems, because in our society everyone is like this to some extent. They just happen to be in a field that rewards this particular orientation more than most other fields.

What this means is that they fundamentally are addicted to status. These are people who want to matter and who want to be relevant at all costs. I would consider people like xQc, Asmongold and HasanAbi to be good examples of people in this group. However, these are just examples, and I would probably say the majority of streamers fit this group to varying degrees. Essentially, one of the problems streamers and youtubers have, which game developers do not generally have, is that it’s very hard for their goals to not be focused on numbers.

If you’re in this field you have to pay attention to the number of viewers and subscribers you get, because if you let it fall too much you run the real risk of it just collapsing really fast since there are network effects at play. When you spend years having to worry about numbers like this it is very hard to detach yourself from them, and so if you’re not careful what this does is that it makes you a slave to popularity.

In this context, popularity is the same as status which is the same as power. So at the end of the day they are addicted to power. But they are addicted to power through their audience, because their audience is what allows them to have this power, which is an interesting dynamic.

This problem also affects game developers sometimes, and I would say this is the same dynamic that affects GGG when it comes to their 3 month league promise. It’s not surprising that Chris mentions Kripparian as the source for the idea of having a rigid schedule, because it’s the kind of thing that makes perfect sense for streamers. These are people who need, for their professional safety, to make the number their primary concern, so they will do everything to at least make sure that the number doesn’t fall too much.

Because they have to care about the number not going down, it’s also hard for them to have goals that are not tied to the number going up, because who doesn’t want the number to go up, right? And so this dynamic played out over the years corrupts them in this very ugly way and makes them become very status hungry.

So from the perspective of a developer, the way to deal with people like this, essentially drug addicts, is the same way you deal with the polluted lake or with the corrupted subreddit, which is to never let them taste power in the first place. You have no control over most aspects of their lives, of course, but you do have control over their interaction with you and your game. And so that interaction should be primarily one of denial.

If your game is popular these people will want to play it because they’re addicted to popular things and they don’t want to miss out on the new hot popular game that came out. And if they really like it and play it for longer than just 1 day then, like everyone else, they’ll start looking to you to fix problems they have with the game, but because they’re addicted to power and status, they’ll also look at you as a sort of status source.

A recent example of this is this interaction of Asmongold with a viewer, thread here. A lot of people focused on how he’s mean to the viewer but I don’t really have a problem with that. What happened here is that Asmongold was talking about how streamers should be able to reserve their names in an MMO so that people don’t steal it and pretend to be them. Then a bunch of viewers pointed out the obvious, which is that that kind of special treatment is unfair, and then the argument followed from there. You can see another streamer defending those kinds of ideas here.

And even in the past you also had similar situations, specifically with GGG and PoE, that started the same debate:

Now, from a pragmatic perspective, it’s totally reasonable for developers to do these kinds of things for streamers. As Soda and Asmongold mention, these bigger streamers and youtubers actually provide a lot of value (although it depends on the game and it also depends on the streamer), so it makes perfect sense for them to be treated in a special manner above other players. But as you might guess, I disagree with that strongly.

The problem is that most of these streamers asking for special treatment are not asking for it for practical purposes, they’re doing it for status purposes. Deep down they want these special treatments because they want to flex/increase their status, and so they’ll naturally gravitate towards situations where this can happen.

They’re operating in an environment that has been thoroughly corrupted by power, and so in the same way that PoE’s redditor’s feedback is not selected for truth and only selected to justify the usage of power, streamer’s requests will not be selected for truth and will similarly only be selected to justify the usage of power.

This is, again, a very important but subtle distinction to make. Even though the requests these influencers make will often times seem reasonable and pure, just like often times the corrupted redditor feedback seems reasonable and pure, it fundamentally comes from a different place, and that place is generally worse.

So once the streamers are given power, they’ll want power again and again. If you comply they will play your game, and if they enjoy it they might even make your game extremely popular. But if at any point after that you decide to refuse them their drug they will get angry and they might even decide to retaliate through their audience.

Asmongold has been doing that with Blizzard lately and it’s absolutely vicious. In case you didn’t know, he basically spends a few hours of his stream going over Blizzard’s recent lawsuit and just shitting on them in every possible way imaginable. He’s started playing FF14 and has been trying very hard to drive as many people from WoW as possible. And he also interviewed Chris as a way to clearly do the contrast of “this is what a good developer who listens to their community looks like”. It’s just attacks from every direction imaginable.

I personally think that Blizzard deserves it, I really don’t think there’s a company that I actively dislike more than them. But from the perspective of a developer this is a perfect example of what a status-led individual in his position might do if they decide they really don’t like you.

What this all means is that when it comes to influencers, if the only types that existed were the normal guys then everything would be fine. As a developer you’d be able to interact with them to your heart’s content and all would be well. But because there are a lot of status hungry ones out there, once you agree to interact with a normal one, you signal to the hungry ones that you’re a status dispenser.

And if you’re a status dispenser but you decide you don’t want to interact with the status hungry influencer because, frankly, he kind of smells and his pupils are dilated, then you’ve potentially just bought a fight with someone for no good reason. In his head (and in his audience’s head) it’s not the fact that you refused him, it’s the fact that you refused him but you didn’t refuse the normal guy. This makes his status inherently lower, so now he doesn’t like you and decides to trash your game for everyone to see because you dared refuse him his drug.

And that, in itself, will get him his new highs. You see, he doesn’t need your game or your status. He just needs his highs. And he can create it out of thin air if he wants to, he just needs a really good excuse. You just happened to be it this time. So, the lesson here is that you want to be a machine, but you want to be a machine of game development, and not one of status distribution.

Note that pretty much everything I’ve said in this post happens subconsciously. No one is going around consciously thinking about these things. These are the basest and most fundamental drives in people and I only get to think about them like this because I’m extremely smart & handsome, but most people just go through life without really thinking about things from these perspectives too much.

Business risk #

There are also more practical reasons to consider more careful interactions with influencers, but these are slightly less interesting. One is that just from a business risk standpoint, if influencers are having a large effect on your game’s popularity then that means the game isn’t yours anymore. This can be both a good thing and a bad thing.

For instance, because ARPGs are kinda niche (they aren’t actually niche but they also are never going to reach Fortnite levels of popularity I don’t think), streaming ARPGs is a risky thing. If you’re an ARPG streamer you’re kinda having to hop from game to game, and all of these games are kinda bad in their own ways, and so it’s very hard to “make it” in that space. But if you have a company like GGG, who has a good game that has updates every 3 months, providing some stability to the space, then it’s much easier for you to make a living.

And so what happens is that there are quite a lot of streamers who are primarily PoE streamers. These people will generally not act like the status hungry streamers might and just casually trash your game, because they understand that your game’s success is the same as their success. In this situation you’re winning because you get a baseline of streamers playing your game (since they depend on it), and you also get a baseline of players because people watching the streamer play the game are generally also playing the game.

So it’s very tempting to cater to these influencers as much as you can because if you’re both happy with each other’s work then you both win. But again, there’s a lot of nuance to it.

For instance, Asmongold has been a WoW streamer for as long as possible. He was in the same situation as the PoE streamers, which is that whenever he tried to switch to another game his viewcount would drop dramatically. So he would always have to go back to WoW. But now that he has found another competing MMO that gets him a higher viewcount, he has easily switched from WoW and decided to trash the game on his way out. Again, Blizzard deserves it, but I don’t really see why the same dynamic couldn’t play out in the future for PoE.

When you initially cater to these influencers who are already sort of frustrated that they have to rely on your game to make a living, and then for whatever reason you have to stop catering to them (i.e. with lots of nerfs) and now they have run interference for you even though they don’t really want to, this creates tension. And if conditions are right, like another patch that is very controversial is released, eventually this tension might just lead these people to quit the game in this very drastic and aggressive way.

On the other hand, if you don’t cater to them at all, like for instance, you don’t release updates every 3 months and thus you add no stability to the situation and make it harder for people to plan around you, then you lose out on a lot of very obvious potential good, but you also give less control of the game to them. And this also has all the other benefits I already mentioned in the promises section.

Cancellings #

Another reason to consider more careful interactions with influencers is that we live in age where people overall are hungry for status. And this manifests itself a lot in the form of status destruction, or “cancel culture”. Influencers live in an environment where status management is important, therefore they’ll want to distance themselves from you if you’re deemed by the population at large to be someone that deserves to be taken from your high status position to a low status one.

I happen to be someone that could very easily fit this mold, as I like going on tangents of all sorts that may be wrong/impolite, and so from my perspective it just makes a lot of sense to not rely at all on influencers since I know that it’s all fake, and that it’s going to go away the moment people just read the kinds of things I’m writing.

At the same time, what matters to people cancelling others is that they’re taking someone from high status to low status. But if you’re low status in the first place then you’re not an attractive target. So one strategy is to just try to keep your status as low as possible, and that requires very few interactions with influencers, since interactions with influencers by default are high status events.

Both of these reasons, “business risk” and “cancellings” are weaker manifestations of the more core issue, which is status/power. Because this is such a core issue for human beings it’s something that’s relevant everywhere, and analyzing everything from this perspective is bound to get some good results. I feel like my analysis of community interactions through this lens has led me on a pretty different and weird path, but I feel like it’s the correct one for what I’m trying to do.

Guidelines #

With all of this said, I can expand on what my guidelines regarding community interactions should be for me. These are guidelines that are valid for games I make now, but that also should work for the eventual MMO.

Promises #

No promises of any kind. No plans for future updates, no roadmaps, no describing what’s going to be in the next update, no dates, no anything. This is a very radical and risky stance but it’s clear to me now that it’s the correct way to act. This means people won’t be thinking about the game’s future, which means that feedback will be less corrupted. It also means that the focus on me will be even lower, as there’s no expectation of anything from the developer.

I will also try this with SNKRX going forward. The only promise I’ve already made is that the game will be rewritten from scratch and that I’ll keep supporting it for at least 1 year. These are still promises but they’re not as bad as “there will be one update every week” or “the rewrite will be done in 1 month” (lol, what was I thinking).

One obvious thing here is updates have to come and the faster they come the better. If people are just left without updates for too long then that will have a predictable effect which is that the game will be played less, and the secondary effect is that all the supporting structures for the game, like streamers and youtubers, will also play it less as they notice that their analytics for the game aren’t as good anymore.

So even though there should be no promises, updates should obviously happen as quickly as they reasonably can.

Small places #

Whenever people have bugs or feedback and this happens in a “small place”, I can interact freely without worrying about leaking power. What I mean by a “small place” is a place that isn’t distinctively a community. A good example of this are Steam forums.

No one considers the Steam forums for a game to be a community. It’s just a place people go to to figure out how to solve a problem with the game or to report bugs or give feedback. There is no recognizable class of “PoE steam forums users” like there are “PoE subreddit users”.

In practice what this means is that if I implement people’s will from there, the effect of leaking power will be much much lower than if the same happened on reddit or other places that feel bigger. Which means that as a general rule I don’t have to worry about it.

E-mail #

Because e-mails are primarily private there’s also no problems with interacting with people there. So far most e-mails people send me are pretty standard stuff, but some people have asked questions that if I knew were going to be posted publicly I would respond to differently, things like more concrete plans for the game’s future and stuff like that.

If I ever notice that e-mails I’m responding to start getting posted on social media then it means I should filter myself more there, but until then it’s fine to reply as honestly as possible.

Discord #

Interacting with people on Discord via private messages is fine, interacting directly in a public server isn’t. When you interact with people too much publicly in a Discord server it just naturally becomes this power leak situation where a few people in the Discord server that you enjoy interacting with start getting favorable treatment (i.e. their opinions on the game are heard more), and this signals to everyone else that if they make you enjoy interacting with them their opinion will also get heard more.

Over time this has the same effect as the subreddit situation, where the entire dynamic of the community becomes centered on you, and this very obviously quickly makes it an undesirable place for you to be in. It essentially becomes a circlejerk. Unlike reddit though, it degenerates into a circlejerk that isn’t even related to the video game, so it’s not a case that the Discord server will have corrupted feedback, it will have no feedback at all because most of the time people won’t even be talking about the game anymore.

Twitter #

Twitter is the most status focused social media site that exists, so it should generally just be avoided. I will never make any efforts to grow my Twitter account to any degree and I’ll mostly use it as a personal gamedev account for the most part.

If any game I make gets big enough I should just create an account for that game just to post news about it there, since the service is genuinely useful for that use case. For everything else though it’s pointless.

It’s also the place where people get cancelled the most so just from a defensive perspective it’s just a good idea to not be too visible there.

Meta posts #

Meta posts are posts like this one. They’re posts that talk about issues that if shared too widely will have a negative effect on the subject that the post is talking about. For instance, if in 10 years my MMO is popular and someone decides to post this post to the game’s subreddit because some sections of it are relevant to whatever is happening at the time, there’s enough stuff here that might cause problems, like mentioning Moldbug by name or slightly insulting some streamers.

However, there are a few things that work well for posts like this. The first is that most people won’t read it because it’s too long and in a way too abstract. Most people aren’t willing to follow along with these kinds of posts and they just get bored before going too far. This doesn’t stop people from sensationalizing things and taking quotes out of context, but it helps to decrease the post’s overall transmissibility.

The second thing is that I can always hide behind the fact that the post was written in the past and that I have changed my mind about whatever issue, even if I haven’t. One example of this is a post of mine that made its way to Northernlion’s sub:

And this opinion of “maybe he’s changed his mind” and “you’re just trying to create unnecessary drama” will generally be present and sometimes be the prominent one whenever these issues happen. Coupled with the fact that I’m not actively trying to be relevant or to increase my status, this makes me a very unnatractive target to the kinds of people who are in the mindset of creating threads like that one.

Cancellings #

In general whenever issues like this happen the response should just be nothing. Like the corrupted redditors or like some of the streamers, people who are looking to cancel you are addicted to power. They want to matter, they want to change the world, and their means of changing the world involves punishing you if they don’t like something you’ve said. Curtis has the correct analysis again:

If power is not an enemy but a predatory animal, like a bear, we see the asymmetry of the relationship. Theoretically, a moose can kill a bear. This is just one way of getting the bear off its case—usually not the easiest. And a moose would never eat a bear. Of course, you are a lot more defenseless than a moose—but nor are you hungry for bear.

If you are securing yourself against an enemy, you need to prevent your enemy from being able to harm you—as you do your best to harm him back. If you are securing yourself against a predator, you only need to prevent the predator from choosing to try to harm you. The bear doesn’t want to eat you. He just wants to eat.

This may be as easy as persuading him to select another moose. We all know the parable of the bear and the running shoes—any physical-security specialist will explain it to you. Even a consistently policed commitment to nonconfrontation gets you quite some distance beyond your “fedposting” friends.

Power is every bit as dangerous as you think. It is nowhere near as intelligent and malicious as you. It is a hungry animal, not an evil genius. Ultimately there is never a formula for dealing with evil geniuses—but hungry animals can be quite consistent.

Human predators are sadists. It makes them feel good to hurt you. They sublimate this evil emotion into an ideology which convinces them that, by hurting you, they are making the world a better place. This is only part of the political formula which tells them that, by supporting the regime, they are making the world a better place.

For by hurting you, they are supporting the regime. And by hurting you, they gain honor, status and power. Who doesn’t want that? Don’t you want it too? While this seems awful when we look at it this simply, they cannot look at it this simply.

These sad people are just slaves—slaves to Moloch. You cannot fight back against them. And ultimately, you have to realize—they want exactly the same thing as you. They want to make the world a better place. The only problem is that they grievously understand the nature of the task—and their understanding of it includes hurting you. So: don’t hate them, don’t fear them, just work hard to stay out of their way.

And in the case of game development, working hard to stay out of their way roughly translates to simply not trying to be relevant. Having that taken care of, it also helps to position yourself such that such cancellings have minimal effect, and I already went over that here.

There’s an argument to be made that gaming is the one entertainment field that has the biggest opening for a developer to really just not worry about this stuff at all. For instance, if you look at what happened with Factorio, threads here and here, and developer’s comments here, and you look at the immediate effect of that whole ordeal:

It looks like attempts to cancel developers who are being reasonable generally fail, as Factorio is not the only game this has happened to, it happens all the time. But this is a kind of naive way of looking at the issue. It’s obvious that the majority of the population thinks all of these cancelling attempts are stupid, but if the general population winning popularity contests mattered then things would have gotten better a long time ago.

As a general rule these issues run much deeper than you assume and you’re not saving Western civilization by confronting these people, you’re just marking yourself as a target for their endless attacks. And for what? For a few thousand reviews of people who won’t play the game either way because they’re just using it as a prop to win some insignificant cultural skirmish? That looks like a bad deal to me, so just pure avoidance and non-confrontation seems like the correct long term way to act.

Influencers #

I already went over how the policy with influencers should be one of denial. There should be nearly no interactions nor special treatment given to them. The only exception is if the interactions are private.

If you can ensure that interacting with some influencer is going to be a completely private thing, and that the influencer won’t mention to others that he interacted with you, then it means that the interaction isn’t happening for status purposes, and it also means that other status hungry influencers won’t be put off by it, since they won’t know it happened. In that case there’s no downsides to it, so it’s fine.

Developers #

Developers have a similar dynamic as influencers, although because gamers as a class generally don’t care about who makes their games (they only care if the game is good or not), it means that the field hasn’t been as corrupted by status as streaming has, at least from its inherent dynamics.

But on the other hand there are way more developers than there are streamers who are completely politicized, and politics in our current age is the same as trying to matter, to be relevant, and to change the world, and so a great number of devs are also addicted to power and status, just from a slightly different angle than the streamers.

This means that as a general rule you don’t want to interact with them purely for defensive purposes, as if they find out you don’t agree with most of their political opinions there’s a higher chance they’ll take you to Twitter to try to do you in.

But because it’s generally not the case that other devs would want to interact with you for status distribution, you can very easily just vet people first and only interact with the ones that seem like reasonable people, as this won’t cause any negative responses on the status hungry political ones.

Reddit #

I also already went over reddit. The perfect subreddit is one where 0 threads are aimed at you and all threads are talking about the game, and you can achieve this by just never interacting with it. This doesn’t mean that you should 100% ignore it. For instance, I read SNKRX’s subreddit every day as well as all its comments, it’s your job to know what people are saying about your game and to act on that knowledge.

But it’s also important to not signal to reddit that you’re reading reddit. You shouldn’t just implement every suggestion reddit has upvoted to the top every week because that will teach redditors that you’re listening and then they will become corrupted.

The main drawback of this very detached approach is that the subreddit will be less popular than if you were more active, but you know, that’s a price I’m willing to pay. I’d rather that my game became popular because it’s good rather than because it has a well managed subreddit, or because I’m friendly with streamers, or because I’m doing some really good marketing on twitter, or whatever.

Bans #

While I didn’t have to do any banning for SNKRX, having a well thought out idea of how to handle extreme edge cases of behavior makes sense if I want to eventually make an MMO.

As a general rule I want to do less work, so I want to ban as few people as possible. This fits with my personality very well from an enjoyment perspective too, as I generally prefer less moderated places to more moderated ones. I think a good guideline to follow is to give people tools filter their experience, and only act and ban people on really extreme cases.

All bans should have multiple warnings first, and they should have increasing sentences based on repeat offences, such that if it’s your third time getting banned you should be banned for a longer period than if it’s your first time.

Bad words #

There are certain forbidden words that people don’t like, but I personally assign no real negative moral judgement to them. From a pragmatic perspective if someone is just going around repeating those words all the time then they should get warned multiple times and eventually banned if they don’t stop. But casual uses of it, especially if it’s not in a global chat, should be tolerated.

I should also make sure that the game has a chat filter that’s on by default so that people aren’t hit by a fat n-word the moment they log on to the server, but by and large, unless I’m forced to by a higher force I don’t want to have to police people’s speech too much.

Spam #

If someone is spamming the chat then there are automatic solutions I can create to prevent them from doing that, like limiting the number of messages sent per minute. But should they find a way to bypass that and to generally ruin other people’s experience via spamming, then they should probably get banned.

Consistent harassment #

People should have tools to block and mute others that are annoying them. I think a feature I haven’t seen in many games is one of IP blocking someone, such that even if they create other characters they will still remain blocked. Even then, some people are on dynamic IPs so this could fail.

If there should ever be a case where someone has done their best to block someone in all ways imaginable and they’re truly not engaging with the person, but the person continues to try to harass them across multiple accounts in the game, then that person should get warned and eventually banned. This should be a deeper ban than just a single account obviously, since it’s clearly harassing behavior that goes beyond what’s reasonable.

Arguments, insults #

As a general rule I don’t want to get involved in people’s arguments. So someone insulting someone else’s mom or stuff like that will be treated as just internet arguments that have nothing to do with me. If someone is insulting others using any of the special bad words then they will get warned and eventually banned if they don’t stop. Otherwise it’s all fair game.

Stream snipers, reserving names #

You can guess my position on this. On top of just being a bad decision for all the reasons mentioned before, it’s actually quite a lot of work to consistently and correctly ban “stream snipers”. And as I just said, I want to do the least amount of work possible on this front.

It’s also worth mentioning that this is another situation where the argument from streamers seems reasonable. If a streamer really wants to tryhard the game and do well, and he can’t because he’s getting stream sniped, and there’s something you could realistically do about it to help, it’s a complete and perfect win-win. They get to showcase your game in a better state, their viewers get better content, your game gets more players, and all you have to do is ban a few people.

But again, this comes from a place that isn’t good. And this is because even though this particular streamer might need help with stream snipers for a legitimate reason, another streamer might not. Another streamer who has a big audience might just decide that he wants people banned in these really awkward situations that are very dubious and not really applicable, and now you have to give this guy a big “no” because you don’t want to ban people unfairly.

And then it comes back to that status thing, where he might just decide he doesn’t like you, and on top of stopping playing your game, he just completely trashes it on the way out and does as much damage as possible. And you know, the chances of this happening are low, but they’re not really as low as you’d think.

And then you also have to apply this to all other possible things streamers might ask for, like the example mentioned before of reserving names. Let’s say you decide to let streamers reserve their names ahead of everyone else so the names don’t get sniped. What’s the cutoff for a streamer being able to do this?

The most reasonable one would be someone who’s partnered on Twitch, but that’s literally thousands of people, and you definitely don’t want to let thousands of people reserve their names ahead of everyone else, that’s totally unfair. So now you’re having to decide who’s popular enough to get the privilege of reserving their names, and now you’re fully involved in streamer status battles and you’ve made some streamers your enemies for no good reason at all.

Anything you decide to do involving special treatment to streamers will lead to disaster one way or another. There’s no logical reason to do it. Most developers do it because, for the most part, most developers want their games to be more popular because they need more security, they’ve hired tons of people, there are families depending on the game doing well, and so on. I can’t really blame them.

But as for myself, I’ve already made enough money out of SNKRX to last me a very long time, so I’m not going to keep making games because I need more money or more status. I’ll do it because I want to make good games. And there’s nothing, absolutely nothing, about spending my time banning stream snipers that leads to a better game. So I won’t do it.

Cheaters and hackers #

I have no idea how to deal with these. This is actually probably the thing I’m most scared of because it just seems to be an endless black hole of effort. In a way dealing with cheaters and hackers is similar to balancing the game’s difficulty, it’s not something you’ll ever get 100% right, you just have to get it right enough that most people aren’t affected by the problem.

One idea I have is making an MMO that has classes that cheaters/hackers would want to use. So instead of people making bots to auto-farm things, you have a class that specifically can build robots to auto-farm anything. This way the game would be balanced from the start with robots in mind and it would diminish the incentive for these kinds of cheats to be made in the first place.

Similarly, to decrease gold selling you could make it so that people can buy items from the cash shop for real money, but they can sell those items to other players for in-game money. This way someone who wants to spend real life money to get a gold advantage in-game can do so, but people who can’t spend money but can spend time (generally kids) also get access to items from the cash shop. This relies on the cash shop having no P2W items, but that’s obvious enough.

Other than that I feel like dealing with this issue will require hiring additional programmers if the game gets popular enough so it’s just a pure hit with no real benefits.

Interviews #

Interviews and podcasts to streamers, youtubers and journalists are a pretty unconditional no. As I’ve already explained, I generally have nothing against individuals in these professions, but the dynamics of it make it so that giving an interview to one person, however reasonable they are, signals me incorrectly to their peers and that potentially leads to problems.

Interviews with players, such as AMAs, Q&As, etc are undesirable but I’m not 100% against them. This is simply because any kind of mass event that brings a lot of attention to me at the same time is something that raises my status, and anytime my status is raised I become a more attractive target to people addicted to status. I would much prefer to answer questions from players on a 1-to-1 basis privately.

Interviews with other developers are more conditional. Ideally anything that is aimed at other developers, so essentially gamedev talk, is fine. This is because sharing information is good and being available to other developers as a source of information is also good. However, it also depends on the developer. If the developer is status hungry and it seems like he will not be able to handle my more contentious ideas then I probably want to avoid talking to that person.

Dev streams #

Development streams like the ones Jonathan Blow does are good. I’m not really going to do any of them in the near future but I’m not against it. The main thing is that it has to be clearly focused on game development. If the people watching it are doing it because they’re players of your game then that’s a bad sign, since it can easily lead to power leaks, as in streams people get pretty mentally close to the streamer, perhaps more so than in any other situation.

Talks #

Presentations in general are undesirable but I’m also not 100% against them. The main reason is that like with AMAs and Q&As, they tend to raise your status, and more status = bad. I’d much rather write long posts like this one that most people won’t read than give a talk.

However, sharing information with other developers is good, so that’s why I’m not 100% against them. Obviously there are certain events like GDC for instance which I would never give a talk to, primarily because I don’t like how the talks are hidden behind a paywall, but also because they have all these weird rules about what you can talk about, like the talk can’t be too negative, and you can’t mention other games by name, and so on. I understand why people from San Francisco would come up with these rules, but you know, it’s just not for me.

Creativity #

This entire post is somewhat focused on me not wanting to draw too much attention to myself. And as I’ve explained this is because I want to keep my status low. But I see how someone might get the wrong idea, that I don’t want attention on myself because I don’t want to deal with all the drama or the insults or whatever that comes with it if I fall on the wrong side of having attention on myself.

However, I am someone who actively enjoys stuff like that. I like conflict, I like arguing, I like debates, I like being the center of attention. There’s nothing that I like more than everyone very visibly telling me I’m wrong and then it turns out I’m right. I would feel extremely at home being the center of some big drama and I would be able to defend myself well. But I’ve also come to realize that this is somewhat of a flaw in myself, and I’ve been actively trying to reign in these instincts that I have to engage in conflict and trying really hard to not give in to them.

At the same time, I’ve also come to realize that these instincts are what allow me to be creative in the way that I am, and likely what allowed me to be successfully creative. Creativity is about useful idea generation. And from a personality perspective, to be creative you have to be high in a trait called openness, as this will allow you to create useful ideas faster than other people.

But it’s not enough to generate useful ideas, you have to generate useful ideas that haven’t been generated by others. And to do that you need to be willing to go where other people haven’t gone before. Part of that will be controlled by a trait called politeness, and if you’re low in it this will correspond to your ability to be out doing your own thing without much concern for what others will think of it, some real sigma male stuff. More seriously, this idea is also captured by this series of tweets.

So when I want to reduce the amount of attention on myself I’m both rejecting this inherent instinct I have for conflict and attention because I’ve noticed that it’s often times counterproductive, but I’m also protecting my ability to be usefully creative, as I need to be able to go for ideas and on tangents that might be wrong to eventually get to something that is uniquely right. And I wouldn’t be able to do that if I had to worry about my general social standing as a developer.

And as if it was a message from God himself, the moment I truly internalized all these thoughts and that I decided that I would just make games, regardless of how popular they got, the first game I end up releasing gets extremely popular primarily because the idea is good. So it’s just very hard for me to not conclude everything I’ve concluded so far.

In summary, as a developer it’s best if you’re invisible. By default gamers don’t care about who makes their games, they just want to play good ones. To the degree that you have to be visible you have failed, and you should work to fix things so that you can go back to not mattering. This will both lead to better and uncorrupted communities, and to better and uniquely right ideas. And that will lead to better games.