Arges Systems is a micro-studio doing game consulting and application development in Unity – their specialty is on game logic and AI. Arges Systems has been partnering with companies specializing in 3D graphics and visual design, as well as doing contracting for companies building games with Unity. In this article, Ricardo J. Méndez (founder of Arges Systems) shares some insights on their just-released Hairy Tales.
I initially founded Arges Systems to take advantage of my experience running remote teams and projects. We had been doing contracting on various game projects for a couple of years before I decided to switch gears and start working on our own stuff. I had just decided to pull the plug on a turn-based strategy game for which I realized the scope was too ambitious and was chatting with Yuriy Mazurkin, our concept artist, about possible themes for a follow up game. The conversation drifted to Russian illustrators – somehow we ended up talking about Ivan Bilibin and it got me thinking about action/adventure games.
A sword-wielding octogenarian riding a warhorse charges forward from a hill. Nordic forests and evergreen trees spread before him. He’s also butt-naked.
It was a simple, straightforward picture by Russian illustrator Ivan Bilibin, from his Marya Morevna series, and my first encounter with Koschei the Deathless. He’s an archetypal antagonist of Slavic folklore, an even more evil version of their better-known Baba Yaga.
I took one look at it and with its combination of adventure and absurd, I thought “damn, this would make for an excellent slavic Zelda-like game”.
You can’t get there… from here
I experimented with gameplay styles while Yuriy drafted some concepts. We drafted several ideas for scenarios, including a story, but it quickly became apparent I had nothing new to say about the adventure genre. I wanted to do a game that was different, and this wasn’t it.
Partly as a way to cleanse our creative palate, I started experimenting with mixing puzzles in. You had to maneuver the main character through a corrupted land, frozen in time, but elements got un-frozen as you approach them. This had pros and cons, since you could activate machinery just by being near it, but enemies also came alive. The puzzle aspect was to figure out what to do when. I decided to discard this version as well. It seemed like a one-trick pony, with the sort of read-the-designer’s-mind approach that I hate, and lacked replayability – once you know the solution, that’s it. Also, the game was taking on a somewhat stoic tone that I felt dragged it down.
You will notice I haven’t mentioned anything about the modeling side of things. We had overenthusiastically already started modeling before I was done with the design, because I wanted to get the time-consuming assets out of the way – or at least properly estimated. Despite it being a bad strategic decision, it had a positive side effect: it made me realize early on that the 3D artist we were working with just wouldn’t cut it. The quality of his work had been in decline, he wasn’t paying attention to details, and both Yuriy and I kept having to bounce work back to him with notes. Eventually I had to let him go and start looking for a new hire. Fortunately we found Ash Barnard, from the UK, who meshed with the team perfectly. He has an eye for detail, very expressive animations and more importantly, just the right sense of humor to make the Hairy Tales animations memorable and peculiar. Ash also brought in a good eye for gameplay, and helped criticize mechanics as I was coming up with them.
Through several experiments and iterations, I ended up landing on something close to the current approach. The first few iterations were fixed stages, based around arrows that directed them and fences that made them turn two sides to the right. It also featured a first draft of the spreading corruption, with the twist that if it spread to a tile with a fence, then it got corrupted and the fence turned into a deadly wall of flame. I can hear the thought gears as you try to figure it out. Playtesters weren’t getting it, and even when they figured out stages the reaction often was “I know this is how it’s supposed to be, but don’t know why”.
That would not do.
Dragging it there
I started paring down the elements. At this point we’ve been in a production iteration limbo for months, and all the associated hair pulling is starting to take its toll on me. Everything is self-funded, so Arges is hemorrhaging cash while we experiment, and my focus is split between the game and the client work that is funding the process. I started trying different games to relax – mostly playing demos, so I didn’t get too involved and lost track of the project. One day I was playing a demo of Atlus’ Catherine, moving Vincent around, pulling and pushing blocks into place, and then a light bulb went on. After simplifying the elements, the stages had felt too straightforward, and the new levels depended mostly on size for their complexity. What if players could drag tiles from one place to another?
I didn’t tell the team, just sent them a build where some later stages required them to move tiles. They were rather surprised at first, but immediately saw the possibilities, like tiles that drag Hairys from one place to another, weapons you can re-use or teleporters. So finally, after months of iterations, we had a design we were happy with.
The 90-90 rule
It took a lot to get from a game’s design to the final product, of course. We still had to design the look for the various tile elements as we were going, which kept Yuriy involved while Ash created the models and I both coded the behavior and came up with the stages. Yuriy was also helping with the texturing. His true love is painterly work, however, so he came to me when we were about to enter the final stretch and brought up that he wanted to move on.
As sad as this made me, since I enjoyed working with him, I helped coach him for interviews and gave him a sterling recommendation. He ended up getting employed by Yager in Berlin, who recently published Spec Ops: The Line, and I expect is right now working on their next project.
At about this time I brought on board composer Levan Iordanishvili as a contractor to work on the game’s music. He liked the game and offered to take care of the sound effects as well. To ensure both were cut from the same cloth – he did a smashing job of re-creating the sound that Ash’s animations made in your head when you looked at them, and his scoring of the three worlds and bosses was top notch.
I had initially planned to release with five worlds and five bosses worth of content, for a total of 75 levels (15 per world). Playtesting had demonstrated that players needed a gentler level progression than the breakneck pace we initially had, so each world had increased to 24 stages. If we kept the same number of stages per world, we were looking at 120 stages total, plus the extra time it would take for the two other bosses and possibly new enemies to keep things lively. The scope was getting out of hand.
I made a judgment call. We would be launching with 72 levels and three bosses, using some minor characters as mini-bosses. Once we saw the initial reactions to these levels from our players, we could release a couple more worlds as add-ons and expand on those qualities that players enjoyed the most. The team agreed, and we geared up for polishing the worlds we had fully designed.
The initial stage sequence introduced one concept after another, presenting a more concentrated experience which gave the player little respite, with no stages that they could use to experiment with the mechanics they had just learned before throwing a new set of concepts at them. After various rounds of playtesting, I introduced some intermediate levels that presented the concepts they’d just learned in different contexts, so that they could play around a bit more, which made the initial learning process smoother. However, it also led to the initial stage sequence feeling a bit drawn out, so I then had to adjust the sequence once again. This process went on over several iterations, even after we had released.
Launch and everything after
We wrapped what we considered to be version 1.0, went back to talk to some publishers we’d been in contact with, and settled on Forest Moon Games and BAM! The game was out the gate.
It was exceedingly well received by reviewers – sites like TouchArcade, EuroGamer and Gamezbo gave it glowing reviews, praising its flexibility, difficulty and character. The game is currently sitting at a Metascore of 81, the highest Forest Moon Games has gotten (and one of the highest of its less-experimental sister brand Crescent Moon Games). Apple picked the Mac version of the game as an Editor’s Choice on the Mac App Store, and gave it a humbler New and Noteworthy feature on iOS.
However, commercial reception was merely lukewarm. We were aware that the characters, not being your traditional cute-and-cuddly puzzle game stars, would be an issue. But we were not sure what was the main problem. The initial game difficulty was rather high – a throwback to the old school puzzle style – which might be turning off casual game players who get it expecting an easier time and hurting word of mouth. At the same time, the visuals are cartoonish, which leads players who would appreciate the challenge to dismiss it as a merely a casual game. Once we get it in front of players they usually love it, but doing so takes a fair amount of effort.
We’ve also had an overwhelming amount of piracy – 95% piracy rates on iOS and Mac, with Windows being well over 99% (Windows sales are comparatively a rounding error). It’s great that players are enjoying the game but having gone with a design oriented towards making it a premium app instead of a freemium game means we get no benefit from those playing it for free – not even a ranking increase.
We have continued supporting the game, releasing so far three minor and one major updates, including an adjusted difficulty curve and 12 new tutorial levels (bringing the total up to 84), but as of this writing that update has only been out for a couple of weeks – it still remains to be seen what effect it will have.
Where do we begin?
Target it, goddammit - We focused too much on the gameplay and on crafting characters with a personality we enjoyed, without considering if we were sending out mixed messages that could confuse players, alienating precisely those we wanted to rope in.
Cut the dead weight early - the initial modeler and animator just wasn’t working out, and keeping him on board for longer than I should have was not only an expense I could have saved, but risked losing us an invaluable team member. Deal with these problems sooner rather than later.
Be ready to trim - we have a plethora of characters and elements we just didn’t get to use because we simply had no time to properly develop them. As much as you love your designs, chances are a lot of them will end up having to be left out.
If you’re indie and working on a premium app, reconsider - there doesn’t seem to be a middle market on iOS apps anymore – it’s almost all either huge AAA-quality projects, or simpler one-mechanic freemium games (and some freemium games have been getting shinier and more elaborate). Do you really want to bet the house on a business model with a piracy rate higher than 90%, when the market is flooded with competing titles that players can get for free?
Where to go from here
Stubborn as we are, we find ourselves already working on our next title, after going through multiple prototypes and even more concepts – this time applying the lessons we learned from Hairy Tales. I expect we’ll manage to retain the spirit of experimentation and sense of humor that we imbued the game with, while setting it into a game design more fitting for today’s game climate. Wish us luck!