My 2c on the scaling / high scores issue: It seems like the only continuous measure of skill is WPM, and the problem seems to be that maximum WPM varies enormou

My 2c on the scaling / high scores issue: It seems like the only continuous measure of skill is WPM, and the problem seems to be that maximum WPM varies enormously across the population. It would be interesting to see if making score proportional to log(letters typed weighted by WPM they were typed at) helps balance things.