Dev Blog of Madness

LILYMODEL03

Terror Chasm Development Recap 4: Playable Characters-


Of the major elements of this game, nothing has probably been changed quite as radically from our original plans, as the playable characters have. Right now Terror Chasm is currently in the alpha stage of development, and while we have the foundations for most of the systems and features that will be in the final game, there are a few exceptions to this. One of the most notable, is the lack of multiplayer functionality that will be in the release version of the game. Although we only have one character in there right now (Lilly), Terror Chasm is actually designed to be primarily a four player online co-op game, with the secondary option of also being single player (just in case you hate dealing with other people).

But our original concept was actually for the game to be a 4 v 1 asymmetrical multiplayer style of game closer to other online horror games like Dead By Daylight. The first idea we came up with was to have the four human characters going up against the plant monster, which would also be a playable character. This was our working premise for a little while as we were initially doing the design documents. However, when trying to design the system for how a playable plant would operate, even in theory, we realized that there were some major issues with this. For one thing, to get the plants to act the way we envisioned (being able to rip their way out of the walls as they do now in the game), while simultaneously allowing a player to control them properly, we realized that we would need a completely different user interface system for the person controlling the plant than we would have for the rest of the human players. We came up with several ideas including giving the player controlling the plant a top down view of the level and making their role/gameplay method much more akin to a strategy game rather than the action platforming of the other players. This would allow them to plan out how they were going to attack the human characters as those players moved around the level.

Another idea was to just have the vines have a camera attached to them and the player who is controlling the plant would see everything from the vine's perspective as they're moving through the walls to pursue and attack the other players. This would make the plant's gameplay seem more action oriented, as they would be able to actively go after the other characters in real time, and on their level. Of course to do this we would have to create a system for that player to actually be able to see through the walls to track the other players and that can be very messy and complicated for multiple reasons. Basically, the problem we kept running into with all of these approaches for having a playable plant monster, is that we would essentially have to design a totally separate game for the person playing the antagonist, and then jam it together with the other gameplay of the human characters and hope it works well together.

colorconcept2

But the chances of that happening and working out well are pretty slim. We realized that the gameplay style for the person playing the plant would be so radically different from the rest of the game, either no one would want to do it, or it would be so much fun that everyone would want to just be the plant. It would be almost impossible to make it even with the other player's experiences in terms of enjoyment and have it be about as fun to play as the other characters, because the gameplay would be so radically different. In fact we knew there was a good chance it would seem very weird and disorienting to have gameplay that is so utterly asymmetrical as that. With a game like Dead By Daylight it kind of works because the killers operate and run around mostly like the survivors do. But the plant would have to move completely differently, it wouldn't even be vaguely humanoid at all. And even with DBD, one of its most common complaints from its players is lack of balancing between the killers and the survivors. Can you imagine how much worse that might get if the monster had such a completely different gameplay style? As a developer, I don't even like to think about how to try to balance that properly.

In addition to all of this, designing and implementing that kind of gameplay for the plant would essentially double (at least) our development time. And for a small two man studio like ours, that is bad. This is why we decided to scrap the idea of having a playable monster and any type of PvP (player versus player) and just stick with the four human characters as the playable ones, and make the plants completely controlled by A.I. This also eliminates a lot of the balancing issues that you see in many PvP style games. In Terror Chasm, everyone is on the same team and must work together to beat the level. And even if only one player survives and makes it to the end, the entire team gets credit for the win. Cooperation becomes mandatory and essential for victory.

colorconcept

However, ditching the asymmetrical PvP gameplay wasn't the only major change to the playable characters. Their premise and back story have also gone through several iterations. The first idea that we had was that these characters were going to be explorers, going through the temple as the result of a university funded expedition. But we decided that this was kind of boring. As a company that loves villains, we wanted these characters in this Twisted Jenius game to be darker than that. We wanted them to have some edge to them. So instead we decided that they would be thrill thieves who ended up getting mysteriously teleported to the temple after breaking into the wrong house. This concept was mostly inspired by the real life "Bling Ring" who famously stole from many high-end and celebrity homes. I felt this would open the door for more interesting options with character development and make them more amoral and flawed. But not only did this decision make these characters more "edgy", but it also solved a few other problems as well. For instance, I didn't want them to look like the typical explorer style of action hero like Nathan Drake or Laura Croft. Although the kinds of outfits that those characters wear might be very practical for stumbling around dirty jungles and ancient ruins, they don't look very cool to me. I wanted the characters of Terror Chasm to wear something more interesting, and so by having them somewhat obsessed with living the L.A. lifestyle, and then magically and unexpectedly teleported to the temple, I could more easily get away with having them dressed in something closer to high fashion. This also has the added bonus of explaining how they could even have a chance of navigating these traps and obstacles and surviving this situation in the first place; being experienced burglars who know how to get in and out of awkward places where they're not supposed to be.

Characterconcept

The four characters are named Skeeter, Lilly, Daisy, and Sharky. However this wasn't the original lineup. Back when they were going to be explorers from a university, the character of Daisy wasn't a part of this. Instead, that character was originally going to be an elderly professor who was supervising the archaeological exploration. This is important from a gameplay standpoint because each of the characters has their own unique skills or advantages that they can add to the team. Lilly is the smallest and lightest and she can balance on things better and fit into places that other characters can't. Skeeter is the tallest and can jump the highest. Sharky is the strongest, and heaviest. And Daisy is the fastest and can jump the longest. But Daisy wasn't in the original plan and the professor character had different attributes. Despite being at a physical disadvantage, the professor was going to be able see and interpret puzzle clues and hieroglyphics that the other characters could not. But much like the problems with the playable plant monster, this presented some major challenges such as having to give the player whose playing the professor a different U.I. than all of the other characters. We could also imagine there would be some serious potential problems with this character having to communicate with the other players quickly and accurately in such an intense and fast-moving game. And once again the gameplay for that character would be very different from the other characters which could create issues with how desirable playing that character would be. One of the advantages of having a lot of game design experience, is that you can see these things coming before you even have to begin actual development and so we didn't waste any time even prototyping a system like that and just scraped the professor character in favor of the speedy Daisy instead. Her gameplay is much more similar to the other characters and of course this worked much better once we changed the premise from university archaeologist explorers, to thrill thieves.

LILYMODEL01

Right now, Lilly is the only character in the game and this version of Terror Chasm that we have is just a single player experience. Part of this is for development purposes (it's easier to set up and test for one player) and the other reason is because we wanted to make sure we had a single player option. But this is designed to be a primarily multiplayer game and that is one of the biggest things that we still have yet to do for development. Most of the other systems of the game are things we've already started, have an established foundation for, and are currently operational to one extent or another. But this one needs the multiplayer system functionality, as well as the actual characters themselves and plenty of testing and balancing to go along with that. Certain aspects of the puzzles and even the level generation will also have to be altered to accommodate multiplayer.

This blog marks the end of the four-part recap series (here's part 1, part 2 and part 3) designed to catch you up on what we've done so far. We've accomplished a lot and we still have some more to go. I hope you'll continue to follow along with us on this dark development journey. Video games take time, but this is happening. Terror chasm is coming...

- False Prophet

munchweed04

Terror Chasm Development Recap 3: Plant Monster-


At Twisted Jenius, we love our monsters. And one of my personal favorites has always been man-eating plants (due mostly to my obsession with Little Shop of Horrors as a kid). However, with few exceptions this type of creature is usually treated as a side threat, rather than the primary menace, in most media that depicts it. Even in The Day of the Triffids (1962), it could be argued that the global blindness in the movie is a much bigger threat than the titular plant-life itself.

This is why we really like the idea of making a plant the main monster of our game. This also really reinforces, and even epitomizes the idea of PvE (player versus environment) as a game design style, since plants are so often heavily associated with the concept of "environment". Not only are there dangerous traps and platforming chasms that you have to traverse in regards to the temple itself, but there is literally murderous vegetation coming out of the walls to attack you! From a symbolic standpoint, it's very blatantly telling the player that the environment is against them. I'll also mention that another inspiration for using a botanical antagonist in conjunction with the idea of an ancient temple, came from the 2008 horror film The Ruins, though the plants in Terror Chasm are much more active, tearing through the walls to pursue you with the speed and ferocity of a hungry predator.

vine01finle

From a gameplay standpoint the plant monster is designed to create a sense of urgency and limited time. You can't stay in one place for very long without dying and even when you're not jumping over the chasms platforming, you're still never safe. The vines can rip out of the walls or ceiling and come to get you from any direction. Your only choice is to keep moving and the longer you remain in a level, the greater the chance that you will die. This creates a constant feeling of suspense and tension for the player, with ominous music playing as the plant grows nearer. And in addition to this, you can't even fight back. All you can do is run from it and this creates a feeling of helplessness in the player. Terror Chasm is not a power fantasy style of game, and is instead emotionally closer to survival horror in the sense that you are at the mercy of the environment and you must struggle and fight to survive against overwhelming odds. If you push forward too quickly you could easily fall to your death in one of the chasms or get caught in a bone crushing trap, but if you hesitate or stand around too long the plant will get you. Basically, as a gameplay mechanic, the plant monster functions as a device to keep the player moving and keep them on guard constantly by adding another variable that they have to contend with in addition to the puzzle/platforming aspects of gameplay. And from a symbolic and even marketing standpoint, the ferocious vegetation serves as a more distinctive antagonist and icon for the game that is a little bit easier to pin down and solidly depict than more vague ideas about the environment being your enemy. It's a tangible monster.

vinewallconcept

These botanical predators come in several forms. The most common one in the game right now is the simple vine that whips out to tag and pierce you. The idea is that these vines grow through the walls like a normal jungle vine, but at an accelerated rate and do so consciously, with the deliberate intention of pursuing the player. They rip through the stone walls of the temple like a train barreling after you, and if they catch up, they get you. When they shoot out at the player they don't instantly kill them the first time, but they do inflict damage.

Of course being fans of the classic Venus Fly Trap style man-eating plants, we naturally had to have one of those in Terror Chasm. This is the dreaded Munch Weed. In the current version of the game, which is still in the alpha stage, the Munch Weed acts as a sort of finishing move for the plant. When it decides it is going to stop playing around and finally kill you, it quickly grows out of the wall and snaps around the player's avatar and promptly gobbles them up. While the act of eating the player is not likely to change, it's possible we might expand the Munch Weed's role in later iterations of the game. While death-by-Munch Weed is not the most common way for players to perish in Terror Chasm, we still consider this fiendish flytrap to be the most notable mascot of the game.

Annotation 2019-04-02 032513

The most challenging part of developing the plant monster for Terror Chasm was designing its navigation A.I. The vines have to be able to chase the player while moving around all of the surfaces of the walls and ceilings. This is more complicated than you might imagine, because it has to be able to navigate a truly three-dimensional space, while simultaneously sticking to certain rules (namely it has to run along the walls perfectly), and on top of that it has to locate, chase down and attack the player (which means it also has to be consistently facing the correct way, while it's doing all of this). This is a bit more challenging than most enemy A.I. navigation that only runs like a human along a two dimensional floor. That type of A.I. navigation only has to worry about moving along 2 axes (X and Y, or if you prefer, front and back and side to side). But the plant's A.I. has to be able to travel the Z axis (up and down) as well, and that creates much more complicated math. The standard navigation A.I. from the Unreal Engine that we're using to develop the game, wasn't able to do what we needed. So we had to program our own custom A.I. system. We even set up a virtual test room within the engine to make sure that the A.I. knew how to properly navigate all sorts of different wall configurations.

Munch-WeedAri

In the current version of the game the plant's A.I. only has one chase behavior, where it relentlessly pursues the player. As development progresses, we plan to give the plant monster a wider variety of behaviors and different ways of doing things (for instance, sitting and waiting to ambush a player, or choosing to go after only one specific player once we integrate multiplayer into the game). We also have plans to create even more and different types of plant attacks and different forms the plant can take. By giving the plant A.I. more ways of behaving, this should make this enemy even less predictable and the game even more "interesting" (a.k.a. freaking and disturbing).

- False Prophet

enviormentblog2

Terror Chasm Development Recap 2: Environment -


Quite a lot of thought went into planning the aesthetic look and atmospheric feel of the environment in Terror Chasm. We like to think of the game as PvE (player versus environment) to the extreme. The entire environment is very literally your enemy when you consider all of the traps, puzzles, and deep dark pits that you must platform over (not to mention the man-eating plant that rips through the walls to come get you!).

The idea is that in many ways the temple itself is a character and arguably even a literal antagonist in the game. This is why we wanted to make sure it had a certain amount of distinctiveness and personality. In a sense it is "alive" due to the fact that it constantly changes, offering new combinations of deadly obstacles to traverse every time a new level loads.

The basic premise for the game was inspired by the opening to Raiders of the Lost Ark with Indiana Jones and we decided very early on that we were going to keep the temple as a setting. But we didn't want it to look too much like a standard "realistic temple" that you see in many other games with a similar premise. We wanted to make it a little more stylized, while still being recognizable as being an ancient temple. Part of this was also for marketing because we wanted to be sure that there would be no confusion whenever anyone saw screenshots from the game; we wanted it to be instantly recognizable as being from Terror Chasm.

1

In order to make it more interesting we decided to take inspiration from Frank Lloyd Wright's textile block "Maya Revival" style of architecture. I've always felt this had a very unique look and I'm quite fond of it. Originally the plan was to make the surface gray, but after a lot of feedback about the first concept art, it seemed that everyone unanimously said it needed to be a different color so we made it brownish beige instead. Technically this color scheme is probably more accurate to what actual temples looked like.

Certain other elements of the temple, such as the wall sconces which hold fire to light your way, were also roughly inspired by some of Wright's work.

One of the many aspects of the art that we still have a lot of work to do on, are the many props in the game. In fact I'd say most of the ones that are in there now are going to be either updated or completely replaced as development continues. But one idea that we've been playing with is this notion of sort of a "temple punk" kind of aesthetic where we take certain objects and other items in the game that may not have necessarily existed during the time of the Mayans or Aztecs (or any ancient civilizations), but we give them a more "temple" kind of feel, in much the same way Steam Punk does with its Victorian take on some modern technology. I'm not sure how far we're going to take this yet, but it's an idea that we are considering and will likely implement to one extent or another.

colorconcept1

The lighting of the temple was something else that we struggled with. Originally we were just going to go with very realistic lighting, but that seems kind of boring and once again not distinctive enough for what we wanted. Eventually we decided to go with "comic book" inspired lightning, to make the entire setting seem a little bit more vibrant and colorful. The main idea here was to use various in-engine lighting tricks to make the dark or shadowy areas of the game appear blue lit, instead of black. This is the kind of thing that you will see in comic book art. In some ways I think this also helps with gameplay a bit, because it allows the game to feel dark while still allowing you (the player) to see what you're doing. This is especially important for things like puzzle solving and platforming.

Another thing that we really wanted to do was to have sort of a black light effect on the players as they went through these dark areas, making them appear to glow as if under a black light when in these blue lit sections of the game. We thought this would be kind of cool and surreal as well as have a potentially positive impact on gameplay, and the player's ability to see what their avatar is doing. It's also another thing that would make Terror Chasm more visually distinctive. The parts of the game that are not shadowy blue are lit by the torches which run along the walls of the game. When the player is in the vicinity of these, the character reverts back to their more normal lighting scheme and ceases to glow. The idea here is to try and replicate the same thing that happens in real life when you have something glowing under a black light but then introduce a normal light or candle to that object. The glowing either stops or diminishes.

enviormentblog-min

Since the player is running down these mostly linear and somewhat claustrophobic hallways of this structure, the end result is that they periodically stop or start glowing depending on their vicinity to the torches. This is not meant to be a gameplay thing, but just a cool aesthetic. As if the characters themselves are phosphorescent in this bizarre, almost otherworldly temple that they have been inexplicably teleported to and are now fighting to escape.

- False Prophet

TEMPALSTART6

Terror Chasm Development Recap 1: Level Generation-


Over the course of the last few years we've been working diligently to bring you our up coming platforming game which we call Terror Chasm. We are very pleased with our work and love the progress that we've made on it. However, due to equal parts needing to focus our energy on actually making the game, and the simple acknowledgment that it's likely no one has really been paying attention until now, we haven't been keeping up with the devblogs. So to remedy this and get everyone caught up, here is the first of a multi part recap of what we've done and our progress up until this point.

One of the most important features of this game, especially once you begin playing, is the procedural level generation. The idea here is pretty simple- gamers like replayability, and since we are a small studio we don't have the resources to hand-make a whole lot of content such as massive open worlds. So the solution that we came to was to design an A.I. system that can essentially produce infinite different levels for the players to enjoy. While procedural generation has been around in games for a long time, we had to create an entirely new system for this, in order to achieve what we really wanted. What we needed was something that would be more akin to A.I. driven level design, where the A.I. is acting more like a human level designer, rather than the standard kind of procedural level generation that other games have used in the past. The kind of extreme procedural generation that we are going for is not what you typically see in most 3D games, which often use much simpler and more basic methods to procedurally place objects in a level and rarely use it for anything that could dramatically affect gameplay.

We needed something that could create levels in a more literal sense. A system that could "intelligently" place, swap out, and interchange any and all of the separate elements of the game (traps, puzzle elements, platforms, etc.) and reconfigure them into different levels, which would of course create different types of challenges for the player in terms of gameplay. This meant that not only would the separate components of each level have to be intelligently placed by the A.I. in a way that the player could navigate, but the physical structure of each level (walls, floor, etc.) would also have to be custom configured for every level that is generated. The truth is, until very recently this was technically not even possible. It's only with the most recent advancements, with things like the Unreal Engine, that this sort of thing can be reasonably done for a commercially available game like this.

profgood03

We spent nearly two years designing this system and figuring out how it would work. We wrote up extensive design documents talking about how the game would handle puzzles, traps, the platforming obstacles and other items in the game. The next thing we had to do was test this idea, but rather than jumping right in with extensive programming, we decided to test the concept on paper, literally. We created a D&D style turn based mockup and used something similar to what we were planning for the computer to use to design the levels, only in this case instead of a computer algorithm we just used some multi sided dice and corresponding charts to determine how we would create the level in this 2D, hand drawn test. We then used little rubber animals as avatars to play through the levels and see how well this idea would actually work. This first test went pretty well and we learned some valuable things from it.

The next step was to create an actual playable game prototype, so we used the same process to create a level that way and mocked it up in the engine as a simple level that we could play through just to see how this would work. This gave us a great sense of what the gameplay would actually be like and allowed us to correct and refine things even more.

demolevel02

The most significant task was yet to come. This was of course creating the actual code and systems that would allow this A.I. level generation to function. It took many months but eventually we had a system that could intelligently create playable levels for the player to navigate. The very idea that we could instantly generate hundreds or thousands of different original levels with just the click of a button was extremely exciting. Because we want to provide these levels to the player automatically, the A.I. has to be able to generate these levels entirely on its own with absolutely no direct oversight from us, the developers. The A.I. has to be able to generate millions of these types of levels and each of these has to be unique, playable and winnable. This was quite a task but once we had this system up and going, we were extremely proud of it. It's not the sort of thing you'd normally expect from a 2 man indie team.

Since we've got this system up and running we've continuously added more complexity to it including giving the A.I. the ability to generate multi-story levels, where the player can go up or down in different parts of the environment. We've also added more procedurally generated decorations like plants and moss that you will find throughout the levels that you are trying to beat. And even more recently we've added other types of physical obstacles that you must navigate within the hallways of the temple, in addition to the traps and puzzles that we have lurking around every corner.

But this is not the end, and in fact we have lots of plans to continue to add to the complexity, uniqueness and challenges of what our level generators are able to supply to the player (our plans include the ability to generate more complex, and less linear level layouts, among other things). As development continues, what our A.I. is able to provide for you in terms of various gameplay and platforming challenges will only become more interesting (and terrifying!).

- False Prophet

Horrorgame Graphics vs. Gameplay: What's Best For User Experience?-


Which is more important for a game developer to focus on, graphics or gameplay? What do gamers really care about? Here's a new devblog (actually kind of a rant) about designing games, realizations about user experience and the issue of gameplay vs. graphics and what I'm learning on my own indiedev journey to make the next great horrorgame. Basically I just felt compelled to sit down and share my thoughts on this (hey, the original version was, like, twice as long; but it was a little too ranty, so I did everyone a favor and edited it for time),

- False Prophet

First Dev Logs Previous Dev Logs
Next Dev Logs Latest Dev Logs