Dim Sum Fusion Assassin post-mortem

Posted by (twitter: @@mike_cullingham)
May 11th, 2015 12:38 pm

Wow, can’t believe 3 weeks have come and gone since I finished my first Ludum Dare.  I really, really enjoyed the process and I’ve been giving it a lot of thought (seriously, it’s been a significant distraction).  Since I plan to do this again, I’ve compiled this list into a post-mortem on Dim Sum Fusion Assassin.

What went well

  • Overall scope
    * I really estimated the scope well… and by well, I mean I only went 50% over what I figured it would take. I originally calculated that it would take about 10 hours to code the bare minimum game, and I finished up everything in the scope somewhere around the 14 or 15 hour mark.
    * I also was really good at only working on things that directly enhanced the game. I allowed myself a bit of a break to play with the background pattern, but other than that I didn’t go down any rabbit holes.
  • Framework choice
    * Coding in C#/WPF/XAML is my day-job, and the type of game I was making worked really well with my existing skill set. I even managed to learn some things that you don’t use with enterprise software, but I didn’t spend a lot of time figuring out _how_ to do something.
  • Graphics
    * I made the decision really early on to just use stock emoji rather than spend time searching for free clip art or drawing my own stuff. It was limiting in some ways, but I saved a lot of time doing it this way and think the variety of animals and foods at the very least were clear in what they were supposed to represent. I’m also happy with how some of the subtle gradients turned out.

What went wrong

  • Name
    * As pointed out in the comments, the food served doesn’t really make one think of actual Dim Sum. And on reflection “Fusion” shouldn’t cover replacing 90% of the food generally available. I think I got attached to the title before selecting the graphics and then didn’t step back to reflect that the graphics no longer supported the title.
  • Balancing / Tutorial
    * Originally I figured the game would play much faster and you could handle 8 customers from the start and levelling would only make it go faster. By the time it was ready for full playthroughs late on the Sunday, I realized that even though I understood every part of it I still couldn’t play it successfully. In the end I had to cut the number of customers to 3 at level 1 and slowly increase it. I think this is better for the game going forward, but I spent too long trying to get it to work with 8 and the intro to a new player isn’t quite right.
  • Order of Implementation
    * Although I got all gameplay elements done in the original scope, I really didn’t spend enough time planning out the order of implementation and so couldn’t actually play it until late on Sunday. To some extent there weren’t a whole lot of items that I could’ve juggled, but even getting it playable an hour or two earlier would’ve exposed the balancing issue earlier.
  • Sound
    * I’ve only played with sound in WPF once before, so I knew all along it wouldn’t be a 5 minute addition and would require some time to experiment. In the end I didn’t think it would add as much as trying to fix some major gameplay bugs, but the game definitely suffers from lack of any sound.


  • Port
    * I’d like to continue with this project and get it onto at least the Windows / Windows Phone stores. This should be fairly simple since the core game logic can be used directly, but so far I’ve discovered that there are a lot of differences between WPF XAML & Universal App XAML.  Hopefully I can maintain my interest in the project to get past this hurdle.
    * Another possibility is that I’m setting a goal for myself to learn a enough of Unity 2D to create a playable game using that framework, so this might be a good simple project to experiment with.
  • Name/Gameplay Changes
    * I’d rather keep the simple emoji graphics than spend a lot of time soliciting or searching for accurate graphics, so have looked over what’s available to me and currently I’m leaning towards renaming it to Sushi River Assassin or Sushi River Defender.
    * A bomb may be more evocative than poison.
    * The identification of the target isn’t a huge part of the gameplay. Perhaps the target should always be the ogre emoji. Also prevents squeamishness about poisining cute bunnies, etc.

Thanks to everyone who played it and gave me feedback.  It is incredibly valuable and I appreciate every comment, good or bad.


Leave a Reply

You must be logged in to post a comment.

[cache: storing page]