I am trying to stay sane. I’m just perplexed. I looked at the top 200 WD clears (of the total 8,000 leaderboard era GR clears) worldwide currently that would grossly overestimate the “average” scaled player at 5K if this is true (specifically these players represent “the highest level of play” as Nev stated). I agree.
You do not need to scale the data if you have actual player data @ 5K. Even with this current leaderboard where several new records were set since early December, the actual WD average for players (n=30) between 4.5K and 5.5K paragon is GR 128.0 in era 12/non-season.
I have reread Nevalistis’s post too many times. It definitely is an issue to reproduce a method given limited details and lack of knowledge of what data points were considered.