Machine Learning Model trained on 1v1 Multiplayer Ladder Games and Server Data

Projects Personal Projects
1

Recently, I requested data, which was granted to me (ask before you scrape!), from the Ladder server of an open-source multiplayer game.
I have my own ideas but I’d also like to ask the community for suggestions regarding what machine learning models I could possibly train from the data.
Here are the columns from the tabular data containing the ladder statistics where each row represents a 1v1 game. What would be nice to classify or predict?

image
image549×754 16.7 KB

I also have every single replay of each of those 74358 rows of games, which has the potential to be datamined.
I feel like this is a great natural next step for me after Codecademy’s OKCupid project because the data set (well at least the existing tabular data) is similar in the number of rows and columns, and also has columns which can be NLP candidates (the winner and loser comments).
Also, I wonder if this piques the interest of Data Scientists here.
@eismaus

1 Like
2

I’m definitely interested! I’ve been meaning to start working in a bit of reading about machine learning models for game theory, and this will be a perfect excuse.

I’ll be in touch over Discord, as well as keeping on eye on how much interest is showed here.