Xeodrifter Postmortem

Since the release of our first self-published game, Mutant Mudds on the Nintendo 3DS eShop; my outlook on game development has changed quite a lot. It was truly the first time I created a game for me and not for the market. And, it was an amazing experience!

Now, don't get me wrong; our debut game, Dementium: The Ward, was a meaty slice of putrid heaven to create and I love that game to bloody bits, but the reason we chose to develop that game was somewhat market-driven. Lots of people like survival horror, and the DS market was sorely lacking in that genre. We had decent experience with the FPS genre and a good grasp on what we thought the DS could handle.

With each game we developed after Dementium, I continued to evaluate the market and try to find a space where the "three P's" would be satisfied: player, press, and publisher. In those days we were reliant on publisher money advances to cover the cost of development, and as such needed to convince the publisher that our games were worth investing in.

We were extremely fortunate to have developed three original first-person shooters and two original racing games before the market presented us with the opportunity to self-publish Mutant Mudds. It was not a bad run at all, but nevertheless, each of those games was guided by what we thought the publisher would believe to be marketable. And, the retail market has got increasingly competitive and risky over the years - especially in the Nintendo handheld market. Mutant Mudds, for example, is not a game that I believe any publisher would have put on the shelf for $30 in January 2012.

Why am I going on about this? Because it led me to the insane decision to stop development on one of our eagerly anticipated games - Treasurenauts - and start on something completely new. Why would anyone do such a thing?

Let me take you back twenty years, to 1994, when a bold new game was released on the Super Nintendo. It was, of course, Super Metroid. And, it was unquestionably a significant milestone in gaming. It presented so many new ideas, emotions, and experiences in one game that it just blew the doors open in so many ways. I could probably go on about Super Metroid for a few more paragraphs, but I should save that for a different time.

What's important to note is the affect it had on me, as I am sure it did on many others. Super Metroid is a wonderful adventure where you write the story. Where you discover the mysteries. Where you grow as a person, and subsequently increase your scope of exploration. Not to mention the art, audio, and... ok, you get the point. Epic game!

Fast forward to early 2014 and we were developing the enhanced port of Moon, now called Moon Chronicles on the 3DS. I have always been fond of the concept of demakes (Dark Void Zero? Yes, please!) and thought I'd try to create a mock-up of what a 2D demake of Moon Chronicles might look like.

I spent a few hours here and there tinkering with a variety of different ideas and then moved onto "real work", which at the time was finishing Moon Chronicles, work-for-hire gigs, the Cult County Kickstarter, and trying to find the time to continue development of another passion project, Treasurenauts. And thus, the Moon Chronicles demake mock-ups were left dormant in a folder on my PC.

We had already been working on Treasurenauts for some time, and unfortunately delayed its release due to our need to work on paying gigs to keep the company going. I don't typically announce any financial hardships we face because it isn't fun news, and it is really just a natural part of business. However, the reality is that it affects our ability to work on what we want, when we want.

We poured a lot of time and effort into our Cult County Kickstarter campaign, which ran the month of April, 2014. That took a considerable toll on us, and resulted in a failed KS campaign in the end. Our next move was to be an important one. I was feeling pretty low after the campaign ended and needed to do something that made me feel warm and fuzzy again.

We were at the point of having some time ahead of us to work on Treasurenauts, but at the same time we needed to release it before the end of 2014 for it to help us out financially in 2015. When I looked at what was left to create, it simply didn't fit in the time remaining. We were faced with two choices: simplify Treasurenauts and release it in 2014, or keep the design as is and delay it. Bearing in mind that delaying it still meant we needed to release *something* in 2014 to make some cash.

I spent a few days focused on thinking about our next move. What would a simplified Treasurenauts look like? What type of new game could fit in a 3 month timeframe and retain top quality? Perhaps there's something wrong with me, but I revel in these sorts of challenges. I had created what I thought was a decent Treasurenauts Lite feature set, and also a list of random ideas that have a small chance in hell of getting completed in 3 months.

My partner in crime with Renegade Kid's 2D games is Matthew Gambrell, who is the solo programmer (and designer, composer, audio, and wizard) behind our 2D offerings. I would frequently discuss ideas with Matthew, and we would both celebrate and shoot-down each other's ideas until we had a list of mutually agreeable realistic potentials to consider. Neither one of us wanted to cut anything out of Treasurenauts, but at the same time sometimes hard decisions have to be made and you just have to make them and put your best foot forward.

Our first decision was to create a bizarre overhead RPG-esk dungeon thing, which we spent about 3 weeks on. It was shaping up pretty good, actually, but in the end I just wasn't feeling it at the time. And this (finally) brings me to the reason I mentioned Mutant Mudds at the beginning of this post.

So much of what makes something - anything - worthwhile is how you feel about it. The emotions behind the creation of something can be the difference between producing something boring or magical. After all of the thinking, planning, and prototyping, it finally hit me.

The 2D demake mocks of Moon Chronicles jumped into my mind. I felt joy. Could it be possible? Could we make a Metroidvania game in a few short months? Oh dear, logic was being overtaken by sheer passion and desire. I was in trouble. I had passed the point of reason, and subconsciously decided that I needed to make a Metroidvania game in an obscenely short amount of time.

When thoughts of Super Metroid and the 2D demake mock-ups of Moon Chronicles combined with the realization (and disappointment) of the fact that there had been no 2D Metroid released on the 3DS, it was all over for me. I had no choice. It had to be done. I had to create my own Metroidvania game for the 3DS.

The first thing that I worked on was the overall flow of the player's journey through the game. To me, the heart of a decent Metroidvania game is the experience of coming up against an obstacle that you cannot pass, but somehow realizing that this is the way you need to go. And when you collect a power-up, you know where to go and use that power to gain access to a new area.

This is achieved by doing a few things. First of all I needed to put these obstructed pathways on the player's main path so they were easy to find, and ensure the mini-map communicated this location as somewhere to revisit. If we look at the first area in the game, the player is immediately greeted by water. Can I swim? Will I drown if I jump in? The player may choose to avoid finding out by jumping over the water to land, and that is OK. Some players may choose to jump right in and find out, knowing the save spot is very nearby if they do meet their death.

On the mini-map it shows the outline of the map to have a broken line, heading down, below the water. Maybe there's a way to get down there? The player can explore to the right, or turn around and explore to the left of their start location. Left leads to more water, and another mini-map indication of something below.

Continuing in the only direction available to the player, the player makes their way to the right, up a few platforms, encountering a few enemies and quickly to the first boss encounter. There are many carefully placed setups from the starting location through to the boss encounter, but this is more of a postmortem than a design breakdown, so I'll try to not dive too deep too much.

Defeating the boss rewards the player with the first power-up; a submarine transformation that enables you to explore the watery depths. Having the boss encounter early on helps establish a nice sense of scale in interaction and more importantly firmly cements the relationship between perceived obstacles and power-ups that enable you to overcome them. Exiting the boss room presents you with an unavoidable situation where you are required to utilize the submarine ability to proceed, which is another vital element with a Metroidvania game.

So, I spent probably about a week focused primarily on the map flow of the entire game, moving this over there and back again, removing that and then adding this before landing on something that seemed to resemble a decent start - knowing that it would all change as soon as I started implementing this for real. Figuring out which power-ups to include in the game, and how they would present obstructed pathways was a constant juggle that we tried to nail down as soon as possible.

One of the main factors that made a Metroidvania game plausible in any way within a short period of time was the fact that we not only had both Mutant Mudds and Treasurenauts to build from - and borrow from - we also had scraps of a Run and Gun game we messed around with in 2013. In addition to providing a solid foundation to start with, my first pass at enemy designs was to see what I could steal and mutate from our library of goodies, or baddies I suppose. This proved to be a huge advantage.

This provided the basic AI for more than 50% of the enemies in Xeodrifter. From their base parameters we could change a single number, such as movement speed or a pause delay, and produce a very different interactive experience. Combine that with different sizes, hit-point damages, and of course unique sprite art and you have an enemy than can feel very different from how it started out.

From a resource perspective, I knew I could not create all of the art assets for the enemies, player, and levels - it was just too much to do in the time we didn't have. Fortunately, one of my good friends, Michael Veroni, was available for some side work and helped out by creating a number of the enemies and all of the boss animations. This allowed me to focus on the levels and the player, which was great.

What Went Right?

1. Genre.

Sticking with the 2D platformer genre was probably the smartest thing we did. Both Matthew and I love the genre, which is a great start, and we have a decent bag of tricks to pull from.

We went through a lot of growing pains with the development of Mutant Mudds, because we were starting from scratch in pretty much every way. Not only was the genre a new thing for us - so was the 3DS hardware! We were able to skip right over all of that business with Xeodrifter and get straight to the fun stuff!

We had our own engine. We had our own tools. We had knowledge, not only from the development of Mutant Mudds, but also Treasurenauts and a host of other experimental prototypes we had messed with in the past two years. Being in that position was incredible. Mentally I was able to start thinking about the game itself, and not worry about how we do it, or even if we can do it.

In addition to providing a solid foundation to start with, my first pass at enemy designs was to see what I could steal and mutate from our library of goodies, or baddies I suppose. This proved to be a huge advantage.

This provided the basic AI for more than 50% of the enemies in Xeodrifter. From their base parameters we could change a single number, such as movement speed or a pause delay, and produce a very different interactive experience. Combine that with different sizes, hit-point damages, and of course unique sprite art and you have an enemy than can feel very different from how it started out.

Another key factor in choosing the 2D platformer genre is the fact that Mutant Mudds is our biggest seller. That may have been a fluke. Maybe people were just starved for games, or maybe people love little Max. Or, maybe, the eShop audience just enjoys platformers that have been crafted with love! We're hedging our bets on the latter.

2. Art style.

I love pixel art. I like looking at it. I love creating it. And, I like playing pixel art games. There is a wealth of different pixel art styles that have been created over the years in both the 8-bit and 16-bit eras. And, of course today we see even more pixel art styles emerge using modern capabilities and traditional limitations. So, choosing the art style for Xeodrifter was not an easy thing.

I already had the mock-up of Moon Chronicles demake that inspired this whole thing, but I am not one to stick with that for the sake of it. I always give myself the opportunity to explore the possibilities before committing, just to make sure it is the right choice. I have to consider what it means for the project as a whole.

The most important element with this project - as with most projects, really - was time. I was going to be the only, or one of the few, artists creating the art for the game so I had to make sure I could accomplish everything within the time-frame.

Fortunately, the demake mock-up was already simplified down to almost the simplest style that I would implement into a game. Colors were minimal. Palette was simple NES. And the tile complexity was also very simple.

Adopting the thought process of minimal-but-potent design worked out extremely well. Each planet is made up of only a few key elements to produce large levels of variety. This also translated nicely to the enemy designs, user interface, menus, and the player animations.

Another key element that helped with time and effort is the fact that the game is scaled up 200% in the near layer - just as it is with Mutant Mudds. The player sprite in Xeodrifter is about 12 pixels high, but results in taking up a space of 24 pixels high in the default near layer, which reduces the amount of detail and pixel density required to execute any art assets.

This also has the benefit of looking super high res when jumping into the background layer, where it is displayed at 100% resolution.

Even though the simplified art style helped keep the art creation doable, I was lucky enough to enlist the help of my good friend and tremendously talented artist, Michael Veroni, for the creation of some enemies and the boss animations. His assistance was a huge help and added a great sense of variety to the game's overall aesthetic.

3. Audio.

We took a slight gamble with the music for Xeodrifter and asked someone who we had not worked with before to create some chiptunes for the game: Roth Sothy. His previous work was great, but that doesn't always translate into good results for your game.

Fortunately, he knocked it out of the park. Roth's music for the game fits wonderfully, and continues our tradition of great chiptunes.

In addition to that, our very own Matthew Gambrell composed some delightful tunes for the game - including the title tune - that work perfectly. Where he found the time to create these no one knows, but I am thankful he did.

The moment we announced Xeodrifter to the world, my pal Brian Altano pinged me about a tune he had just finished. He thought it would be perfect for the game. He was right! It was not a chiptune, but it was awesome. In the end we used it for the end credits, as that felt like a great way to incorporate a new style of music, and it worked perfectly as a way to send the player off into the cosmos after their accomplishment of finishing their adventure.

Despite Matthew already being taxed with the solo programming task of completing Xeodrifter, and the creation of some great tunes, Matthew also found / created the time to provide the sound effects for the game - and they're great! Piece by piece, one sound effect at a time, the audio slowly formed over the course of the project. And, each time a new sound effect was added to the game it improved the experience.

4. Delaying Treasurenauts.

Not without its downside, delaying Treasurenauts to create Xeodrifter was ultimately the right thing to do - for us. It was not an easy decision, and it was not a popular decision. But, in the end that decision enabled us to release a game in 2014 with the potential of earning the company some much-needed revenue.

Treasurenauts could not have been completed in time for a release in 2014. So, the need for a smaller scoped game presented itself. Xeodrifter was that game.

Not only did Xeodrifter become the little game that could, it added to our bag o' tricks. The efforts put into Xeodrifter can be utilized by the continued development of Treasurenauts. This will ultimately make Treasurenauts a better game - albeit a very much delayed one.

I am also super excited to have a traditional metroidvania game in our catalog of games. It is one of my favorite sub-genres, and something I have wanted to create for a very long time.

5. Approvals.

One of the things that went right, that I was perhaps the most relieved about, was getting it approved by Nintendo on the first submission. The only time that happened for us before was with our first game - ironically - Dementium: The Ward for Nintendo DS, thanks to Bob Ive's wizard programming skills.

There are so many small things that can be buggy or slightly wrong and cause a fail at Nintendo's lotcheck, evening including mistakes in the eManual, which is one of the most gut wrenching reports to receive. We always try our best to get our games clean, but don't always achieve it.

The fact that Xeodrifter was approved first time is a testament to Matthew's approach with the programming of the game. Much fine-tuning and clean-up and/or testing was performed throughout the entire development process, and not left to the end of the project. Combine that with a healthy dose of good luck, and you have a chance at passing first time!

An aspect of the business side that some people may not be aware of is how when we get paid the revenue is earned through sales of our games. Releasing Xeodrifter in the fourth quarter of 2014 means we receive payment 30 days after the end of the quarter, which in this case is the end of January. Whereas if the game was released in January, payment wouldn't be received until the end April.

What Went Wrong?

1. Planning / time / scope.

Well, Xeodrifter was supposed to fill the gap of a 3-ish month project, but ended up taking nearly twice as long. Sure, a five month project is certainly not a bad thing, but we had originally hoped the game would release in the third quarter of 2014 - not at the end of the fourth quarter!

Even though I am happy with the results of the game, we allowed ourselves to add to the game in terms of features and overall scope throughout the course of development. If our lives depended on releasing the game in the third quarter, I would not be here now to type these words.

Fortunately, we were able to limp by without the revenue the game could have earned in Q3 and subsequently arrived in late October, but not without some financial hardships. The latter part of 2014 was very tight financially due to the slippage of the launch.

Due to the delay in releasing Xeodrifter we took on some contract work to bring in some cash, which was equally distracting as it was helpful. The contract work did not initially affect the development of Xeodrifter, but after a month or two of the contract project Matthew was required to help out, which resulted in his split attention.

To avoid this in the future I need to make sure the game we are developing is properly scoped out before assuming a street date. If I had spent a proper amount of time scoping out the game it would have been obvious early on that the game would not fit in anything less than a 5 month time-line and we could have made decisions based on that information instead of taking on contract work as a desperate band-aid.

2. Delaying Treasurenauts.

Even though delaying Treasurenauts enabled us to release Xeodrifter in 2014 - instead of rushing out Treasurenauts - it had negative effects on us too. Those that were looking forward to Treasurenauts were understandably upset that it was being delayed - again - and we lost some of their trust as a result.

It gave people the impression that Treasurenauts was in trouble, and needed to be delayed to fix something wrong or broken, even though the truth was we wanted to preserve the potential of Treasurenauts by not reducing its scope and rushing it out.

Another negative effect from this is how it makes us, as a company, seem indecisive and wishy-washy about what we work on at any given time. This is one of the dangers of sharing so much information about what we're working on instead of keeping everything secret until it is ready to release.

How this might affect the reception of Treasurenauts when it is released remains to be seen, but there is a chance that those that were once excited about the game might be cautious now about taking the game at face value.

To combat this I feel as though the best thing to do is to make sure Treasurenauts is as awesome as it can be to justify the delays. Hopefully by making a great game people's concerns will be addressed and we can all hug each other and move on about our day.

3. False start(s).

We knew that we needed to develop a game in a short amount of time, and we wanted to make sure the game was scoped to fit within that time to produce a good experience. This resulted in much work upfront exploring different ideas, and some false starts.

We started work on an overhead action RPG type of game. It was not strictly a RPG in the traditional sense. There was to be no story or dialog to speak of, but more focused on roaming the land and killing enemies.

We spent about 3 to 4 weeks developing the idea in prototype form, and managed to create some decent results we were happy with, but in the end it didn't seem to fit well in the time-line we wanted, which was a real blow when we had to make the difficult decision to axe it and pursue other ideas.

There's really no way to avoid these sorts of things when trying to develop something original. If we were exploring sequel ideas or licensed properties, it would be a very different approach, but with original games you have to spend at least some time exploring the idea on paper and/or in gameplay.

4. Work-for-hire complications.

This is something I touched on above. We had to take on a contract work-for-hire gig to bring in money to pay the team so they could pay their bills, etc. On one hand I was very appreciative to have the opportunity to generate some cash by performing a service for another company, but on the other hand I couldn't help but feel the pain caused by the fact that our focus was divided.

What we wanted to do verses what we had to do were at odds. We are a small team, so dividing our attention between too many projects and spreading ourselves that thin is not an effective method of bringing in cash. Fortunately, we were able to be effective in producing quality work for both the contract work and Xeodrifter, but it was a strain.

In the future I need to find other ways of generating cash when needed in desperate situations that do not tax the team that is passionately working on a game. The hope is that one or some of our games sell enough copies to give us a wad of cash that is padding for desperate times. We have had a little in the past, but it can easily be eaten away with just a couple of month of no income.

We discovered that banks only give out loans, line of credits, and credit cards to companies that already have cash and don't need extra cash at that time. This makes sense, for the banks, but it is completely useless for a company that actually needs money in the short-term to get over a hump. I guess we just need to stock up on all of that surplus credit and good will from the bank when we don't need it so we have it on a rainy day.

Conclusion

Xeodrifter was a really fun project to work on. Developing an original game in a period of 5 months from inception to completion was an incredible experience. Many things went right and many things went wrong, and in the end we created a game I am super proud of.

Moving forward we will be developing more 2D platformers - some announced and some not yet revealed - as well as getting back to our polygonal roots with at least one first-person shooter in 2015.

Our hope is to generate enough revenue from self-publishing our original games so we never need to do contract work to survive, but only do it by choice when great opportunities present themselves.

Data Box:

Developer: Renegade Kid.
Publisher: Renegade Kid.
Release Date: 12/11/14.
Platforms: Nintendo 3DS (eShop) and Steam (PC).
Nintendo Game Page: Xeodrifter on Nintendo.com
Steam Game Page: Xeodrifter on Steam
Number of Developers: 2 fulltime / 4 part-time.
Length of Development: 5 months.
Budget: Under $100,000.
Lines of code: 40,400.
Number of Dr. Pepper drinks consumed: 203.

My Twitter: https://twitter.com/JoolsWatsham

Renegade Kid website: http://www.renegadekid.com

Xeodrifter website: http://www.xeodrifter.com