Graduation Project

Overall thoughts and reflections

I’m doing a game for this biography, and I actually started out with the idea of having a complete experience of what it takes to make an entire game. Because I have never written the code to run a whole game before. So I can call it a hot-headed decision.
I started the production with the planning, I designed a very interesting world view about because modern people are too efficient, thus a company called 2424 was created to allow people to generate money even while they sleep through emerging technologies. In this context people’s lifestyles and perspectives have changed dramatically. Because people’s purest rest time is being capitalised, it actually represents the whole process of people being capitalised. I thought it would be interesting to explore this part of the story, intending to assume that the player is one of the company operators and embody social change through this lens. However, because this was my first time making a complete plan for a game, I focused more on the plot and background, but was a bit vague about the gameplay (because I thought I had thought it through when I was writing it, but realised that what I had planned was nowhere near enough when I was making it). Also, because I’ve never tried to make a game all by myself from start to finish, my estimate of the workload was way off. I went through five or six versions of my plan, and I revised the flow of the first level three times. A lot of the details were not even considered until it was time to make it. To be precise, I only started thinking about it when I was writing the code.

I wasted a lot of time in the planning part due to settling on the story of the game instead of the game mode first, a step that led to a lot of detours later on. Because I wanted to combine both plot and management elements, but I felt that the two were too fragmented in my arrangement. In the first version, I designed the game to have a small level at a time like a traditional management game, with each level having its own independent goal. But then I actually wanted to emphasise the part of being a plot character, so that gave it a very confusing feel. Then I looked at long-term management games like Civilisation and Victoria, and realised that I couldn’t make such a large system and art resources in a short period of time. So I went back to the drawing board, and I added a time system to the game, where the storyline is triggered by time, and then different choices trigger different endings. Of course this was already in the last few weeks, so I wasted a lot of time and resources on this part.

My second step was to try to make some features first, then realised that all I had mainly made upfront was the umg, which is a simple but easy part of the game to produce results. Since I never made a blueprint game before, I spent a long time reading the official ue documentation and the official examples provided. And when I really started to make it later I found that my plan was written in a mess, no wonder the programmers in the game company insulted the plan every day, even if I wrote the plan myself I wanted to travel back in time through a time machine to beat myself up. Suffice it to say that my planner has nothing but variables arranged. And I had to learn to write data, blueprints and arrays. At first I was plagued by a very silly question for a long time: how am I going to interoperate data in different levels. Then I looked through the data and realised that I needed a common instance of the game. And all the logic needs to be offloaded inside this instance, while the main function of blueprints in other levels is actually triggering and referencing. While I’m being very clear now it actually took me almost a week to figure out the logic, and that’s just to figure it out not counting the crafting part. I also crafted the quest system under this principle. I’m sure everything I’ve said should be pretty basic to a dedicated programmer, but to me, a new learner of blueprints, it’s all very interesting! For the first time, I experienced the sense of accomplishment of making a complete system, and I got a better understanding of the internal logic.
I made three levels for the game, shop, indoor, and outdoor, then linked them through game instances and unified the umg. It’s still all function and no playability but I still feel happy that I learnt a lot.

Next came the art resources. I started by drawing a furniture design on paper, then I realised that it took me three whole days to make the chair and it was still asymmetrical, so I immediately dismissed the idea of making all the models myself started downloading free open copyright models online (and then ran into no less trouble at a later stage than making my own models). But I couldn’t find any suitable city materials, so I had to make my own, but since my modelling skills are not so good, I mainly used cubes and faces to build directly, without topology bending complex shapes and a series of processing, I directly determined the location of the main city through a few cubes and started to “build the pieces of paper”. I started to “build the paper” by identifying the locations of the main cities from a few cubes. But because there are too many small cubes and surfaces, my ue crashed, I had to export part of the ue build first, and then re-import it as a whole into ue. the whole process can be said to be super confusing, and in order to make it more like a city rather than a bunch of paper and cubes, I downloaded some wire models for decoration, the effect is excellent haha.
But another problem faced by downloading models online is that it’s difficult to unify the styles. Because of time constraints, I didn’t have time to redo all the textures, so I made post-processing materials to unify most of them. The post-processing material gives it a mesh texture and shadow layering, which adds richness to the overly simple image.

Next came the part of creating the 2d footage. This part also took much longer than I thought it would, I drew it for at least three weeks. Before I made it, I thought the icon would be quick, and the only thing that would take more time would be the cover. But after I made the cover, I thought that the icon on the back should be in the same style as the cover. So I made a simple pattern from ai and then imported it into ps to add and draw textures, I tried dot mapping and hand drawing textures, but after I made it it was very strange to add umg to the game and combine it with 3d, and it was especially cut off from the screen. So I had to redraw a part of the drawing that was made purely in ai, and then layer and slice it through ps. And since I had to merge the layers of the icon and 2d footage I made in the first two weeks, then layer and slice them again, the pixels and edges were a bit rough and not as recognisable as a straight ai vector image. And since my bad planning was always being fixed, my art resource needs were always being changed. Doing it later in order to hurry up I made it quickly using the vector tool directly from ps with tracing, including some of the character artwork in the later updates was also much worse quality than when I had plenty of time in the early stages.

The next step is to modify a lot of trivial details, music, clicks, specific character arrangements, and so on. Each one is small yet all very trivial, including the uniformity of colours for all the fonts. In fact, there are a lot of details I can’t tweak in time, like I said, it’s a working game now, but not a fun game. But this process of going from 0 to 1 has actually made me happy and accomplished and feel like I’ve made something fun.

Of course, in fact, I encountered more than this problem, but if I encountered every problem I specifically describe that I’m afraid that this record will become a 10,000-word long speech, so I only mention a few impressive discovery and modification.

In fact, this whole process I found that I have a lot of shortcomings, for example, I like to set some big goals. My planning was not clear enough, which caused me a lot of trouble, and my time schedule was too tight, which led me to stay up late to make drawings and even got sick. This process has taught me the complete process of game production, not only the art, but also the programming and some simple music. It also taught me that game production is not only about the exciting and interesting parts, but also about the details, the standardisation of production, and the trivial things. I crashed and burned many times while making 2D icons, many times just on making the same icon with inverse colours, different colour schemes, and whether to stroke or not, and layering a few of these small issues over and over again, exporting changes and comparing them. This process allowed me to be more realistic about the specifics of game production, rather than idealising everything. And I also tried to write my own tunes and bgm for the first time in this project! I think I found a lot of deficiencies and realised a lot for the first time in this project, even though the final work is not a complete game, I learnt a lot of things, and I will surely improve a lot next time I have the chance to make it again. And I also plan to continue to improve this project after graduation, and strive for the day when I can actually let people play it from start to finish.

Progress during the first two weeks

In the first two weeks I mainly wanted to make sure I had the ability to make this game and the ability to implement features using blueprints. So I made a demo with no art resources at all. it only has functionality, which is to start the game jumping to select the data store. But it was a pain to write because on the one hand, I’ve never worked with blueprints before, and on the other hand, I’ve always felt that the blueprints way of thinking is the opposite of the way you think when programming. So even though it’s a simple function, it took me a long time to write it.

But the feeling of accomplishment when it’s done is unmatched haha

The first umg I did, I didn’t even know at the time that I needed to include a canvas. Also didn’t know that the viewport for jumps was to be SET before adding it. For a while I thought the problem was caused by the version being too new, but then I finally looked it up and found out that I hadn’t set, and it took me two days and four version changes to realise that this error was my problem.

2d style references and reflections

I found all the reference from the artstation

This is the style I wanted to reference, with simple colours but vivid shapes, but after thinking carefully about the amount of work involved eventually abandoned the idea

Then I started referencing this relaxed line and simple sketching style, using a simple handful of colours

Again I’m referencing this simple blocky generalisation of a bit of a manga style, the idea is that all the 2d stuff has to be as simple as possible as all the ui elements in the game have to be split up so it’s going to be a lot of work.

After referencing and thinking about it I think my 2d style should be like a manga with an emphasis on blocks and lines, at the same time I’d like the hand-drawn feel to be obvious, it would make the image more vivid. Since the whole background of the game is cold and sarcastic considering, I’m thinking a style like a satirical comic would be appropriate

So I started by creating this main menu screen! I particularly like this image, I wanted to express the reality of when people’s purest downtime is used to generate value. So I drew a lot of clocks to express this, mobile ones, planet-like ones, and digital ones, plain ones. For the colours I chose blue as the main colour, wanting to give the feeling of a cold future.

But I’m having problems making the ui inside the game, I’m making the ui buttons in the style of the main menu, but the combination with the 3d inside the game is very strange and cut off. Instead simple images are more appropriate. So I had to recolour a lot of drawn icons. Later even made them directly from vector graphics. But there are still a lot of problems, such as after recolouring the edges are very strange, and some of the icon size and degree of refinement due to my time schedule in the early stage of a lot of drawing is very delicate, and later gradually minimalist, so that led to some cut style.

Also I tried a lot of different texturing methods while making it, I used dot mapping, brushes for square textures, as well as comic book textures and my own hand-painting, so it looks a bit dissonant. But after I actually used it in the game I realised that all these details would be ignored and instead the overall structure and level of detail was more important

(I also tried to put a cot shape on the body butter, which ended up looking like an unidentified pattern because it was so small, which made me laugh)

3d style references and reflections

These are the 3d styles I’m referencing, I want it to be simple with an emphasis on geometric forms

This style fascinates me, I love the split shading and vintage textures. Since my models will be simple, I want to make my visuals a bit more varied to make the images more interesting

So I made my post material