The Finals (but not the end)

THE FINAL UPDATE

An entire semester's worth of work has gone into this project and it has come a very long way from a mannequin running around a flat plane throwing rocks around. From learning basic enemy AI, throwable items, blockouts, to lighting, sound, and polish everything has culminated in this final submission! I'm very happy to present Cavern Crawler. A stealth game where you find yourself in an underground prison looking for your gear and a way out. In this devlog we will discuss some of the changes between our midterm submission and the final.

Basic Enemy AI

In the beginning we had a simple AI that would randomly patrol around the plane but we wanted to give patterns the player could read in order to plot out their movements from spot to spot. We ended up placing down patrol points that the enemies would move from however their speed was to fast and during playtesting players found it unforgiving that their was no way to escape once a guard locked on to you. So we toned down the speed but then it became it was to easy for players to ignore the guards and just run laps around them. We did end up coming up with a middle ground that was giving the player an ability to crouch hide making them go invisible and the guards to return patrolling. In playtesting players ended up just perma-crouch walking so we then added a timer and a cooldown to force players to use it as a tool to escape instead of just avoiding guards all together. While some players did find fun and creative methods to "break the system" most found it useful and was the exact middle ground we were looking for in regards to making the enemy AI threatening without making the players feel totally helpless. We polished this ability by making the player visually become transparent and adding VFX particles around the character to visually display when they were/were not in hiding. Lastly we included a HUD icon that displays when the ability is available for use as well as a graph depicting it "recharging" while on cooldown.

Terrain/Blockouts/Level Design

While building we had originally used simple blockouts but during the time between midterms and finals we took the time to update our orange blocks and clear blue skies with a more fitting underground setting. This did present its own unique set of problems (See "Lighting" for more on this) but it also made the game feel like it was starting to come out of its prototyping phase and into something more intentionally designed. The floor terrain was a challenge since the smallest bumps and valleys caused movement issues with the models getting stuck beyond little ridges or trapped in small valleys so it took many passes in order to get the floor right. We also decided to add more context to your environment by adding a giant "hole" in the ceiling with light coming down to help the player feel immersed in their new underground environment while simultaneously providing a landmark the player could use to gauge their position in the prison. 

Custom Characters/Movement

With a lack of proper models to use and no time to build/animate from scratch we decided to re-material the base models so you could tell them apart from each other with a distinct glow that can make them easier to point out from a distance. We also added movement to each model. Guards, as mentioned before, will patrol their points with a particular affinity for chasing down the player if they come in sight. Prisoners will follow the player unless the see a guard, in which case they will run the guard down, eliminating both actors from the playing field. Players can use these prisoners to act as a body guard from the guards or they can go down another path with a new trick in the bag. We updated the Player Character with a unique hide skill (See "Basic Enemy AI" for more on that process) so they could ignore the prisoners all together and look out for only themselves! This skill felt very thematic for the game and was a natural evolution of the "throw rock" mechanic previously used. During playtesting players didn't enjoy having to find keys, find gear, find rocks, find prisoners, find exit, etc and the rocks being a one time use and then forcing the player to go hunt down more didn't feel fun. We tried adding more rocks and making the reusable been then the mechanic felt cheapened so we decided to tackle a new idea. Plus sometimes the rock wouldn't correctly throw in the direction we were looking (sometimes it would go in the opposite direction!) so instead of tackling and re-coding a mechanic we weren't 100% in love with we decided to do something different! We hope you enjoy it

Lighting/Audio

This was probably were we spent most of the time given. What was originally suppose to be a "screw in a light bulb and plug in the aux" turned into a multi-week ordeal of sorting UV's, mixing/layering audio, and trying to make everything feel natural to environment. The first attempt at adding Lighting to the environment caused the area to be lit with the full concentrated power of the sun 

It also caused crazy shadows to run all over the floor and the walls so we had to experiment with different lighting types and positions because the primary light source needed to come from the giant hole in the ceiling with smaller secondary light sources coming from the torches on the walls. The torches were all handmade by kit bashing some already existing model's and VFX's together to produce soft lights to break up the shadows. With some work we managed to improve the lighting quality from its starting point (top) to now (bottom) while still not entirely perfect the game was running much smoother and not throwing up angry warnings every time it loaded. Players reported a more natural feel to the environment without some areas feeling super bright or super dark. Postprocessing also went a long way in helping smooth out the lighting by increasing the contrast and adding some color temps for a more underdark aesthetic. 

After lighting was sorted Audio became a whole new beast. Doors opening were to loud and all the footsteps from every character could be heard at the same volume simultaneously. We wanted players to be able to hear the footsteps of approaching guards without having to swivel the camera around all the time to keep them all in sight but the constant reverb of loud footsteps made it hard to determine what was actual near you and what was your own footsteps. We also wanted to include some soft ambience in the background so players felt the cold wet nature of the under ground lair they found themselves in. So after tinkering with attenuation and adding anim-triggers to the player model we were finally able to get all the noises and a level that didn't have the player feeling like Tetsuo from Akira

Playtesting

Playtesting was such a massive part of this project. I had some close friends play thru multiple instances of the game and they gave me some great feedback every single time. It was this feedback that helped find some wonky terrain, game breaking mechanics, or simple blunders (like accidentally leaving all your Booleans as "true" allowing the player to gun it to the finish line without actually playing the game). In the end it was these play testers that encouraged me to constantly try new things, like adding a hide mechanic, clearer text for the dialogue by adding some color to it to stand out from the white flooring, a simple quest tab so players knew what their tasks were, etc. and help get this game across the finish line. Id be remiss to not include some of the best feedback I received from a play tester.

Conclusion

In all this assignment did so much more then show me how to blockout a level or implement some sound. Learning how to troubleshoot certain parts of a game, think about how players will interact with the product you place in their hands, how to take in feedback and implement it into something useful, and how to prioritize what's worth saving and what needs to be scrapped/redone to fit a deadline are all super important skills for anybody pursuing game development. This level submission has so much potential for more and the longer I worked on it the more I found myself wanting to expand it. Why did the player find himself in this prison? Who are these guardsmen? What awaits for them after escaping? So much potential lies within a short game made over the course of a semester and I'm eager to dig deeper (pun intended) to see what more can be done.