Jarate Chop!

Community & Sustainability

| Comments

After ~14 months running Smurfy Fortress I thought I'd share some of our experience in terms of building a lasting TF2 community. This post is also, in part, prompted by the many recurring threads on reddit regarding premium or donator perks.

Firstly, I'll say that I don't particularly care what other server operators do or don't do but our own players have suggested various perks at various times so this post seeks to share our thoughts around this.

As I said back in November 2012 the goal with Smurfy Fortress was to host servers that I want to play on. Broadly, this means fair & fun games, free of abuse. Above all else, game play is paramount. This isn't only a personal preference but I think it's the key to the sustainability of any gaming community.

Why is community important?

Well, firstly, multiplayer games aren't much fun on your own.

More importantly, anyone can put up a server for $20/month. It's neither expensive nor technically difficult. The only thing that differentiates one server from another is the players… i.e. the community.

Ok, if community is so important, how do you build a community?

One approach is simply to bribe, or buy, players. You would've seen this approach in many forms. e.g. join our server at

This may be okay to kickstart a community but long term this approach is fundamentally flawed as you're essentially paying players to play on your server. Unless you have unlimited funds this is obviously unsustainable, and when you eventually run out of funds your players will migrate to the next "community" offering perks, hats, and whatever else.

Let's think about this another way.

We know that players like to play with their friends. So maybe the easiest way to grow a community is to ensure that the players that we want as part of our community consistently play on our servers. Then, over time, their friends will become regulars, then their friends, and so on. Growing a community this way is essentially free (monetarily, it certainly takes time & effort).

So then the question becomes: what makes players stay on a server?

This seems simple enough. Players leave if they're not having fun.

This brings us back to donator perks. Such perks might be good in terms of fundraising, and are almost certainly fun for the donors, but are terrible for everyone else in your community. If you assume your players aren't completely stupid, the vast majority of players who aren't donors will never play again on your server once they realise they're being raped by players who paid for advantage. Long term, your "community" consists only of donors and new players who don't know any better.

An optimisation of this model is to entice new players to join your servers through giveaways as mentioned earlier. This may be sustainable since the donator perks drive revenue which allows the server operator to continue offering enticements to new players. But this isn't the sort of server I care to operate.

What we've chosen to do with Smurfy is focus on game play.

We started with basic things such as auto-assigned teams (to combat team stacking), and class limits (no one wants to play with 5 snipers, though 12 snipers is hilarious!).

As the community grew abuse became an issue. So we added more admins, and also created the Elder Smurfs. Essentially we give long time regulars some recognition (a server greeting) and the power (!votekick and !votemute) to help us keep the servers clean… for everyone's benefit. Players have asked why Elders don't get other perks, e.g. the ability to have Unusual effects added to their hats. This comes back to our optimisation goal - we want to build a lasting community of friendly players. i.e. we don't want players to play and rack up hours because it will get them perks (any server can offer perks), we want players to play because they enjoy the games & community.

With a scalable way to deal with abuse the main problem we have now is team balance. Although the random assignment we used previously was fair in the sense that no one was allowed to bias the teams, the resulting teams weren't always skill balanced and the game would be one-sided.

A side note – some players are happy with, or think that they are happy with, an easy win. Afterall, a win is a win. But my contention is that what players really want is a challenging win. Consider this – if you play tennis with your best mate and you beat him in straight sets 6-0, 6-0, week after week, how long do you think he'll keep playing? He'll feel terrible, and you don't get any sense of achievement, it's like beating a 5 year old. One of the challenges in making games fun for everyone is remembering that winning is fun but not if it is a walkover.

To resolve this we recently introduced SmurfPowder which ranks players based on whether their teams win or lose, and attempts to pick skill balanced teams based on these rankings. The system needs time to establish rankings so we're unsure how much it will help but we will certainly share our findings when we can.

For the business nerds I see parallels with marketing & product strategy. If your product (games in this case) is shit you'll have to spend a bunch on sales & marketing to sustain your business (community, in our case). OTOH, if your product is good, word of mouth will go far. I think both models are valid from a business viability perspective but I know which one I prefer.

Also, on differentiation – anyone can set up a server and implement any of the technical features mentioned, including the TrueSkill based rankings. Long term, the only differentiator for gaming communities is, unsurprisingly, the community itself.