Midnight Ultra: Devil's Pass
As with my previous work with Forward Instinct, I served primarily as Level Designer and QA Lead. I was primarily responsible for Levels 4-2, 4-3, and the second boss arena, as well as the design for both boss battles. I also assisted in the design and layout of levels 4-1 and the first boss arena. For the photos below I will be using the expansion's main color scheme, the blue and white "Ice"
If you want to learn more about the game in general, information can be found at the following links
After Midnight Ultra released and feedback came in, a common trend we noticed was that players enjoyed the second world more than any others, due to how open and fast paced they were. Taking this feedback, I set out to expand upon these designs.
A common grievance that was heard was that the final level featured too many gates and was very linear, a complaint that, in hindsight, I agree with and wished to avoid repeating. As such, every level was made with alternate paths in mind. While the specifics will be outlined in their respective sections, levels 4-2 and 4-3 were both designed to have at least two different ways of approaching the ending trigger, and in the case of 4-2, several different ways of approaching those two paths.
Early on there had been talk of using the Palms Beyond in the later sections and have the player swap between the Witch's magical dimension and the grounded, human dimension, but I ultimately decided that the foggy mountains were a better setting for the story we were trying to tell, and there was no need to complicate it with dimension hopping. In removing these sections, I was able to make more cohesive levels and allow for more player freedom in how they chose to approach each path.
Level 4-2 is by far the level I scrapped and restarted the most, as early playables of it date back as far as when the base game was still in development. Those versions were unfortunately lost to a merging error, so no pictures exist, but the core of it was a small forest area and a grouping of free standing cliffs. The forest had a few walkways and an elevated section, containing several jump spots to islands within it, and basic enemies that would attempt to follow your jumps. The Cliffs on the other hand were spaced out just far enough from each other that you could easily keep momentum through the entire area with only single jumps, as some players expressed frustration at the double jump sections located in the late game. While the concept map did not exist at the time of starting Devil's Path, the cores remained and were expanded on.
The first working idea for 4-2 is pictured above on the left, and was intended to be split between a ski lift and forest pathing, with the ski lift focusing on platforming and the forest on combat. Issues with this design quickly came up, as I had trouble coming up with a decent way of mixing the two sections with the layout I had in mind, and there were concerns that the diagonal verticality would get players caught on edges and slow gameplay. A few tries later resulted in the image on the right, which became the true prototype for 4-2. Abandoning the Ski lifts, I decided to instead focus it on tight cliffs and wide open combat sections. This version still had trouble with pacing and difficulty, as hugging the left of the map would get you through without seeing much combat, but it was getting closer to the intended experience.
The "Alpha version" of the map was decided on after I took the base of the yellow image above and edited the terrain directly. After a few hours of raising and flattening it, I had finally settled on a map.
The new design started in the forest area, an open low land with multiple ways out that were marked either by pick ups or enemies that could not be encountered in the forest. To the left, players could climb up steep hills and platform across a broken path guarded by the new Snipers, Yetis and Thunderbirds, take the central path to face off against Witches and shamblers, or stick right to go up against the usual crowd of ranged and melee enemies. The Center path was designed to split off to whichever way the player saw fit to follow, though due to the fog it may not be immediately obvious which section of path lead where. After a few more passes to improve game feel, the level finally reached it's final state, as shown below:
One change that you may notice from the previous maps is that the arenas have been completely replaced with large enemy clusters, so the only limiting factor to game completion is how fast the player chooses to move through the level. There are also more plentiful ammo and health pick ups, but as they're more spread out it makes it less likely for the player to end a level with 250+ ammo, as was the case in 3-2.
As mentioned earlier, a big change from the base game was the addition of fog, which drastically reduces player vision. While this allowed me to build a more open labyrinth, it also posed a few challenges, such as how to guide players to the paths forward. To get around this, I carefully placed certain enemies and items right at the edge of player's sightlines, which prompted more investigation and subtly guided them through the level.
When the player spawns in, they need only take a few steps to see several enemies. Depending on which enemy the player chooses to engage, the level will invite them to continue moving in further. To the left, a Sniper will fire at you from just beyond your sight line, prompting players to move up the hill to engage him and investigate the Cliff Path, and to the right players will see melee enemies who have cleared out part of the forest, with some Shamblers on the area above them, leading towards the Camp path. For clarity's sake, I will follow the Cliff Path first, then loop back to the Camp Path.
Should the player take the left hand path, they will climb to a higher plateau and engage with several enemies. Directly past them is the cliffs themselves. While they pose no real threat to any player with platforming experience, they are made more dangerous by Thunderbirds and Yetis, both of which can knock players backwards to their doom. The cliffs are also littered with ranged enemies who have no problem engaging the player while they're occupied with Thunderbirds, so good enemy management is a must.
This section was more inspired by the original layout of 4-2, with high momentum plateaus towering over deadly drops. The addition of Thunderbird enemies was absolutely the biggest influence on this section, as they changed it from a simple skip across to a dangerous sprint where you could be knocked to your death at any second. The Yeti is less of a threat than any other in the level, but he serves the purpose of teaching Cliff players that they can rush in at any moment, giving them an easy ambush before the tougher to see one later on.
Once the player moves ahead past this point, they will come across a curved path that contains stronger close ranged enemies. The gap between the sides is spaced so that a good jump can clear it and give the player breathing room, but it's a risky move as failing will drop them to their doom. A final yeti lies in wait past the bridge, just behind some rocks. Once the player has taken care of these threats, they can move towards the exit, which is blocked off by some sinewy substance that can be destroyed with some minor damage.
Should the player take the right-hand path, or the lower path, they will set out towards the camp. As seen in the pictures above, the low path will lead the players towards a Witch and her Shamblers, which can be cleared out and leads to the area before the camp. Taking the right-hand path will lead to a straight away filled with various enemies. The player can either avoid them or fight through to the camp path. Should they so choose, there is a lower area that has a Yeti and several Shamblers, as well as a large HP and Ammo pick up. Regardless, the player can continue on towards the right down another path that leads to an enemy outpost.
This outpost features a fairly large and varied group of enemies to fight through, but the camp is also littered with pick-ups, allowing the player to clean house and come out resource positive at the end. Rock walls block the area behind the camp itself, forcing the player to go around the other side alongside more cliffs. Contrary to the left path's cliffs, these are more connected and have a higher concentration of melee enemies. The area is arranged to allow the player to take advantage of their movement abilities, double jumping over the pits and keeping enemies at bay. As the continue, the player will come across the sinew blockage, and can shoot or melee it to pass.
This section is far more straight forward than the alternate path, and is more suited for players who prefer combat over platforming. The cliffs share a similar threat in the Thunderbirds knocking players to their deaths, but the greater amount of pickups means that death is less likely, despite the higher enemy count.
Level 4-3 also went through a variety of changes and alterations. Due to the same merge error that erased the original 4-2, 4-3 had to be nearly restarted from scratch, save for some of the geometry that managed to stay intact. In the original version, the player would start in a cave, and would have to climb vertically to one of two exits. The first lead them deeper into the cave's interior, past the momentum hops that were present in the test level, and ultimately into more enemies. The other would have them go along the exterior and be exposed to sheer cliffs and death defying jumps .
The rebuilt 4-3 attempted to rebuild what was lost, but never quite made it to a testing stage. Keeping the original idea of climbing up and out, 4-3.2 brought back the outdoor towering jumps and the interior's stacked caverns, but the two now met at the end in a cave that worked it's way up the mountain to a final arena fight before the boss battle. After several day's worth of trying, I decided that my current design wasn't up the the bar I had set for this expansion, and I opted to restart the level.
This design was ultimately scrapped for several reasons, the lesser of which being the data loss. The second, and main reason was that 4-2 had progressed quite a bit into its final state, and 4-3 was now just an inferior rehash of the same concepts. I decided to toss everything but a single room, which still appeared as is in the final build. The new version of 4-3 was no longer based on climbing, but instead on leaping. The single jumps finally found a home, and are quite prevalent in this level, once again interrupted by Thunderbirds to prevent them from being too simple.
Additionally, this new layout allowed for what I had originally wanted from 4-2, a criss-crossing path that allowed you to choose between heavy platforming or heavy fighting. This worked out well, as both play testers and fellow developers noted the difficulty of 4-2 and the Final Boss to be higher than the other levels, so 4-3 allowed for a bit of space for players to catch their breath, without feeling like they were being let off easy.
My approach for the final version of 4-3 was also different from the other levels, as I opted to create all of the arenas first, then link them all together in a way that made sense. Not only did this new method allow me to complete the level faster than any other, it allowed me to better space out the encounters, allowing for some better flow.
The player will start 4-3 in a small cavern filled with Shamblers and a single Pistol Scumbag. The only way out is through an elevated section in the back, so the player will have to maneuver around the rises in the cave to reach it. Some players may notice, however, that you can kick off of the stalagmites near the elevations, saving some time and potentially avoid Shamblers altogether.
Once outside, a Sniper will attempt to hit the player, but they can easily be rushed and taken out. Players will also notice the sinew from the previous level is quite prevalent here, and it hides a new type of enemy, the Phage Turret and the Floating Phage These enemies behave similarly to the Witch but are slower (and in one case, immobile) and tankier. Ammo is plentiful in this section, so disposing them should not be an issue. Health is rarer though, so care should be taken, especially if the player has taken previous damage in reaching this area.
From here, the path splits entirely and there is no one way to reach the ending. For this write up, we will first follow the "Main" path, then go back to the "Alt" path, but these names are just for clarification's sake, and the correct path is purely up to player preference.
Upon starting the main path, players can chose to either jump straight for the pillars, or continue to the wall to find a large HP box. Once on the pillars, they can jump directly to the nearest path, or take a side route. The side route has more difficult jumps, but reaching the end will reward players will a large ammo box. However they choose to arrive, players will encounter the second new enemy type, the Floating Phage. These variants move around freely and shoot fast moving projectiles, but can be dealt with swiftly. There are more standard issue enemies on this path, and as you reach the end, the player will be guided down cliffs that rapidly change elevation, forcing them to fend off enemies that appear from their peripherals. Following through leads to the final arena, which will be discussed after the Alt Path.
Should the player opt for the Alternate path, they will ignore the pillars to the side and progress towards the mouth of a cave. Upon entering, they will find Witches and pentagrams spread about, as well as a number of tougher enemies. Players can dispatch these, or continue through the cave, breaking through another Sinew barrier. Alternatively, exploring the cave will reveal a Large HP and Ammo crate guarded by a single Yeti.
Once the cave is cleared, the player will find themselves at an elevated vantage point above the center of the map. They can move straight towards the Slavedrivers to get back to the Main path, or head left down to the second Cave guarded by a Sniper to continue along the Alt path. Should they continue down the Alt path, the player will enter the cave and be ambushed by a Giant Yeti. This miniboss has higher HP and damage than normal Yetis, but moves slower. He can easily be kited around the elevations on the cave floor and riddled with bullets until he drops. Alternatively, again, you can just leave through the back. Once you leave, there are a few pillars to hop through before you reach the lower section of the Final Arena.
Regardless of how you reach it, the Final arena is one of the tougher segments to clear out. Players need only clear the Sinew to exit the level, but the enemy mob will divide their attention before they can do so. Options to do it quickly include sniping the door from the Main path's entrance, or unloading with the Sten Rifle as you rush towards it. Should the player decide to take on the mob of enemies, priority usually goes Witches, Birds, Shotgunners, everyone else, as Slavedrivers and Bat wielding Scumbags can easily be jumped over to avoid taking damage.
From the design side, this was my favorite area to work on, the variety of enemies gives the player a legitimate challenge without being too overbearing, and upon replaying it, gives several options of how to approach it. This level also helped us solidify more story elements, as Phages only appear in or around craters, hinting at their extraterrestrial origins, and the increased amount of Sinew offers a metric for the players to see how far along they are in the level.
When we first talked about the expansion and how it would play out, only two things were set. It would involve the Biker character from the base game, and the Final Boss would be a giant lumbering Yeti. This boss fight went through a few iterations before we settled on what we have now. Originally, the fight would take place in a sealed cavern, or the mouth of a cave, and the Yeti would only be damaged by dropping Stalactites on his head. Later, you were able to climb up to his head, but risked falling and taking Earthquake damage if he slammed the floor. While these weren't bad concepts, none of them seemed to quite reach the level of overkill camp we were shooting for.
During a discussion between myself and Daniel, our lead, we mentioned that it would be perfect if we could set the fight up so the Yeti could only take damage after his skull was cracked open and his brain was hit directly. We ended up dismissing this as we couldn't come up with a feasible way of doing this, until the next week when our Concept Artist, Zach, suggested it be a two phase fight where you enter the Yeti's skull to finish him off. We took that as the perfect solution and completely reworked the fight to accommodate this.
The new Greybox for the fight worked much better, but as Daniel was the one who was programming and testing the Yeti King, final level duties fell to him, and I shifted my efforts to cleanups in the Yeti King fight, and design of the Brain fight.
My designs for how the King worked remained intact howver. The main way he always worked was that damage could only be dealt to his head, and he would use each arm for a different attack. His right was a windup punch that dealt 25 damage and knocked the player backwards, and with his left hand he would slam the ground to shoot out 3 shockwaves that dealt 10 damage. His final attack, which was new to the 2.0 King, was laser eyes that dealt 25 damage and could be used in conjunction with other attacks. Early playtests of this fight proved to be entirely too easy, as unlike the Biker, the King couldn't move and used extremely telegraphed attacks. As a solution, we decided to play to our strengths and drastically altered the once flat terrain to be broken up. We also added regular Yetis and Phages to the level, in order to divide the player's attention and increase the difficulty. These changes were successful, and the fight now presented a decent challenge.
Once the Yeti King has fallen, you are able to enter his eye socket and begin the true final battle.
Once the Brain was decided to be an enemy by itself, several questions were posed. How built to scale should this section be? How will it attack the player? How should we justify this in universe? The answers were, in order, Witch Magic, Witch Magic, and Witch Magic. With that out of the way, I set out on constructing the level itself. Several prototypes were build, all that was certain was that the brain would be the main weak point, and Sinew and Synapses would block areas and damage the brain when broken.
As I set up the brain, I originally thought of it as a maze that lead to the brain, and at the end the player would just mag dump it to end the level. This was an incredibly anti climactic ending, especially once Daniel showed us the eye explosion animation. With that, I returned to the drawing board and started on the platforms you see at the end of the above slideshow.
The new and improved fight had the platforms set up in a way that the player would never have to touch the floor, instead being able to infinitely strafe around the brain. Satisfied with how this played out on its own, I went in again to add additional challenges to the fight.
In the final version, the brain is decently protected by synapses, but they can be destroyed to deal a small amount of damage to the Brain. Its only method of defense is the Phages it will regularly spawn in. They can be destroyed, but deal no damage to the brain, and a fast moving player should be able to avoid their blasts altogether. Once the brain has been dropped to half health, it will spawn in a pair of eyes that will track the player and unleash a rapid-fire barrage of lasers. The eyes have 300 HP each, and once destroyed, deal that much damage to the Brain itself. Gore is turned up to 11 in this level as well, as blood leaks from the ceiling anything that is destroyed results in a shower of gore.
Personally, I would say this was the most fun part to work on, as we got to really cut loose and get into the over the top feeling of the game.
At the end of the project, we had gone slightly over the estimation we set for ourselves, but I believe that it was absolutely for the best, as it meant that we had the extra time to put in and polish up some things that weren't quite up to our standards, as opposed to the base game that was made on an extremely tight schedule. There were quite a few takeaways I had from this, such as better design flow, a reminder that scrapping a scene can be for the best, and a new appreciation for lo-fi art in general.
Overall, I'm very happy with the work I put out, and feel as if I managed to improve on every aspect I was lacking on in the base game levels. The filter we use comes with certain limitations, but working around those was a good design challenge, and while I don't think I got it quite perfect, I certainly feel as if I improved drastically from the start of this game.
Currently, we're finished with planned content for Midnight Ultra and hope to start on a new IP later this year, so if you want to hear about that, keep an eye here or at Forwardinstinct.com for updates!