cesaregarza / SplatTop

4 stars 1 forks source link

Clarify Gini coefficient #16

Open zhxie opened 1 month ago

zhxie commented 1 month ago

The Gini coefficient is 0.6132, indicating a slight imbalance in analytics. But in my education, a coefficient of 0.6 is generally considered relatively high. I also asked ChatGPT and found that a Gini coefficient below 0.3 is indicative of low inequality, while above 0.4 suggests high inequality.

cesaregarza commented 1 month ago

You're not wrong, a gini coefficient of 0.61 is definitely extremely imbalanced and in economics would be considered among the worst in the world for income inequality.

I made the decision to slightly bias the interpretation of the gini coefficient on the website to reduce the amount of panic players might exhibit. I don't really have a justification for this, and I might change the values to be more reflective of reality.

zhxie commented 1 month ago

Okay, I also calculated the Gini index of Splatoon 2 X rankings after the final update before the release of the 3, and found that the value is 0.6775, even larger than the 3.

In my mind, weapons in the 2 are more balanced than the 3, so maybe the Gini index cannot prove something that I thought.

cesaregarza commented 1 month ago

I wonder how much it’s being affected by the zero usage weapons. I had considered calculating gini throughout the history of the game before, I’ll probably move that up. I’ll also ask top players what they felt the most balanced meta was and calibrate the interpretation based off of that.

zhxie commented 1 month ago

Yes, I also did a further look into that. I added weapons with zero usage and merged Hero/Octa Replicas with the original ones. The calculated Gini index is larger than 0.7.