cl_smoothtime

cl_smooth "1"    // def. "1" # Prediction error smoothing, if prediction was wrong the model studders. This counters it with smoothing

cl_smoothtime "0.05"    // def. "0.1" # Prediction error smoothing time (0.1 = 100ms)

but I've been asking around and when I've heard everything from 0.01 to 0.05. Anyone know what it does exactly and what the best setting is for it? I know it's probably like cl_interp where people have varying personal settings for it but any tips would be appreciated. By the way, when you google it you get 0.01 as the most used option.

Thanks for your help!