​Post mortem

Type: Product 
Genre: Side scroller/platformer
Engine: Unity.
Released: 11/6/2025.
Last updated: 13/6/2025.
Learned from: Originally from Diploma of Digital and Interactive Games (2018), but recently decided to expand it.

Before moving further towards my game dev journey, I wanted to exercise my game and level design skills. I decided to dig out one of my old assessments from my hard drive and expand on it instead of starting a new project from scratch. I also wanted to make a project that can demonstrate the 3D assets I created and my creativity in action.

Originally it was going to be a quick exercise, but due to some issues, it end up taking almost a year to expand it. I had to make sure the game was polished and running as smooth as possible.

This time: I didn’t code as much, and a few of my 3D assets are in the game such as the Glue Cannon, Lightsaw, Security Bot, Security door, and Giant pistons.

Originally the project had 1 level and few sounds. Now the new project has many sounds,9 levels, main menu, and a finish scene.

WHAT WENT WELL?:

• Despite some of the 3D models have a higher poly count than others like the Pistons, I was able to make the levels short and run above 60 fps as possibly.
• I managed to make and edit particles respectfully like smoke, fire, Glue Cannon particles and sparks really well due to what I learned about making particles last year. Doesn’t include electricity sparks for Tesla coils since I followed a tutorial to help me make it years ago.
• For the first time I added a level select menu and a button that takes Players to my home game site.
• Despite doing minor code changes, I managed to work around the code that my past programming teacher did. For e.g., I tweaked the checkpoint system to make a sound once when the Player collides with it instead of it sounding multiple times. I changed code for the big button by the Security Door to only press down and activate when a crate has collided with it instead of the Player.
• I worked around making sounds activate within the animation instead of code. Good e.g. is the Security doors, once the door activates the animation, it enables a Game object to play a machine moving sound. Once it reach’s the end, It disables the game object and enables a new Game object to play a machine stop moving sound.
• I created a endless loop of the player’s character running animation on the platforms on the main menu with the help of a project I made few years ago called Roller ball.
• I edited some sounds with Audacity to fit with the game as possible.

WHAT I LEARNT:

•    Not much new things I learnt this time, but one thing is I have to think twice before working on a new project. I need to think about... - What good would it do for me? - What new things I will learn? - Is it worth taking months or years on the project.

WHAT WENT WRONG AND WHAT COULD I HAVE DONE BETTER?:

• I didn’t realise at the last second that the Pistons had big poly count, I had to tone down the amount of Pistons in the levels so it doesn’t affect the frame rate. I could have reduced the Polycount of the pistons, but it was already animated and I didn’t want to risk  anything happening to the Pistons like it not animating or something like that.
• I did a lot of breaking the prefabs that were attached to what I made were level prefabs containing all the 3D assets each level, next time I will not make level prefabs and just make a group contacting each type of prefab like I have done before. Once the prefab is broken, I had to manually update the prefab nearly each level rather than clicking “apply” each prefab .
• The Player jumping on Pistons is a little sluggish, I could tweak the jumping code to be more responsive, except its not my code, so I don’t want to be breaking any major code. I may need to make a mechanic in future that the Player can still jump and move well on moving platforms.  
• I tried to do Occlusion Culling, but it only seem to do minor improvements with the fps, Maybe I didn’t do it correctly or the assets already in the game are not designed to be LOD and blend with Occlusion Culling. I will need to understand it more in order to make noticeable fps improvements.
• Making the short trailer for the game for the first time was a last min result, I need to plan and prepare more of what the trailer may showcase, sounds, music, possible cutscenes and what sells it compared to its similar recent competitors.
• In the trailer, I don’t think I recorded the right wxh resolution of the video which made the it look outdated, I will need to adjust the record settings to fit on modern wxh resolutions that will fill the entire screen.
• I accidently recorded the Mouse in the trailer, I should have the game turn the cursor off or move the cursor off screen and not touch it until recording is done.
• Some sounds that that should be heard first (e.g. touch Finish flag), should have a bigger priority to be heard than other sounds, I will need to make sure sound priorities are set properly in the future.
• I noticed they were missing faces on the Piston, at the time I thought players wont see that error, but I was wrong. I am going to let it go this time.
• I think I should have spent more time getting players parameters and Pistons polished, it has few minor issues that cause the movement including jumping to be sluggish. I spent most of the time making levels rather than Player movement testing. I should REALLY spend more time with the core mechanics like Player movement as possible.
• I could have turned the Gear and tank HUD off since it is not used and wont mean anything to the Player, but I decided to leave it there because I thought it fits with the Industrial/Dystopian theme. Next time if the HUD is not going to be used, I will remove it from the game completely.

OVERALL:

  • Despite this taking longer than expected, I learnt what I can do better next time, things I should consider improving, and what not to do.

• It was a fair good exercise, but I do not want to dawdle any further, since I’m still working towards a commercial product rather than a free product.
• This project is a definite reminder for me to prioritize Core mechanics rather than level mechanics since I have been building test projects for a long time.