Magpie - Post Mortem
Back to the game


To conclude our Ludum Dare 28 adventure, we (Moose and Vadinci) wrote a post mortem about the game that we made during the weekend of 14 to 16 December.

The concept of Magpie
The theme for Ludum Dare 28 was "You only get once" (YOGO). Both of us downvoted this theme during the voting round, not because we thought it was restrictive, but because it didn't open up potential for really creative ideas. As we had expected, a lot of games took this theme in the fashion of "you only have one bullet" or "you only live once". Of course, it's hard to avoid these ideas, since "having one" implies some physical restriction.

We tried to stay away from those ideas. We both had our own idea when we first met after hearing the theme. The first one was a strategic game where you only have one turn to plot out your entire battle and anticipate your enemies' moves. The other played with the restriction of just holding one item at a time, while you had a whole bunch of items with different uses. This is the idea that we went with, and eventually became Magpie.

The name is a result of "Hoarder" being a name already being used for some overlord game in process on indieDB, and Thesaurus.com suggesting Magpie as a synonym for 'someone that hoards objects".



What went well
According to the artist
The style - There are two styles of art in the game, which came to be by accident. I started with making a ground texture that I spent way too much time on. I was happy with the result, but quickly realized that there was simply no time for this style to continue. I then switched to the style you see the characters and assets in.

The pace - As a person who is never satisfied with their work, it was quite difficult to create art and then 'just go with it'. I had to force myself to just save the asset and move on to the next once. Fortunately, that did somewhat save the style as a whole.

According to the programmer
Using FlashPunk (once again) - FlashPunk keeps being my favourite framework to crank out games with. Even though I completely abuse the system in the 72 hours that I have to make a game, FlashPunk is a solid base and provides a lot of features that are essential for building a game.

Re-using old techniques - A lot of things that I did to handle the tile rendering, Z-sorting and dungeon-generation are things that I've toyed around with a few times before. Reapplying techniques that I've developed earlier saved me a lot of brainstorming (and therefore time)

Creating a base for rapid content addition - Even though I find it scary to work on something for a day that doesn't really change anything on the surface, time and time again it proves extremely useful to create a dynamic system to support your content. On the second day, I had the easiest time adding new items, because of all the groundwork I laid down the day before.

According to both of us
Realizing the concept - The process of transforming the idea from paper to game went really well. We both had a really clear idea of what we were aiming to create. There were some doubts about the pacing, but at the end of the first day those troubles disappeared. From that moment on we could focus on creating more content for the game and not lose time fixing the mechanics.

Teamwork - This was not the first time we've worked together, and it has been another positive experience for both of us. It is easy for us to give honest feedback without having to worry that the other person might feel offended. Furthermore, we agree on a lot of points and have a similar working ethic.

Fun - Having fun with your colleagues and the game is something that went really well this time. Genuinely having fun while playing your game is a huge boost in morale and made it easy to work on it throughout the weekend.



What could be improved
According to the artist
The style - The style isn't coherent. Not in any way. This could have been avoided by actually giving it some thought instead of just starting somewhere. In hindsight, I would have opted for pixelart. Not being a very seasoned artist myself, I find pixelart easier to set up, especially when creating tileable environments. Creating good pixelart is another story, but in any case it might've worked out better. The type of game lends itself quite good for that style.

The workload - The workload itself wasn't particularly high, but the lack of an easy style that I was happy with made it difficult to quickly create assets. Even though I could create them relatively fast, the quality was low.

According to the programmer
Not providing enough player feedback - Even though this might be a flaw of mutual origin in the game, I feel as if it were my responsibility. As the one who makes everything 'come alive' I should have provided the player with more feedback as to what's going on. Right now, it's not really clear that actions can be performed in four directions, even though the character only faces in two. Battles with enemies might feel random (even though there was quite some timer-tweaking there) because you can't really tell the underlying mechanics. I wish I would have done more in this section as I feel it would improve the game a bunch

The AI - "//I CAN'T GET F*ING PATHFINDING TO WORK FOR SOME RETARDED REASON" is a line that resides somewhere in "Enemy.as". I wrote it after trying to make a simple a* work - which obviously did not work out. In the end I found another solution to make the AI look less stupid, but it is still pretty easy to trick them, and that has actually become the sole mechanic to fighting them. The boss only has one, really simple charge attack as well. I could have improved a lot here, but time-limits made me focus on more importing matters (like implementing an 'ending')

The code is a mess - This is a thing that always happens with harsh time limits (at least it does for me), but I've really outdone myself this time. The code that handles the white out to the win screen, which is triggered by walking on a certain tile, is handled in the class for the boss. I am going to leave it at that.

According to both of us
Lack of items - The main mechanics of the game have really proven themselves to us as well as others that we showed it to (and the ones who commented on the Ludum Dare site so far, thank you so much :D ). Still, we feel part of the concept is having a buckload of items, all with different, unique, interesting properties that you would hoard and use in different situations. At this point though, there are just about 9 items in the game, of which 4 are weapons and 2 are currency. The fault here lies mostly at the lack of time and the fact that coming up with unique ideas requires bursts of inspiration, which you can't really force.





That's all folks
We hope you got something out of this read. The next time we participate in a jam (which will most likely be the Global Game Jam of 2014) we will take all our successes and mistakes into account, to ensure we keep improving.
but really, stop reading text and go back to playing Magpie!


Back to the game