Pong
Version 1: Pong Remade
Failed First Attempt
Despite reading the introduction on The 20 Games Challenge website, I still lacked a clear understanding of how to begin the implementation process. Therefore, I decided to follow a tutorial video in order to recreate the game step by step. The website provides six embedded tutorial videos, one of which is private. I chose the one posted by the creator with the highest number of subscribers (5.41K on YouTube) as my starting point for learning.
During the learning process, I found that the tutorial video I chose primarily focuses on “how to do” rather than “why”. It shows the tools and steps needed to achieve specific results but provides little explanation of their functions or when to use them. As a result, I spent significant time simply locating tools and understanding basic operations such as adding, copying, and pasting elements. For example, when addressing the paddle movement mechanic, the tutorial only presents the final code without explaining how the code was developed or the reasoning behind its structure.
During the coding process, I encountered an unexpected issue. Despite following the tutorial exactly, my result differed from the one shown in the video. Strangely enough, the effect I created seemed to skip ahead and directly achieve the result that the tutorial intended to reach in a later step. However, this surprising success introduced a new problem that was not covered in the tutorial. This left me extremely confused. I had no idea what I had done differently, nor did I understand why I encountered an issue that never appeared in the tutorial. It highlighted a limitation of tutorial-based learning: while tutorials often demonstrate ideal outcomes, they often fail to prepare learners for unexpected challenges and provide limited insight into the underlying logic, making it difficult for learners to diagnose and resolve problems independently.
Finally, I decided to seek assistance from AI. Although specialised coding tools such as Cursor have become increasingly popular, I chose to use ChatGPT because it was the AI tool with which I was most familiar. I described the issue I encountered and asked for a solution:
I am currently recreating Pong using Godot. While implementing the ball’s bouncing behavior between the two paddles, I noticed that the ball gradually slows down after each bounce and eventually stops moving. Why is this happening?
Although ChatGPT provided several technically plausible explanations, none of its suggestions resolved the problem. I suspect this was because the AI could only respond to the symptoms I described rather than analyse the full logic of my project. Without complete context, its recommendations were limited in effectiveness. After reviewing my work against the tutorial again and still failing to identify the cause, I decided to switch to a different learning resource.
Trying Another Tutorial
This time, I directly searched “Godot remake Pong” on YouTube and selected the first video recommended on the webpage. Compared to the previous tutorial, this one was much more convenient because it directly provided open-source fonts and assets through GitHub. I no longer needed to spend extra time searching for resources myself and could simply follow the tutorial step by step.
This tutorial was much clearer, and by following the instructions step by step, I quickly completed the scene setup in the editor. However, like the previous tutorial, it focused on demonstrating procedures rather than explaining the underlying reasons behind them. As a result, although the task itself was straightforward, I was concerned that I might struggle to work independently in the future, as I had learned how to recreate this specific game rather than how to use the engine's tools more generally.
Once the scene setup was complete, I moved on to coding. This process was both straightforward and challenging. On the one hand, I could follow the tutorial line by line, and apart from a few minor debugging issues caused by careless mistakes, the game ran successfully. On the other hand, the explanations of the code were brief and closely tied to the specific implementation of Pong. As a result, although I was able to recreate the game, I struggled to understand the broader purpose of many code segments and found it difficult to transfer this knowledge to other projects.
