Yeah, I'm not a math guy. The goal was to weight mutes against follows in a reasonable way. Someone suggested I use tan instead. The nice thing is no one has to use the same formula.