GMod Stories: Trouble in Terrorist Town

I keenly remember the first time I played Trouble in Terrorist Town. I was idly flicking through GMod servers when I stumbled across a game where arguing seemed to be taking precedence over bullets. At least for a while. Eventually bullets came into play. I tried asking what was going on, but ended up getting killed in a dark corner. I was truly an Innocent. Returning after some serious Googling I was a bit better prepared for a gamemode that felt years ahead of its time: where players were give roles to enact as murderous Traitors, Innocent victims, and the Detectives trying to sort out the good from the bad. I'm not the only one who was taken in by TTT. Over the years it's been one of the most enduring Garry's Mod modes, if not to play then certainly to watch. I got in touch with TTT's creator Stefan (aka Bad King Urgrain) to talk about how it came together, how it sprung from another mod, why people are still playing his duplicitous deathmatch, and what Gabe Newell thinks about it.

I’m old so I played the original Half-Life 2 mod version when it was new. I had the whole “let’s put thrusters on everything”-phase, but didn’t get into building/posing long term. The game as we know it now I didn’t actually own until I decided I wanted to try making a gamemode.

I was already involved in the “make games on the Source engine” scene via a full-fledged mod: Zombie Master. We made that as a (very) small team from 2005 until early 2009. When I released the final version largely by myself in January ‘09 I figured I was done with the whole game development thing (free advice: never try to make an RTS game on Source). There was an active community on the Zombie Master forums even in 2009, with a bunch of regulars that often played together. Around the summer of that year they started playing an entirely different game within ZM, where instead of doing anything with zombies, the players were playing against each other. It was called Serial Killer, and it worked something like this: a server admin would privately message one or more players that they were “the killers”, their job was to kill all other players, who of course didn’t know who was a “killer” and who wasn’t. That probably sounds pretty familiar. All players would run some console commands to turn off some HUD elements, like the kill log, and generally use the honour system to apply the game’s rules, because of course Zombie Master was not built to support any of this. I played a couple of matches and figured it could be a lot more pleasant if all this manual stuff was not necessary. So I grabbed GMod and started making a gamemode. The goal was just something that we could play this game in without working around an entirely different game.

Initially TTT was just what I called my implementation of this Serial Killer game. I was using the Counter-Strike: Source models because they were conveniently available, that’s why TTT is set in this world of Terrorists. If I had known TTT would gain some level of fame I might not have made that choice, because explaining to non-gamers who don’t know Counter-Strike that you made a game in which everyone’s a terrorist is uh, pretty awkward. A lot of the names/descriptions in TTT are jokes about the absurdity of the whole setup (including the game’s title), but it leans pretty heavily on CS to make that all work. Anyway, at the start I only intended this gamemode to be a thing for this forum gang. I was working on it every day given that it was the tail end of my summer holiday period before starting my last year of university. I was putting up new releases every day, sometimes multiple times. The speed of developing in GMod’s Lua environment was a real contrast compared to Source modding. Being able to extract a new zip on a server and have everyone playing your new stuff right away gives a lot of motivation. The very first version already had the round structure and basic Traitor/Innocent stuff. It was built to run on Zombie Master maps, and the weapons were basically straight clones of ZM weapons except with a CS:S look. I added the Traitor equipment store soon after. That made room to add a bunch of cool Traitor weapons. The Detective role came a while after I believe.

I played the game a lot while I was actively working on it, and that gave me some clear ideas of what I wanted players to do and what I wanted to avoid. Then I tried to make the game’s design encourage the “good” behaviour that generated cool or funny experiences for everyone, and discourage the bad stuff. This was never related to the setting, purely on what would be the most fun for everyone involved. A lot of the game is fundamentally about information. Who knows what, and what do players do to get more data. Traitors fundamentally know almost everything, especially if they get certain items like the Radar. That’s their biggest edge and they need to exploit to win. Innocent players know basically nothing, and if a Traitor plays a perfect round it stays that way right until each Innocent’s death. In practice, a Traitor is going to “leak” information: they will leave corpses behind with clues, they will act suspicious, perhaps even get spotted. At the same time, they can manipulate others into drawing the wrong conclusions. The Detective was introduced as a way to focus the Innocent “team” on what they should be doing: gathering information. Their extra abilities and purchasable weapons and items are largely based around that. For the Innocent players it’s important to defend the Detectives and help them do their job. That gives them an extra task while also introducing more conflict, because obviously a Traitor would also love to follow a Detective around to “protect” them. It was a challenge to find gameplay mechanics for Detectives that on one hand let them gather meaningful data, while not ruining the game for Traitors if they make a single mistake. This is one of the places where a lot of iteration happened, for example on the DNA mechanics. It started out pretty complex, with the Detective having to gather samples from corpses, dropped weapons, and players, and then allowing them to match them up to deduce more clues. In a typical hectic round this was way too much and no one bothered. Clicking around in a little menu to match things up is not going to feel like a good way to spend your time when you could get stabbed in the back any moment. Over time the mechanic changed into its current form where finding DNA on e.g. a corpse gives you a direct lead to the killer’s location, but Traitors have ways of disposing of corpses or faking their current location. Plus it’s always possible and encouraged to talk your way out of things.

I think the hook as an Innocent is that you’re in this very paranoid position where you know nothing and there’s people who know everything and want to kill you. And it could be anyone. This creates a lot of weird incentives, such as how on one hand you want to stick to a group, on the other hand you don’t because they’re all dangerous. At the same time it might be a while before you see some action, so they might start messing around… or maybe they’re not messing around because they’re a Traitor. This is all very fertile ground for fun(ny) interactions to happen, and most of the game’s design is about encouraging those things. There’s positive and negative sides to that. TTT is a very social game, and if you have a good round depends greatly on the other people you’re playing with. Play it with friends and you’re almost guaranteed a good time, play it with random assholes and it’s a lot more shaky. This has always been a challenge in designing the game because on one hand I want to encourage chaos, because it can be funny as hell. At the same time, if a random person just shoots everyone they see every round it stops being fun very quickly. I’ve tried to strike a balance there with the Karma system to let the game do some moderation of its own, but most public servers have very strict rules going on that I don’t particularly like. It cuts off some interesting gameplay if you just have an admin slay everyone who shoots the wrong person. At the same time, servers have to deal with trolls and bad players. It’s a problem I have not been able to solve, and it’s a big reason why I don’t necessarily recommend people who know nothing about the game to just go play on a random server. You will not always find a good environment, and even if you do, they might be playing what is essentially a different game entirely. This is also a major reason why I’ve never considered making a standalone paid version of TTT. I can’t sell you a certain experience unless I sell it to you in a pack that includes a set of fun people to play it with. In the end, TTT was developed as a game for a bunch of (internet) friends to play together, and that’s still the best way to play it.

I became pretty good at weighing every idea that I had myself or was suggested by players on how it could be abused or negatively affect the overall experience. People could get very inventive with ways in which they could figure out if you’re a Traitor. For example, finding a Traitor weapon somewhere and then throwing it at people to see if they automatically picked it up. If not, they’d already have a Traitor weapon equipped. I had to change how weapon pickups worked to give Traitors plausible deniability there.

Information It got a huge boost in popularity when Youtube LPers found it, which was actually pretty late (2012/2013 probably). Before that it was really purely a GMod thing, after that it found a much bigger audience (and probably generated a whole bunch of GMod sales!). I think the main reason it works well for them is they tend to have a group of people to play with, which as I mentioned before is the best way to play. Even if they then end up joining public servers and gratuitously griefing, that’s Youtube for you.

Information I’ve never seen The Thing (oops) but it’s definitely a very social game. Being good at the game mostly comes down to being good at lying to people. The funny thing is I never even owned a headset while I was working on the game. You sometimes hear about developers making the type of game they wanted to play, but I suppose in this case I was making a game for other people to enjoy. Adding a cool new feature in an afternoon and then playing a match the same day where people are having fun with it is a great feeling.

Information I always like the maps that do something neat with the TTT-specific stuff to make things like traps or Traitor-testers with interesting gameplay around them. The ones I remember are all going to be very old. One TTT map I remember was (funnily enough) ttt_thething, because it was an early map with a Traitor-testing mechanic that always had interesting chaos happening around it. A lot of rounds I played during development were on CS:S or Zombie Master maps though, so I’m kind of the wrong person to ask. I’m sure some awesome maps were created in more recent years that I have never played.

A brief aside: I decided to take Stefan's advice and asked a few people who still play it: The Yogscast!

You’ve been playing TTT for years. What keeps it interesting?

Lewis Brindley: The amazing community is continually keeping the game fresh with new maps and items - or even new roles - such as the Jester (who wants to get himself killed, which is super fun). Tom Clark: The Steam workshop - New maps, guns, toys to play with.

Has anyone properly fell out over it?

Lewis Brindley: Never - of all the games we've played together TTT has caused the least arguments, which is surprising since the whole point of the game is to backstab your friends. Tom Clark: Zylus and Rythian are sometimes salt lords. There has never been a proper falling out though. Ben Edgar: There have been a few RDM incidents which have taken a while to get over.

Got a favourite episode/moment/plot twist?

Lewis Brindley: This short round shows off a lot of the best stuff. A fan-made FTL map, costumes, good use of traitor weapons (chainsaw) and detective weapons (handcuffs). Ben Edgar: This is one of my favourite recent episodes.

Information The people playing it as a real life airsoft game, that’s definitely the most surprising. In some ways it’s like a return to the Serial Killer origins of TTT, in how more of it runs on an honour system.

Yeah that was pretty awesome, and completely out of the blue. That’s in 2013 too, which shows the degree to which Youtubers spread the word. Though I must say, for aspiring modders who might feel Gabe liking their stuff would be a big deal: it’s not like I was offered a job at Valve the next day. Kinda nothing happened really. The guy is also a big World of Warcraft player so I’m not even sure I can speak to his taste in games (Gabe I’m sorry don’t cancel HL3 please).

Information There was a period after university where I looked for a job in that realm, but the games industry in the Netherlands is tiny. Plus as I had thoroughly experienced, programming video games is actually kind of… bad. You end up writing gross stuff all the time in terrible programming languages to make things run fast, while working around bugs in whatever horrible engine you have to use. Who knows, I might return to making games someday. The only way I can see myself enjoying it would be in smaller indie games, but that’s a really tough space to succeed in right now.

There was not really a single end. In a small way I still work on it by integrating contributions from people submitted on the GMod Github repository. I think the biggest features were hammered out by the time TTT started being included in GMod some time after it won Facepunch’s Fretta gamemode contest, but I still did a lot of polishing through 2011. Since then it’s mostly been occasional fixes for things GMod broke, and the aforementioned contributions from people. I think there are a couple of reasons why I started working on it less. The main reason was simply that I had most of my (good) ideas in there, most of the gameplay systems were working well, and adding more stuff would mostly serve to duplicate existing mechanics or mess with balance. The biggest thing I’d want to do is go in behind the scenes and clean up the code, given that it was hammered out quickly with the focus being on iterating on features, and not so much on stuff like keeping things nice and consistent. Problem is that people have been writing all kinds of addons and extensions, and anything major I change now could break a bunch of them. Another reason is that I finished university and got a job, so I simply have less free time. I’ve spent literal hundreds of hours working on TTT, there’s a point where it stops being more fun than actually playing games. Add to that that I’d been working on Source-based games for years at that point and I’m kind of done with that engine.

I still get a lot of Steam friend requests from total strangers! That’s about it. To this day I’ve never even met a person in real life who had heard of the game (let alone played it) before I told them about it. It didn’t get me a job or open doors for me or anything special. Of course there are intangible things though. I learned a lot from it, and it was satisfying and a lot of fun to make something people enjoy, and have kept on enjoying for a surprisingly long time.