When "Doing Everything Right" Goes Wrong

I’ve been debating whether to write a postmortem of my recent indie platformer Super Win the Game since its launch a little over a month ago. I don’t feel like its story is complete yet — certainly I hope not — but I think it’s important to commit some of these ideas to writing before I’m too far removed to speak earnestly about them. As postmortem formats go, I’m not a big fan of the bulleted “what went right, what went wrong” paradigm, as the reality of development is often subtler; there can be both good and bad facets on every part of every game, and portraying them as wholly right or wrong overshadows this reality. Instead, I’m going to recount the history of how Super Win the Game came to be, then I’ll take a step back and analyze some of the critical points along that path.

In Spring 2012, I released a platformer called You Have to Win the Game. It was designed to recreate the experiences of the earliest games I played as a small child, from its four-color CGA graphics to its monotone PC speaker audio. I released this game for free on my web site, and it was almost immediately a success, with sites like FreeIndieGam.es and Kotaku picking it up within days. Over the next two years, I continued to make incremental improvements to this game, adding new challenges and support for additional operating systems, and eventually bringing it to Steam, still free of charge.

In January 2014, my brother David and I met up at Steam Dev Days and discussed our current and future projects for our company Minor Key Games. At the time, I was working on an unannounced game, a combination of a village life sim and a dungeon crawler. It was becoming apparent that this concept was perhaps too large in scope, and coupled with the recent frustrations of failing to port my engine to Mac and Linux, I was discouraged with my work. In discussing alternative projects, I was struck with an idea: I could make a sequel to You Have to Win the Game, this time taking on another equally formative element of my gaming history, the NES era. This felt like a perfect combination of a smaller scope, a chance to build on existing technology, and a premise I was excited about. David suggested “Super You Have to Win the Game” as a title. I liked it, but the original was already too much of a mouthful, so I shortened it to Super Win the Game, and a few weeks later, I set to work making that game.

When David released Eldritch in October 2013, he had only announced the game about six weeks prior. We felt like this was a mistake in hindsight, so one of my first goals with Super Win was to announce the game as early as possible. I wanted to avoid a situation in which the announcement fell flat for lack of media assets, so I produced a web site, screenshots, and a teaser trailer to accompany it. This put the impetus on me to develop in a very short time a number of solutions to features to distinguish Super Win from You Have to Win. In a little under two months, I produced assets in the style of NES games for the player character and a number of environments. I implemented a new editor from scratch in C# (a language with which I had previously had almost no familiarity) which could support concepts like larger scrolling rooms. I developed a tool to render MIDI music as faux NES synthesizer audio. I prototyped a title screen. The teaser trailer was largely fake in that levels didn’t exist beyond the parts seen, but this work laid the foundation for the development of the real game that followed.

Almost coincident with this announcement, You Have to Win finally made its way through Steam Greenlight, some sixteen months after I initially launched the campaign. I took a brief hiatus from working on Super Win to create some new content for this release. This turned out to be a good introduction to releasing a game on Steam and understanding the amount of content creation involved in supporting features like achievements and trading cards. (Unfortunately, the trading cards and related assets for that game were not able to ship because these features are not allowed for free games.)

After wrapping the Steam release of You Have to Win, I began working on actual game content for Super Win, building off the lessons I had learned from mocking up scenes for the trailer. I expected this to be a fairly lengthy process of building a large game world a little bit at a time, but (as I began to recognize would be a pattern across the entire development), this ended up being truncated and accelerated in order to meet deadlines.

Besides announcing late, I felt like not having an Eldritch presence at conventions had negatively impacted that game, and I wanted to correct that mistake. PAX seemed outside the realm of possibility for a number of reasons, but I found a number of local and regional events that seemed appropriate venues for demoing Super Win. RTX, SGC, and QuakeCon were all to take place in July, each a weekend apart from the next. If I could have a demo build ready by that time, I could bring it to all three events, get some playtest feedback, and spread the world about the game.

In the meantime, another opportunity arose. A friend was hosting a local showcase of indie games in early June. This would be a good chance to do a test run of the July demos, I figured. The downside was that I needed to have a demo ready to go a month sooner than I expected, and that meant crunching on level design.

I’ll readily admit I’m not much of a level designer. My design process for both of the Win the Game titles has been to block out rooms one at a time with little thought to the larger picture beyond a vague mental image of how I imagine the space will look and feel. Sometimes, when I have sufficient time to test and refine these layouts, this turns out well. In this case, I was pressed for time, and some regions of the game suffered as a result. In the end, I made the decision to omit the final stage and secret world from the demo, in the assumption that few players would reach that content during a demo session.

Development throughout June and July was slow and sporadic, but in August, I commenced regular development again, and I announced a release date of October 1 at that time. The next two months were spent developing the remaining regions that had not been included in the demo, as well as making various changes and improvements based on observation and feedback from playtests.

I released Super Win the Game on October 1 as planned, and per player feedback, I produced a non-Steam version available via the Humble Widget a week or two later. I had not planned for that development, but it turned out to be a good experience insofar as it prompted me to upgrade my code and build pipeline to be able to more easily produce Steam and non-Steam versions side by side.

That’s the quick history of Super Win the Game, told largely in an objective voice and void of emotional matters. Now let’s get into the meat of this piece.

Super Win the Game sold about 200 units in its first day, 700 in its first weeks, and 850 in its first month. As of the time of writing, it’s inching up on 900 units sold. That’s not good. There’s no way to spin that other than complete and total failure. So, how did that happen?

In my mind, Super Win was always a $15 game. I fancied it could sit alongside other games at that price point, games like Braid and Super Meat Boy. I started having some reservations in this regard as the schedule began to accelerate to meet demo deadlines, and in taking the game to events, it became clear that there simply wasn’t as much content as I’d hoped. What I estimated to be a five- or six-hour game was looking closer to a two- or three-hour one, and I wasn’t sure I could price that game at $15. Around that time, I started hearing from players who wanted to be able to buy the soundtrack at launch, and that led me to the decision to launch the core game at $12.99, with an option to get the game and soundtrack tlogether for $14.99. This allowed me to have my $15 game while hopefully not being tied to the expectations that accompanied that price point. I didn’t want to price the base game too much lower, as that would overvalue the soundtrack, and I also wanted to avoid being part of the myriad $9.99 games on Steam. (I’ll admit I had no logical reason to be wary of the $9.99 price point. It was a gut decision made without any data to back it up.)

It’s possible $12.99 was still too high for the game I released. That’s a difficult call to make for a number of reasons. I have some data to suggest players feel it’s too high, but I also have some data that says players felt it was appropriate after playing the game to completion. This raises the question of whether the experience of playing the game markedly exceeds expectations. I don’t have any strong evidence to support that, however, and it feels a little self-congratulatory, so I’m hesitant to make that assumption. It’s also difficult to gauge a more appropriate price point on account of the fact that Super Win has yet to go through any big Steam sales. (The 10% launch discount may have been a small incentive, but I tend to feel like that discount is taken for granted. In any case, the sales were already low enough by the end of this discount period that I could not perceive any drop by virtue of its ending.) I feel like there is a “wait for the sales” mentality in PC gaming today, and my hope is that Super Win may find a long tail at discounted prices. Certainly sales have been good to Eldritch, but I’m wary of making that assumption as well, considering how poor the initial performance was.

So, if not the price, what else might have hurt my sales? I can think of at least five off the top of my head. Early October was a notably busy season for games, and launching a no-name indie game alongside the likes of Shadow of Mordor and Alien: Isolation certainly did me no favors. The indie platformer genre feels more than a little passé at this point, and it’s entirely possible many potential customers overlooked it simply by virtue of its genre. On top of this, the game lacks an easil identifiable gameplay hook; its strengths lie in its aesthetic design and solid game feel, and maybe that’s not enough. I also can’t ignore that Super Win the Game is a silly name at best and an objectively bad one at worst, and that may have hurt it. Lastly, for all of these reasons and probably more besides, the game received little to no attention from the press, meaning it launched with low awareness.

So that’s the sobering boat I’m in now. The game’s dead in the water, with only the chance of future sales to revive it. But let’s set that aside for a moment and see what lessons can be learned from this entire process.

I think my first mistake was to assume that because You Have to Win the Game had been a hit, players would necessarily care about (and want to pay money for) a sequel. I tend to think of You Have to Win first and foremost as an authentically retro platformer, but maybe in the public’s mind, the most interesting thing about it was that it was free. I ignored that possibility, and as a result, I was surprised and hurt when Super Win wasn’t greeted with the same enthusiasm. Especially telling was that, despite having announced the game several months earlier, many of the players I met at events were familiar with the original but had no idea a sequel was in development.

In discussing the game’s reception with David before and after release, he mentioned that I was playing it too close to the chest, holding back information that could have attracted attention from both players and press. Maybe that’s true. There’s a darker, more personal side of Super Win seen in the interstitial dream sequences that I’ve deliberately avoided talking about. Partly I avoided talking about these bits because I wanted them to be surprising for first-time players. Partly it was because I didn’t want to oversell those aspects. Partly it was because some of those elements are personal and emotionally affecting and I wasn’t comfortable talking about them. I can’t say with any certainty whether this was a mistake. I’ve heard very little feedback, positive or negative, on those parts of the game. It’s possible they fell flat and left players confused. That was occasionally the impression I got from playtesters, for sure, but I made no attempt to course correct.

Along the same lines, it may have been a mistake to set an embargo on reviews and related coverage (streaming, let’s plays, longplays, and the like). I don’t have enough info to support this one way or the other, but certainly it didn’t have the intended effect. My hope was that, if I could send out a large number of builds to reviewers and YouTubers with a launch day embargo, I could generate enough noise at a time when the game were available for purchase that I could cut through the business of the big name releases that week. Obviously, that didn’t happen. I don’t have any evidence to suggest the embargo necessarily turned anyone off of the game, but it doesn’t feel like it helped in any way, and it was probably a silly and headstrong thing to do for a tiny little indie game.

So what does this mean for the future? Well, I still have a little bit left in savings, probably not enough to finish the development of a second game in itself, but if Super Win performs well over the sales season, it could hold out. I’ve soft-announced my next game, Gunmetal Arcadia, not with the theatrics and bombast I attempted on Super Win, but with an understated development blog (http://gunmetalarcadia.com/). This decision came about after seeing reactions to David’s blog on his next game, Neon Struct. It was telling (not to mention emotionally crushing) that the two biggest pieces of press Super Win ever got — previews in Polygon and PC Gamer — came about not in response to the multiple press releases I mailed out, but because editors for those sites contacted David about Neon Struct, and he had to redirect them to me because my game was coming out sooner. Ouch.

I have a note here about “trying to do everything right.” I haven’t found a good way to fit it into this piece yet, so I guess I’ll just wedge it in here awkwardly. Maybe this is an issue of pride. Maybe it’s about avoiding confrontations. For whatever reason, I spent a disproportionate amount of time and money trying to do things that I felt were necessary to be a successful indie game in 2014. These included sim-shipping on Mac and Linux, supporting Steam achievements and trading cards, providing better Unicode support so that the game could be translated into Russian, and so on. To date, I’ve sold about fifty copies on Linux, which is a drop in the bucket next to what it cost me financially and emotionally to support that platform. I wonder if my experience would have been better if I had forgone these features in favor of making a stronger game exclusively in an environment in which I were comfortable. In a world where Super Win was a smash hit and players were clamoring for ports and additional language support, maybe it would have made sense, but doing that work upfront is hard to justify in retrospect. I haven’t recouped those costs, and it’s questionable whether I will. Where possible, I’ve made these changes across my engine as a whole and not just this one game, so future games can benefit from this work as well, but exactly how beneficial that proves to be hinges in large part on whether I am able to make any future games.

I feel like I’ve been leaning a little heavy on the doom and gloom, which I hope is understandable in my position. I was fairly depressed for a week or two after launch in light of the poor sales, and I still have serious doubts about the future of Super Win, Gunmetal Arcadia, and my role in Minor Key Games. But it hasn’t all been bad. This was the first time I made a game from scratch and sold it, and I learned a lot from it. I’m still learning from it. Notably, this was the first game I’ve made a serious attempt to market. Some of this was abject failure (I simply have no idea how to court press), but some of it went well. Despite the high stress associated with being a socially-anxious introvert trying to demo a game publicly and often entirely by myself, I actually really enjoyed those experiences, and I will almost certainly take Gunmetal Arcadia on the road next year too. I produced a fairly large quanity of free swag for these events, which was a big hit (I’m noticing that trend with things I give away for free), but I am nowhere close to recouping those costs, so probably I won’t do that again.

I’m not sure whether it makes sense to demo an unfinished game again. Personally, I feel like the improvements I made by watching people playtest the game were huge, but I can’t realistically say those improvements had any positive effect on sales. Taking the game to conventions was good for the quality of the game, and it may have had some effect in increasing awareness, but It was also a huge drain on my schedule, finances, and emotional state. I wonder whether taking a finished and released game to events would be less of a drain. There’s no reason I couldn’t still make post-launch improvements based on this feedback, and I wouldn’t be interrupting development to do so.

Lastly, I’ve been tracking some stats since launch which might be of use to other developers. In their first month report on Shovel Knight (http://yachtclubgames.com/2014/08/sales-one-month/), Yacht Club Games mentioned that preorders predict first week sales by 200-400%. I didn’t run a preorder campaign for Super Win, but Steam does report wishlist numbers at any given time. Wishlisting a game isn’t the same as a preorder; it’s not a promise of a purchase, but as an expression of interest, it’s the next closest thing I had. Just prior to launch, Super Win had 2,457 wishlists, and it sold 694 units in the first week, which suggests wishlist numbers predict sales by around 28%. Sales for the first month were 122% of the first week (846 total), and for what it’s worth, about 22% of players opted for the slightly pricier deluxe edition, which included the soundtrack. I should note that these numbers are from Steam only, but Humble sales are negligible; that version was announced late and with little fanfare, and it only sold two units in the first month.

So that's where Super Win the Game is today. It’s been a strange, rushed development and a rough launch, but I guess that’s how it goes. I made a game and sold it, and that’s probably as good a first step as I can hope for.