I have looked into how they determine model errors vs actual forecasts vs observed and god help you if you don’t have a strong background in statistics. I sorta understand it, and maybe could explain it when I have time to review my notes, but it is not intuitive. I suspect it’s more correct doing it how they do it, but I was hoping to find more localized analysis and that seems not possible unless you posses a login for noaa to access certain information .