Game Career Guide is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Get the latest Education e-news
  • Postmortem: Skirmish Line

    - Tony Hua

  • Logistics

    Naming Matters

    Skirmish Line features three modes: Defense in Depth, Sector Defense, and Campaign Mode. When the game was initially released, Defense in Depth was called Arcade Mode and Sector Defense was called Classic Mode. I will be inverting the order of GBU here.

    The Unexpected

    Many of the influencers we contacted only showcased Campaign Mode, despite explicit instructions in our emails for them to focus on the then named Arcade Mode and Classic Mode.

    The Bad

    Campaign Mode, with its troubled development, was not the most polished game mode. With most of the focus on the two endless defense modes, Defense in Depth and Sector Defense, and influencers making videos on Campaign Mode, the best parts of the game were essentially being ignored.

    The Good

    After seeing this happen, I contacted one of the influencers and asked why they decided to feature Campaign Mode without even a mention of the two other modes. It turns out Arcade Mode and Classic Mode just don't sound very good. When we renamed Arcade Mode and Classic Mode into Defense in Depth and Sector Defense, we began getting a better mixture of coverage of the game.

    Build Clustering

    After Skirmish Line was released on Steam Early Access, I began grouping changes together into major builds thematically named after major battles with the Japanese during WW2. While hotfixes and small changes would be patched in, major features were grouped under the named builds.

    The Good

    Although seemingly arbitrary, grouping major changes under named builds yielded a lot of benefits. On the development side, this helped with pacing and morale. Being able to point to what you've accomplished in a set amount of time was a solid morale boost. On the community front, this helped focus community attention towards new challenges and let players know that progress is being made.

    The Bad

    When possible, I tried to group tangible content updates for players. Unfortunately, as in the case of software development, sometimes the work done on the backend is not something players can immediately appreciate. There were a few builds whose release I slowed in order to incorporate more content type changes in order to keep players excited. As a developer, I find it tempting to show my players the newest developments, but it might be worth it to keep some content items hidden or saved for later when you're working on some major refactoring or new backend systems. Alternatively, this would also be a good time to implement a few low hanging fruit ideas, such as the in-game achievement and mutator systems mentioned under the Design portion of this article.

    The Unexpected

    Building cluster was an expected evolution of the development process. With the game on the store front, it's important to maintain some degree of stability for players. Naming the builds though was initially done on a whim because it simply seem more interesting for major patches to have a name alongside the number.

    Voice Acting

    Most of the voice acting in Skirmish Line came from fiverr, a freelancer website with an emphasis on low costs. At the time, our budget for voice acting amounted to only a few hundred dollars, and we weren't able to afford most of the requested fees from more professional websites

    The Good

    Surprisingly, despite very low prices, we were able to get free voice samples from the scripts I had written. A few freelancers denied our sample requests and asked for us to pay for a gig, but the majority were willing to just throw their attempt at voice acting.

    It took about a month for me to finalize my decision on which voice actors to take. I spent a lot of time listening to the voice samples, listening to the samples the freelancers had own their own sales pages, and asking friends to pick between the various samples.

    Many voice acting freelancers seem to enjoy voice acting for a videogame, as it was something very different from the usual commercial gigs that are brought on the site. One freelancer, who wasn't able to provide a voice sample due to a family emergency, contacted me afterwards and asked for me to consider him for future voice acting roles. Another, the Japanese voice actor, even corrected the script I had sent him free of charge in order to get the lines as authentic as possible.

    The end result was that we were able to get quite decent voice acting at very low prices.

    The Bad

    A friend of mine had convinced me to get a non-Japanese speaker to voice act the Japanese lines because he liked how the voice actor sounded. This voice actor, who had asked for my budget, wanted the full budget that I could allocate, a figure I didn't even exactly have at the time as I was approaching the situation from a bottoms-up approach by adding together the cost estimates across the actors. The voice actor was barely able to deliver the requested lines, taking over two weeks longer than any of the other voice actors, and we had to find a replacement voice actor in the end.

    When I had to find new voice actors later on, I spent a lot less time vetting out the samples. Due to time constraints and pressure to release new content, I had invested only a week in listening to voice samples. Subsequently, there was a drop in quality on some of the later voices, particularly on Santa's voice.

    The Unexpected

    An aspiring voice actor on a community modding forum I browse frequently had asked if he could voice act in some of the mods. When it seem that none of the mods were looking a voice actor at the moment, I sent him a private message for a role in Skirmish Line. This turned out to be a win-win situation, as I worked closely with the voice actor to help prepare his equipment setup and he was able to provide a nice set of British and Australian voices.


    As with voice acting, I took a similar approach the game's music. After vetting through a selection of freelancers (this time on peopleperhour), I was able to find a musician who was willing to provide the music that fit within my budget. And similar to the voice acting, I later came across an aspiring composer looking to get into the game music industry. We worked out an an arrangement where I was able to get some low cost but nonetheless high quality tracks.

    Unity Engine

    Skirmish Line was built using the Unity engine. The choice of Unity wasn't made for any particular technical reasons. After I had graduated from university, I was working with a friend to make a mobile game. Since he had used Unity for one of his school projects, the decision to use Unity was primarily because it was the engine I had known about. Although this mobile game project would fall through, I would end up using Unity by association.

    The Good

    Unity is a very solid engine, certainly more powerful and optimized compared to something like Flash, which was what Mud and Blood 2 was built on. Unity allowed the game to run much more smoothly even when there were a lot of units on screen. This in itself was likely a selling point, as Mud and Blood 2 was notorious for slowing down to a crawl on later waves. Documentation on the engine were also very good, with plenty of user support online.

    The Bad

    Unity is first and foremost a 3D game engine. Combined with the engine being a blackbox, we had to figure out Unity's kinks and how to build a game using its systems. There are likely better 2D game engines available and Unity was probably overkill for the project's requirements.

    The Unexpected

    Since Skirmish Line ran without any real issues, players didn't experience any slow down on later waves. A Skirmish Line game that passes wave 100 would usually be roughly 40 minutes of play. By comparison, a Mud and Blood 2 past wave 100 might have taken around 2 hours due to the huge slow down in the game processing. This is a solid benefit for players looking for a smooth playing experience, but some veteran Mud and Blood 2 players would perceive Skirmish Line as being too fast or short of an experience. Such players had already expected to be spending 2 hours on a single game.

    Buying Versus Developing

    For people looking to get into indie development as business, don't be cheap. If you haven't already, make sure you have a sizable sum of money saved up, not just for your own living expenses but to put into the project. If you need high quality art and your own work can't even come close in quality, hire someone else to draw it. If there's a technical feature you need, consider looking to see if it's been done by others on free to use license code or available on the asset store. For instance, the $30 spent on Nested-Prefab for Unity drastically cut down on the amount of work needed. Your own time is a commodity, and it needs to be used for developing everything else in the game.


comments powered by Disqus