The Big Picture

The Pokémon Company holds one of the largest media franchises on the planet, in which games are a major part. These games have become much bigger in scope and globally more competitive, resulting in tournaments with large prize pools.

The Problem

Balancing the game's multiplayer aspect has become a difficult topic. There is a general upward trend across all generations. This is known within the gaming industry as power creep. Newer content is stronger so that new releases and merchandise are made more attractive to potential customers. Much to the detriment of the competitive scene, which suffers under these balancing issues.

A (potential) Solution

A tool that forces any newly created Pokémon's base stats to be restricted within the constraints of previous generations. The strength of a Pokémon competitively depends on many factors, but within this project the base stats serve as a starting point.

The Technology

I programmed the core of this module with Python and built a frontend using Django. All of that was condensed into a Docker container, which I then exposed through nginx so that it can be publicly accessed.

The Conclusion

I still find it an interesting approach to balancing, and the project successfully demonstrates an example of this. The novelty of using predictive models not just for predictions, but to keep stats of gaming entities in line, could potentially counteract meddling through human biases or bad assignments caused by oversight.

But in all honesty, no profit-oriented developer would likely consider this viable. Power creep sells new content, skins, and merchandise. This approach would be useful to indie developers, but they can only use it once they build up a backlog of stats that can be used for training.