Twin Stick Blower - 3rd Winter Game Jam - Post Mortem
This post mortem describes the development of the game Twin Stick Blower which was created during the 3rd Winter Game Jam at the Alpen-Adria University in Klagenfurt. As usual for post mortems, this document is separated into a section that describes what went right or better than expected, and a section that describes what went wrong or could have been done better.
What went right or better than expected
Shortly after the development started it was clear what the main feature of the game would be: Controlling the character movement through the push-back of two leaf blowers, each controlled by one analogue stick of a gamepad, and also blowing leaves around with these leaf blowers. This creates an unconventional, but interesting way of balancing movement and action during gameplay. The focus of the first hours was to balance the main feature regarding how the movement feels and how the leaves physics are handled, so that it felt nice to play.
Early feedback from other developers
Due to our unconventional way of controlling the character movement, we were afraid that it would be to weird for players. Therefore, we decided to gather player feedback as early as possible by asking other game developers at the game jam. Most of the feedback was very positive and the controls were called unusual but interesting and that after a few minutes of gameplay they got the hang of it. Some players during the game jam and after the game jam did not like the controls, but we decided to stick with them because the positive feedback was dominant and we would rather create something new and special than sticking too much with the standard model.
2-player and 4-player multiplayer modi
The main idea for the game was the 2-player mode, but we started with implementation of the single player mode to get the controls right and test the game concept as described above. Doing this was a good move because the game most likely would have turned out completely different if we focused on the 2 player mode from the start. After everything was working for single player mode, the focus was placed on 2-player mode which was implemented very fast and after getting the tweaking done, we even found the time to implement a 4-player mode which turned out to be even more fun than the 2-player mode.
Keeping the game simple
We decided right from the start to avoid feature creep and although having some cool ideas for features that would have made the game even more fun to play, we decided to put them aside and just use the basic leaf blower mechanic. These ideas included automatically refilling energy that could be used to increase the blowing strength or power-ups that would alter the gameplay.
Game graphics, music and sound effects
Based on the low number of experienced graphic designers and music composers at the game jam, I was afraid that our game would not fit the standards I was expecting, with having good graphics, music and sound effects. But I need not have worried. Although my two teammates have never done anything regarding game development before, they were very motivated and by following the rules that freely available assets were allowed, we manage to create nice aesthetics for our game by mixing them up with self made assets.
Focus on game polishing
Due to keeping our game simple and focusing on the main mechanic, the game was fully playable after 24 hours of development. This allowed us to focus the rest of the time on polishing and juicing up the game, which to some degree helped us win the game jam.
Based on my experiences with game jams, I encountered that a team of three persons works fine. One for programming, another one for graphics and animations, and another one for music and sound effects. This way, conflicts between team members are kept at a minimum. As I aimed for this team composition, we had almost no conflicts and could focus more on game development.
What went wrong or could have been done better
Due to the rather random creation of the team, we decided to keep the asset sharing as simple as possible and, therefore, used a shared Google Drive folder. This is not an optimal solution because it adds some overhead because I, as the game programmer, had to implement all assets in to Unity. For further development of the game a better solution should be used. For example, everybody is implementing their assets directly in Unity and uses Git and Bitbucket for version control.
I only made single player games before and going to multiplayer was new to me. I managed to find a good tutorial online, but the way player would register for the multiplayer mode was not solved in a good way. Every player registers by pressing any button on the controller, which could result in players being different characters every game round. As seen during the game jam and at some play tests after the game jam, this created confusion between the players, because they sometimes did not know which character they are and thus time was wasted finding it out. We partly resolved this issue by adding the press of a specific button for each player to indicate that they are ready before the round starts, but for a future version this issue has to be tackled again.
Due to the unofficial theme of the game it was clear to us that the main character had to be a raccoon, but for the multiplayer mode we decided to use different animals. Player 2 is a fox, player 3 a goat and player 4 a wolf. We also created different colors for the leaf blowers, but only used the red and blue ones. We planned to make it possible for the player to customize their character by choosing the animal and leaf blowers, but due to the time restriction, we had to ditch this feature. If the development will continue this feature should definitely find its way into the game.
After having problems with the old input system of Unity when using gamepads for different platforms (WebGL and Windows), I changed to the new input system right away, which I knew works perfectly for all types of gamepads. This was true throughout the game development, but as we created the final, and first, builds of the game for WebGL and Windows, we encountered problems with the different gamepads. For the WebGL version the Xbox controllers were not working and for the Windows version the PS4 controllers only worked when using the DS4Windows application. Due to only having 90 minutes left, we decided to not solve this problem. For future game jams and project the gamepad compatibility should be tested as soon as possible to avoid such problems for the final game.
As seen above, there are more positive points than negative ones, which made the game jam a very pleasing and good experience for the whole team. This also played a role in the final voting, which resulted in our game winning the game jam.