Terror Chasm

Terror Chasm is a multiplayer horror puzzle platformer with massive replayability. You and your friends will navigate treacherous 3D procedurally generated levels as you try to escape a deadly ancient temple full of traps, puzzles and a botanical vine monster which rips through the walls and can come at you from any direction! This is an up to 4 person PvE online co-op style game in which all the players will have to work together in order to beat the level. Each playable character will have different abilities and strengths that they use to help out the team as a whole, as they try to get out of the temple alive.

How Terror Chasm Was Designed

It all came together in one conversation. Just a couple of hours of talking and we had 90% of what would become Terror Chasm; an idea that we were very pleased with and proud of. Of course, that conversation happened after months of researching the game market, agonizing about many other not-as-good video game ideas and over a decade of learning how to actually be indie developers. But yeah, in just one afternoon it all came together.

The story of Terror Chasm began with a question- "How do we make a game that a lot of people will really love?" It's common for small development teams like ours to come up with an idea that they are really excited about and then try to convince everyone else without thinking about whether or not this is something that others really want to play. We had made this mistake in the past with other projects ourselves, but this time we wanted to design something which would provide massive value to the fans from the very beginning. Something that we knew a lot of people would enjoy.

So to answer this question we began intensely studying the game market and for several months we committed to learning what was working and what gamers and fans really wanted. We analyzed the numbers, read articles and reviews, saw what people were excited about and what they were getting value from. This was a daunting process and we ended up coming up with and then scrapping many ideas that weren't quite right. Some obviously didn't have the kind of large appeal that we wanted to provide, while others just didn't feel like a Twisted Jenius game. We are committed to our mission of spreading dark fun and so we won't compromise the unique weirdness and creative spooky elements that are so essential to Twisted Jenius.

After discussing this for many months and writing up multiple possible ideas into game design documents that were ultimately not right, we finally hit the point where we said "we just need to do something!". We were about 5 minutes from settling on a game idea that we only kind of liked and had almost zero chance of any mass appeal whatsoever, when we had a breakthrough. The idea sounded crazy at first – "what if you had a platforming game with just one jump?". The concept didn't make much sense, and in fact it sounded stupid and boring. But the idea was that if you could make an entire game about just getting over one pit, that would greatly limit the scope of the development process, which is something we were concerned about.

The genesis of this idea came from the opening sequence of "Raiders of the Lost Ark". If you remember that film, Indy is trying to steal a golden idol from a temple full of traps, including a rather large spherical boulder and deadly darts. There's also plenty of tarantulas. Plenty! But the thing that really catches him up the longest, if you watch that sequence, is a simple pit in the middle of the floor that he has to cross over. He swings over it using his whip when he enters, and almost falls to his death trying to get over it (and halfway out of it) after being betrayed by one of his comrades while exiting the collapsing temple. This was the starting place for the game's premise.

Because our previous projects were very ambitious 3D games, we knew that we could do a lot with what we had as a 2 man team. But we also knew enough to be careful about scope and try to make something that would be within our means and general skill set. We knew it would be a bad idea to try to create massive environments with expensive worlds, even though we wanted to give fans a really awesome 3D experience. Of course two of the keys to a successful game are often player retention and replayability. This is what made the idea of an enclosed platformer with a simple premise like this so appealing. Have really short indoor levels, each with a pit that you have to cross. We quickly realized that by making the entire game and every level procedurally generated, we could effectively provide limitless gameplay to the players without having to manually create every level ourselves. But to do this right we would have to create a new type of procedural generation with the express purpose of simulating intelligent level design using an A.I. system.

It wouldn't be acceptable to just have the standard type of procedural generation with only certain elements of a level being tweaked to give it just a little bit more variety. For Terror Chasm, each 3D structure would have to be set up completely by the A.I. in entirely different ways for every level of the game, which would massively effect the gameplay and this level design would have to be able to rival the efforts of a human level designer. Fortunately, we consider A.I. our specialty, and we have no personal qualms about effectively outsourcing our artistic vision to machines. We've used our own custom A.I. for all of our previous projects (which has always involved some specific implementation that we needed) and we've even programmed and sold A.I. add-ons to other developers. We knew there's a way to do this, we just had to figure out how.

Of course, being a Twisted Jenius game, we had to include some horror elements and weird creatures. Working off the premise that the environment itself is your enemy, we decided that the guardian of the temple would be a giant plant monster who's vines would snake through the walls and be able to shoot out and attack you anywhere, from any direction. We felt that this would give an adequate level of menace and paranoia to the game, while further leveraging our A.I. skills. From a purely mechanical gameplay standpoint, the plant serves the purpose of a "timer" for the players, making sure that they can't stay in one place for too long and they have to hurry and get out. The choice of a plant as the enemy was due in equal parts to a lifelong love of "Little Shop of Horrors" and a recent, coincidental viewing of the horror film "The Ruins". Other games which we drew inspiration from while creating Terror Chasm include Portal (a puzzle platforming game in a stark, claustrophobic environment), the Tomb Raider series and Dead By Daylight, for simply proving that a multiplayer horror game can be popular and commercially viable.

It took almost 2 years to create the design documents for the game. It has been said that “every minute you spend in planning saves 10 minutes in execution" and we knew from our previous experience that the last thing we wanted to do was to fumble our way through a needlessly long active development phase. Our goal was to front load the planning and strategy, so that we could minimize and streamline the execution, which can often be even more time consuming and expensive if done improperly. Basically we wanted to be extra smart about this. One of the fundamental themes of Terror Chasm's development is to maximize our admittedly limited resources and leverage our creativity, brain power and the tech as much as possible, to get results that far exceed what anyone would normally expect.

The part of the game design document that took the longest was the procedurally generated puzzle section. I'd imagine that just the idea of having procedurally generated puzzles in a 3D action game like this might seem daunting, if not completely crazy to many indie devs. But we were determined to find a way to do this properly. We studied many different types of puzzles, not just from video games, but the principles behind puzzles in general, and tried to break them down to their bare essential elements. This would not only help us to plan them but to communicate how to create them to the A.I. that would be responsible for providing them in the game. It was also decided that the smartest way to procedurally generate the levels would be to create the puzzles first, and then build the physical structure of the temple building around them. This was especially important since this is also a platforming game and many of the puzzles interact with or involve the pits or "chasms" around the level.

During this time we also began doing concept art and decided on the look and feel of the game. Instead of going with the standard "realistic" temple interior that can be found in many different games (Tomb Raider, Uncharted, etc.), we decided to take a more stylized inspiration from Frank Lloyd Wright's textile block "Maya Revival" architecture. Our primary reason for doing this was that we wanted the game to look distinctive and unique. Ideally, we wanted any screenshots or trailers of the game to be instantly recognizable as being from Terror Chasm. This aesthetic choice had 2 additional benefits. The first is that it worked much better for procedural generation (arguably, this was kind of the purpose behind Wright's actual textile block buildings, the blocks were easily manufactured and could be arranged and stacked in any number of ways). The second benefit to this design was that it made for a fairly even and simple interior wall surface for the plant to rip through. This is important because being able to tell where the plant is, as it's tearing through the walls, is essential to game play.

The next step in creating our 3D platforming videogame was obvious- mockup a physical turn based, D&D style board game prototype to test it. No, wait, that wasn't obvious. It's actually counter intuitive and kind of screwed up. But we did it anyway as a proof of concept to get a very quick rough idea of how this thing would actually play based on our game design documents. We used the same principles or set of rules and parameters for the procedural generation that we would eventually use when programming the game, but we did it "manually" (using things like charts and multi sided dice for the randomization elements) to create this 2D board game map that we would use to test this concept with.

We learned at least one important thing from this proof of concept, that the plant would be a bigger menace than we had originally thought. Up until this point we had been working under the assumption that the puzzles would be the biggest factor in gameplay for Terror Chasm. But even in this primitive turn based board game version, it became obvious that the plant would be a very potent source of suspense while navigating the temple. It may have just been pieces of string on a flat, hand drawn map, but this thing was scary! Scary, creepy and relentless. We began to realize that the botanical monster that can come at you from behind any of the walls might just be the game's biggest hook and the most blatantly interesting and iconic element of the game.

It was finally time to test this out as a real video game. We designed a level using the same manual procedural generation that we had done for the proof of concept, and then we began constructing it in the Unreal Engine (the game engine we decided would be best to use for this project, despite us previously working in Unity). One of the first things we had to do was to create proper A.I. navigation for the plant monster. Creating this A.I. was challenging because the plant had to be able to move along any of the walls, in any direction in the 3D space. This was especially tricky because the geometry of the walls themselves could be inconsistent and change dramatically from level to level (they're procedurally generated, remember?). The plant's navigation had to be able to move anywhere it needed to in the X, Y and Z axes, while simultaneously sticking to the walls and also facing in the correct direction at all times (so that the plant can attack the player). This required a lot of math and testing. A special test room was constructed in the game engine just to see how the A.I. navigation would respond to the geometry of difference size walls and spaces.

During this time we also began refining the game's aesthetic. Using the concept art as a reference, we begin experimenting and seeing what the lighting, textures and props should look like within the actual game environment and created them accordingly. This also involved coming up with various rules for how these things will be automatically placed in the game environment that we can then communicate to the procedural generation system. We also began testing physics mechanisms and looking more closely at certain elements of the gameplay such as platforming (which was one element that we, quite understandably, had some trouble testing in the flat board game, proof of concept version). We decided to use a generic stand-in character model for these first test iterations of the game. While the actual playable characters are important, we realized that in some ways they are also the most predictable and standard element from a development standpoint. So it made more sense to focus on creating this hazardous environment and other gameplay elements for showing off and demoing the potential of this idea (basically we figured that anyone could have a character running around a level, but a wall-ripping vine monster was a little more unique and interesting).

And now as I write this, we have just completed the basic first iteration of the procedural generation system that will be a major component of the game and creating nearly infinite levels for the players to navigate through and attempt to survive. We will use this basic first version of the system to create a vertical slice to show off this game idea and hopefully draw in some support for this strange but exciting and impactful project. Although we've successfully laid the groundwork for how the game will play and we now have a vertical slice and can offer a general idea of what it will be like, Terror Chasm's journey isn't over yet and there is still much to be done.


Character backstory-
The Straying 4 were a group of amateur "thrill thieves" that targeted expensive homes. One night they chose to break into an old 1920s Hollywood mansion, designed by Anton Shandor, the son of the legendary New York architect, and rumored to be haunted. The mansion had many occult artifacts and curiosities. In the middle of their burglary strange things started happening. One of them saw a bizarre floating shadow coming towards them and they all decided to book it out of there. Before leaving, one of the group grabbed an ancient idol off of a pedestal with the intention of putting it in his bag. When he did they were all instantly transported to a temple somewhere in the middle of the jungle and are now forced to use whatever skills or creativity they have to try to escape with there lives.

Temple backstory-
The Mkodo were allegedly supposed to have worshiped, among other things, something that roughly translates to "the great dream flower" (or perhaps a more accurate translation- "terrible nightmare weed"). Nothing is known about this great dream flower except that it was supposed to be big and miraculous. Most botanist assume that the plant was either a complete myth of a fictional plant-god, or an exaggeration of some mundane native species which is now extinct. Both assumptions are incorrect.

The Temple of the Dream Flower was by far their largest temple and likely the only surviving structure of the lost and geographically isolated Mkodo tribe.

Check out our developer log for updates about Terror Chasm.


March 3, 2022- Began closed pre-alpha testing.
January 10, 2019- Terror Chasm is announced.