But thats not what Giladis said. And you did not formulate it clear enough too, how much does team tournies data weight in external balance analysis.For me its important because as far as I know team tournies totally dominate the scene, contributing to the vast majority of results. Am I wrong in this assumption?
Why are you asking me for such specifics? You know that I have had nothing to do with the last update. I have literally no idea what tournament data was used and exactly how it was weighted. Maybe you should volunteer to help the data team by providing a public summary of what was done and why?No, he did not. I'll repeat my question: how much does team tournies data weight in external balance analysis? Compared to singles?
I was just talking generally from my previous experience, because the constant arguments over whether or not the project should use team data are tiresome and fundamentally stem from most people not actually listening to what the project says it does.
To my recollection, from the times I was involved, there wasn't a "singles count for 60% and teams count for 40%" or any such thing (this is ultimately not that much different from just throwing all the data in one pot to start with, which is certainly against what I've always advocated).
It depended on each update, exactly what the goals were, and what datasets were available.
Speaking as a scientist, one should never blindly combine datasets without understanding their properties first.
This is why I have always advocated performing separate and combined analyses and examining the similarities and differences, before best deciding how to proceed.
The post was edited 1 time, last by DanT ().