GearVR Jam 2015 Postmortem
The immediate goal was to win a contest (duh), with secondary goals to have a game framework to further develop for release and explore what was fun about VR. I settled on a first-person grappling mechanic. I had an inkling it would be fun, but I knew I had to experiment to see just how it would work. Once the grappling locomotion was fine tuned, the other features kind of snowballed into a big mess of stuff that isn’t quite balanced yet. With more time, I would have liked to focus more on what was fun (or not) about each one like I did with the grappling, but time was an especially precious resource and at some point I just kept saying “yep, ship it, next feature.”
WHAT WENT RIGHT
1. All desired features were implemented.
In other projects, I’ve often cut “non-essential” features, but I felt that this game concept couldn’t be properly articulated without EVERY feature in there. Amazingly, they all made it in (almost):
• Grappling locomotion and UI (feedback on grapple state)
• Objective info reticles
• Intro cinematic scene
• Quest system
• Resource mining/upgrade system
• Combat mode w/enemy AI, upgradeable weapons/shields
• All 3 quests, including turret minigame and “level completed” sequence
• Last-minute voiceovers and character avatars
• All SFX
This resulted in a full half hour of gameplay with several different challenges to master. Though the mechanics need tweaking, the core of the game is very approachable for casual players. The one missing element was a more thoughtful tutorial.
2. I learned a whole lot about rendering optimization.
I wish there were a single silver bullet to tackle this stuff, but there’s not. It was all about finding a balance between techniques. Some thoughts:
• I’d want to use shading “prebaked” in the textures, rather than lightmapping to save on memory.
• Definitely no pixel lights or dynamic shadows.
• Creating multiple LODs decreases polys but doubles or triples batches on those objects; so the jury’s out on the best approach. I went with a single LOD + culled distance.
• Use stencil shaders for alpha HUD when possible (overdraw is greatly reduced).
• Atlas textures for batching.
• Combine meshes only if they will share screenspace often. Otherwise, the unnecessary poly rendering will be worse than the saved batches.
3. I learned a lot about how to design for VR.
The biggest points, aside from performance optimization, were avoiding motion sickness by locking player rotation and easing in/out of movement. Other considerations (that I’m still ironing out) are how to best achieve a HUD/crosshair. Currently, by drawing UI to a render texture and passing to a dome with a stencil test shader, a fairly comfortable HUD is displayed (with only 1 extra batch per camera and limited overdraw). Lerping the HUD and crosshairs also helps smooth out short drops in framerate (but it won’t help sustained sub-60Hz moments). I’m very pleased with the accessibility of the game, as grappling around in zero gravity has the potential to be the most nauseating thing ever.
4. I finished!
Going into the jam, I knew there wouldn’t be very many eligible projects by the end due to the lack of available hardware and the skill and effort involved in finishing an entry. (This, compared to the 3 million phone apps of competition I’m used to). Making it through with a project I’m proud of is most definitely a “right.”
5. Portfolio content, demo for future launch.
At the very least, I have some VR content for a portfolio, and a starting point for a potential store launch. Whether or not I wait until a GearVR market’s there is my current dilemma. There’s a risk in putting even more effort into turning this into a full-fledged product, but it may be worth it to be in the store for the first batch of GearVR consumers.
WHAT WENT WRONG
1. Time crunch at end left little space for getting my “marketing” together.
By this I mean that I knew there was an initial round of judging based on 2D assets only, and I should have devoted more time to calling out the best features of the game rather than do an eleventh-hour screen capture.
2. No artist.
I spent a lot of time optimizing 3rd-party assets that weren’t necessarily built for the tight performance constraints of the GearVR pipeline. The overall look was also limited to what was available, and very little creative direction was achieved due to a lack of access to custom art.
3. Limited “shelf appeal”
I’m a little concerned that the concept and/or project assets don’t seem to engage people’s curiosity. Certainly this could be related to the fact that there was no artist involved, but I also have a project that I abandoned in the 3rd week (a 1st-person jaywalking game) that got about 4x the video views and positive feedback as Galaxy Grapple. That game’s concept certainly seemed to pique interest a lot more than this one despite it having almost zero replay value after the first 2 minutes of (fun) novelty. A challenge in the next stage of development will be to figure out how to grab people the moment they hear about it. (Better name? Better art? Better tagline?).
4. I worked too hard
I was lucky to have a lot of time to put into this project (no day job lately). I did take a couple weeks in the middle to do some music stuff, but that was hard to enjoy with this deadline in the back of my mind. I also (like many others) didn’t sleep enough. The result was that I created a feature-rich project, but the cost was my sanity. I’m not sure I could have done things any differently, but now I need to find a way to narrow the focus of the project in a way that VR early adopters are going to want to experience.
5. Not a finalist
Obviously failing to make it through the first round of cuts is disappointing, but I have to refocus on the consumer market now and hope that VR game sales potential rises as hardware sales enter the market. Hard work does not equal success, validation, or money, but it’s definitely a prerequisite. That’s a pill I thought I already swallowed, but I guess I needed a booster.
1. Get an artist
Like, yesterday. I would never ship with stock art given the quality of work other people are putting out (still drooling over SteamCrewVR’s handpainted submarine concept art).
2. Decide on approach
I’m not sure the lighthearted characters and UI are the right feel for this game (though they might be). I need feedback on whether or not a “hard sci-fi” feel would be better, or if I should stick with the current feel and just plug in better art and voiceovers (be honest, were the VOs an instant turnoff?).
3. Fine-tune mechanics
Many of the systems are in place for a game with large scope, but I need to reassess what’s most fun about playing and modify what’s not. Mining, for example, plays a small role in a demo quest, but has potential to be a crucial part of the main loop if implemented correctly. Similarly, the combat is fun enough but would be repetitive as-is. I could add items or skill trees like many RPGs, but I like the simplicity of a single-weapon combat system.
4. Add content
I’d want to have at least 90 minutes of gameplay available. This would mean 2 more “sectors” of quests, characters, and minigames that make new use of the grappling mechanic. Current ideas include gravity wells, depth charges, races, maybe some kind of time trial.
This jam was a great idea. Post-acquisition Oculus throwing some money into the ring to get content before their hardware launch was a great solution to the chicken and egg situation VR has. Having a “not unsubstantial” amount money at stake, I would have liked the first-round judges to have considered execution and innovation equally (since it isn’t a “brainstorm jam,” right? I have a game that physically teleports your real life body to Mars, but that feature’s rolling out post-jam, pending the solving of some fundamental quantum physics constraints). Minor frustration aside, I really commend their desire to incentivize new ideas to accompany this new input/display form factor. In the next few months I hope to see these projects turn my headset into a real piece of entertainment hardware instead of just a novelty. In reparation for not being selected as a finalist, I will accept a launch day store feature. Just send me the paperwork.