Stage Presence: A No-Hit Wonder
Note: This article is made up of two sections; the history of the Stage Presence’s development and a brief post-mortem. If you want to skip to the gory part, scroll to the ‘Launch’ section roughly 3/4 of the way down.
Introduction
After a few years on-and-off work, I released my first indie game in 2017. It’s been a strange, anti-climactic experience that I wanted to get down on ‘paper’.
Here’s the elevator pitch from the Steam page:
Stage Presence is the anti-guitar hero. A comedy-horror-karaoke game where you need to use your voice and wits to win over a fickle mob!
And here’s the launch trailer which does a good job of capturing the tone and mood of the game:
Pre-Production
In the summer of 2013, while watching the Glastonbury footage on the BBC that year, I felt my annual nostalgia for being in a band and quiet jealousy that I’d never gotten the chance to play in front of thousands of people at a festival. Imagine how cool that would be! Hold on, I’ve got a VR headset and a Unity Pro trial!
Initially Stage Presence started as serious “academic” “research” into the idea of ‘presence’. I wanted to see if you could make people feel stage fright by presenting a virtual crowd to them — would they be convinced it was real and get scared? That failed; the rendering and animation requirements to make a completely convincing crowd in VR were impossible with a copy of Unity and a 2013 spec PC. Everything looked cartoony and video-gamey so I decided to play up to that.
As a failed musician, I have regular performance anxiety nightmares (I’ve forgotten all the songs / we haven’t rehearsed / I’ve got to play an unfamiliar instrument) which is where the basic concept for the game came from. I mined the legendary tale of Daphne & Celeste at Reading being pelted with bottles of piss as the model for the game’s fickle mob.
At the time, all anyone seemed to be making in VR was endless jump-scare horror games — I liked the idea of making something that was a non-traditional “horror” and blending it with humour.
My model for this was Typing of the Dead where players had to type random phrases to kill zombies. In boss encounters, the game throws funny non-sequiturs at the player and the challenge becomes trying to type while laughing. This blew my mind as a game mechanic when I first played it as a teenager — using humour against the player so that it becomes their enemy! What an idea! I aimed to make Stage Presence a similar experience — where the player needs to concentrate on entertaining the crowd but with a constant stream of weird, funny and mildly scary game elements distracting them.
Development
Oculus DK1
The early versions of Stage Presence were dictated by the limitations of the hardware. At the time the Oculus DK1 didn’t allow for “positional tracking” meaning that players were rooted to the spot. As a sitting duck, the crowd could pelt you with bottles and shine lasers in your eyes with abandon. This created the atmosphere I’d initially been aiming for; a feeling of dis-empowerment and stage fright.
Another limitation the game had to adapt to was input. Gamepads are clunky in VR and practical tracked input controllers hadn’t yet been invented so the game was left to rely on microphone input alone. Unlike today, voice recognition technology wasn’t particularly robust so the game simply measured the pitch and volume of the players voice and made the crowd react accordingly. Designing around this limitation I made the crowd grow bored of samey input quickly, (theoretically) pushing the player to quickly progress from authentic singing to panicked babbling.
One of the big challenges of early VR was figuring out how to do UI correctly. Traditional 2D HUD elements sucked for various reasons so VR naturally pushed the game towards diegetic UI. I put the main UI on giant ‘jumbotrons’ in front and behind the player — these felt like natural parts of the world and served the same function as traditional UI.
Audio feedback from the band’s manager provides another opportunity for VR friendly UI and giving the game more of it’s character through humour. I roped long-time friend and one-time video game writer James Duffy into co-writing a harassed sounding band manager character and wrangled a free VO session from the Voice Acting Power Squad.
This early version had the right ‘feel’ but wasn’t exactly ‘fun’ and didn’t offer much player agency. The players ability to affect the game’s progress and world felt limited and ill-defined. Still, I’d roughly achieved the idea I’d had and released the prototype for free on the Oculus forums.
I was done with this daft idea and wanted to go off and play with something new. I’d never look at this dumb project again, right?
Wrong.
On a whim I’d also sent some screenshots to Rock Paper Shotgun and much to my surprise they actually wrote an article about it. Then, even more surprisingly, the prototype got some spontaneous YouTube attention from Hank Green and KSic wracking up 10s of thousands of views.
I decided it might be worth carrying on developing this daft idea.
Oculus DK2
After a few months of this unexpected attention I opened the project again and decided to see how far I could push it.
In the intervening time the new version of the Oculus, the DK2, had arrived and my skills in Unity had improved. The Oculus DK2 introduced a literal new dimension to VR — the ability to physically move around the scene (albeit in a very, very small volume).
This technical change had a profound impact on the fundamental design of the game; where I’d wanted players to feel vulnerable targets they were now able to move. When the angry crowd throw bottles at them or shined lasers at them, the player could now dodge out of the way. The player was now less of a victim and becoming empowered. I was sure this would ruin my perfect design and the game would be less fun.
Wrong.
If anything ‘breaking’ the initial design like this made the game much more enjoyable. What I failed to learn at this point (and what would become a recurring theme) was that VR was still a rapidly developing technology and I’d have to keep altering the design of the game to keep up with it.
I was now starting to think of Stage Presence as an actual game instead of a prototype. With vague ideas of selling the game I dedicated some meaningful effort to making it not look incredibly hideous. While it remained a profoundly ugly game, it at least looked somewhat like an actual game.
Feeling like I’d hit the limit of all I could do with the project in VR, I stuck it on Itch.io for a few quid. I was finally done with this silly project. Again.
Greenlight and Bundles
With the DK2 version of Stage Presence starting to feel sort-of fun and (a few) people actually buying the game on Itch I decided to take the £70 gamble of putting the game through Steam Greenlight. The game was basically finished so I could pass Greenlight, stick it on Steam and make a few extra bucks in a couple of months right?
Wrong.
People liked the concept and the little bit of gameplay I’d shown but a phrase kept cropping up in their feedback: multiplayer. Desperate to please (and pass Greenlight) I declared the game would have multiplayer without really thinking about it. Multiplayer would come back to haunt me; barely anyone will ever get to play it (the game rounds are too short) yet it absorbed a huge amount of development time and energy that could have been spent on making the core game experience better. All it really served to do was add a tickbox on the Steam page and some more footage for the launch trailer. Conceptually the multiplayer is cool but I doubt if anyone will ever actually have the chance to play it.
To try and get more eyes on the Greenlight and make some short term cash to fund development I enrolled Stage Presence in two game bundles. This was a spectacularly bad idea which I talk about more in the ‘What Went Wrong’ section of the article. Don’t do this. It was stupid.
As well as screwing myself over with Greenlight game bundles, I made a gameplay trailer (that was a hellish afternoon) and sent out loads of ham-fisted “press releases” to the games press. I managed to score another article on Rock Paper Shotgun one on Eurogamer and a really nice write up on Kill Screen. Actual, proper press sites were talking about my game! I even managed to sell a few copies on Itch off the back of it!
The press attention was great but sending out the “press releases” and cutting together promotional videos was stressful and time consuming. A massive amount of the time I should have been spending on the game was being spent on the Greenlight campaign and PR. I vowed I’d pay professionals to do my trailers and PR in future, it’d be worth the money. What I needed was a publisher.
tinyBuild (AKA holy shit, I’ve got a publisher now!)
It was also around this time (Valentines Day 2016, very romantic) I sent an email to surprisingly-tall-in-real-life Mike Rose. Mike had a reputation for knowing how to market a game so I sent him some gifs of the game and asked him what sweet marketing tricks I could use to sell a billion copies of the game. Mike, having the same stupid sense of humour as me, liked the game and instead of giving me hot marketing tips he persuaded tinyBuild to sign my game.
Holy shit!
This was ace — me and Mike were emailing each other daily with ideas for the game (Mike is hugely responsible for the game going down the ‘make it weird’ path which gave it a lot of it’s character and identity). The scope of the game started to expand and turn it into a something people would actually pay for and soon it was decided tinyBuild were going to announce it at PAX East 2016 live on the Twitch stream.
Fucking hell!
I ordered up a quite-expensive-but-worth-it announcement trailer from the super talented Marlon Wiebe, added a new level to the game in super record time and gave Mike my Oculus CV1 devkit to take to PAX. I watched the game get shown off live on the Twitch stream and was amazed to see how effective the press release was; dozens of sites picked it up and posted about “tinyBuild’s first VR game”. Mike even swung us an interview with PocketGamer which was a super ‘real’ dev thing to be doing.
I’d made it — I was a proper indie dev now and with tinyBuild’s publisher clout I was definitely going to make a billion dollars.
Then Mike left tinyBuild.
HTC Vive
Just before PAX East 2016, Valve sent me a Vive Pre devkit. I’d been planning to support the Vive post-launch because hey, it’s just another HMD and it’s not going to be better than the Rift, right?
Wrong. Tracked controllers changed everything. AGAIN.
Suddenly I had the ability to interact with the world I was in and it was super compelling. Except in Stage Presence I couldn’t do anything and that was a problem. The bottles the crowd threw at you afford being picked up.
The game had been designed around the idea of the player being vulnerable and dis-empowered — as with the freedom to move introduced by the Oculus DK2 and CV1 I once again pivoted away from that. Now not only could the player dodge out of the way of a bottle, they could pick it up and throw it back at the crowd. The player was no longer a victim, they could give as good as they got.
This was a fundamental change to the original design of the game but I’d have been insane to ignore it. Not only did the player expect to be able to interact with the world it, it made the game more fun.
I was staring down the prospect of writing a whole VR interaction system and adding months more development time when I stumbled across VRTK. This was one of the best things that happened to Stage Presence. In less than a day I had basic hand tracking in the game. In less than a week I had new gameplay centered around hand tracking. 6 months later and it had become a focus with the player unlocking hand tracked game items to help them win over the crowd (mystery pills they can dose on stage for random positive or negative buffs, the BASS DROP you physically drop for a dubstep fueled score boost, a tennis racket for defending against bottles and glowsticks and tonnes more).
Very quickly Stage Presence became a room-scale, interaction based VR game with Vive as the lead platform. VRTK was critical in that happening — if I’d had to write the interaction system from scratch either I wouldn’t have bothered or I’d still be working on it now.
Launch
As the game was getting close to launch I ordered another quite-expensive-but-worth-it trailer from the super talented Marlon Wiebe. That’s the one at the top of the page. It was super fun making the tools for it so that Marlon could get the shots he wanted, he got the tone of the game down perfectly and made it seem super-slick and ‘real’. I can’t recommend him enough.
I also managed to rope in a bunch of friends and family to compose music for the game. Probably the most exciting part of making the game was when Steve Durose, guitarist in my all-time-favourite band Oceansize, agreed to write a song for the game. A literal dream come true.
As one of the final touches, I paid handsomely to have the game localised into 9 (!) other languages (“Karaoke is big in Asia right? Let’s translate into Japanese, Chinese and Korean!”) – pricey and time consuming to setup but it was bound to score some big sales.
Finally, after 3 years of work launch day rolled around. I was variously excited, terrified and relaxed. Only weeks before I’d lost my job and had been working on Stage Presence full time in the run up to release. Losing the job was fine — I was about to become a successful indie developer, right?
Wrong.
tinyBuild launched the game and I waited expectantly for the flood of press coverage and cash I was definitely about to get. Instead I got a flood of users telling me the game was buggy and confusing and a vanishingly small number of sales. According to SteamSpy, Stage Presence is currently tinyBuild’s least successful game by quite some margin. Take into account that a bunch of the ‘owners’ SteamSpy lists are YouTube / Twitch streamers tinyBuild sent free copies to for promo purposes. Another bunch are people who got the game effectively for free in the two bundles I put Stage Presence in during the greenlight campaign. The real figure of actual sales is much lower.
It seems unlikely that Stage Presence is going to make back it’s development costs.
Weirdly, we got almost no press coverage on the launch itself. tinyBuild had sent keys out to streamers a week before launch and had gotten a bunch of cool videos (and some useful feedback) but on launch day? Nothing. Other than a couple of (lukewarm) reviews from the keys sent out the previous week, the only launch article that popped up was another Rock Paper Shotgun article after I mailed them with the info (based on them publishing two articles for me in the past).
I spent the following week frantically patching bugs and adding new features players had expected. Negative Steam Reviews starting popping up and the game dropped quickly from “Mostly Positive” to “Mixed”. One surprising (and extremely heartening) element was the discussion forums. I had several players post that they were disappointed with the game and listing thoughtful, constructive feedback. I talked to these players and quickly patched the game based on their feedback. Despite being initially disappointed with game, several of these players then left positive Steam reviews and even went on to defend me and the game in comments on the negative reviews.
That bit of kindness made me get a little bit teary eyed during post-launch blues.
What Went Wrong?
Pretty quickly after launch I could see the game had been a semi-disaster — at the very least it was a critically lukewarm commercial flop.
Why?
I’m still figuring things out but there were a bunch of factors involved. The major points are:
- The game was buggy.
- The game was confusing.
- The game is niche, on a niche platform.
- Adding a tacked-on desktop mode to a VR game was a mistake.
- The game got (almost) no press.
- Pre-release bundles hurt sales (duh).
- The game wasn’t actually particularly good.
Most of the responsibility for all this falls to me. Sure, I didn’t have the resources of a full QA team or professional playtest analysis but I didn’t do as much on these points as I should have.
I didn’t spend enough time bug testing and the streamers tinyBuild sent keys to were finding massive bugs a week before launch. Really I should have built in (at least) an extra month on at the end of dev purely for QA.
Better playtesting would have revealed that the lack of tutorials — a conscious decision I thought would increase the feeling of ‘stage fright’ and persuade users to experiment to discover how to player the game — was in reality just confusing to players. They fired it up, weren’t told what they were supposed to do and then were left to fail the game.
Stage Presence is a weird game and really only truly works in the context of a drunken party. The whole point of the game is it encourages you to make increasingly weird noises with your voice and embarrass yourself — it’s hardly surprising this didn’t translate into massive sales. This was compounded by being on a small platform (VR HMD sales ain’t huge, kids) with a tacked on desktop mode.
Talking of that desktop mode — that was another mistake. Stage Presence only really makes sense in VR where you feel like you’re actually surrounded by a hostile crowd. It’s experiential as much as it is a game. In an attempt to try and overcome the small VR user base I added the desktop mode. This resulted in more bugs in the game, less time for me to focus on the core experience and bad player feedback. Nearly all of the negative reviews on Steam currently are from desktop mode players. Nearly all of the positive reviews are from VR players.
The lack of launch day press, as mentioned in the last section, was surprising. When we had announced the game in 2016, it got loads of articles and the press release appeared on all the game PR websites. When we launched the game, nothing appeared. Gamepress.com shows no press release for the game on the 28th of Feb. I can’t explain why the press release didn’t get picked up (even by press release aggregators like Gamepress and Gamasutra); tinyBuild think it got buried by GDC news. The result was that we had no print press apart from an article Rock Paper Shotgun posted based on an email I personally sent them on launch day (albeit the article landed a few days later). The irony.
To tinyBuild’s credit, they did get a bunch of streamers playing the game the week before launch (although as far as I can tell, this didn’t have much impact on sales).
Aftermath
So what now? During development I’d made grand plans to support Stage Presence for at least 6 months to a year with new content. I’d planned a bunch of free content (new levels, new unlockable items) and a big bit of paid DLC that completely transformed the game into something else.
None of that happened. The game is unlikely to ever make back it’s (relatively small) development costs — making free content and paid DLC for a game no one is playing would only put me deeper into the hole.
We’d also been planning to port Stage Presence to other platforms but again that seems unlikely given the poor sales on the biggest marketplace (Steam).
What Stage Presence has been successful in is teaching me a lot about the reality of indie development.
The most positive (and surprising) lesson came from responding to Community posts about bugs. Several times, people left reports of (sometimes major) bugs or issues with the game; after replying to them and fixing the bugs for them they frequently left the most positive Steam reviews for the game. It was a real eye opener to see how being an approachable dev and fixing peoples problems could turn the most disappointed players into evangelists.
Another memorable lesson has been that people will have experiences with the game completely counter to what I intend or design; one Chinese user left a Steam review saying how the game had alleviated their depression and helped their anxiety of singing in public. In a game I’d designed to torture people with stage fright. Knowing I’d had this (unintentional) positive effect on someone was heartwarming and opened me up again to listening to feedback and changing my ideas based on how people actually experience my games.
I learned over and over that my assumptions about the game were wrong — I should have rolled that into doing significantly more playtesting.
I’ve learned to calibrate my expectations and be more realistic about what ‘success’ looks like. Am I ever going to get rich from indie games? No. Am I ever going to be able to work full-time as an indie? Maybe. But don’t expect it to happen overnight and from one game. I re-read this blog post and this blog post by Cliff Harris a lot in the months after launch. My next few games are going to have smaller budgets, smaller development lengths and be much more focused on ‘finding the fun’.
I definitely won’t be bundling my games before launch. The two bundles Stage Presence was a part of made me <£500 and meant I had to give out ~5,000 keys for the game. Not a good return on investment.
Would I work with a publisher again? Maybe — it’d need to be the right publisher with the right deal on the right project; I’m certainly more wary about the idea of working with a publisher than I was before Stage Presence. tinyBuild opened doors to some opportunities I wouldn’t have had otherwise; my game was at PAX, my game was the on the Twitch channel livestream, my game got fresh twitter attention, my game got sent to a bunch of YouTube streamers. This is all stuff I couldn’t have done on my own. The only problem is I don’t think any of it really amounted to much in terms of sales.
So what’s next? In the months after Stage Presence I worked on a bunch of small prototypes (here, here and here). I’ve learned to know when to cut my losses and several of the prototypes were ditched early when I couldn’t “find the fun”. I’ve definitely learned to ‘fail fast’ after staying with Stage Presence too long after the idea revealed it didn’t have any real legs. For all my future projects I’m going to keep the scope small, make sure it’s fun, test it to death and release it. And maybe it’ll make it’s money back.