[squeek.]

Originally Posted by beees

Instead of an arbitrary c, why not use a percentage of the total (ie. 1% of all players)?

I'm going to do it (as far as I know) totally correctly, and make the C for each record the global average of the "total" for that record. Like what I said in a previous post:

(global average of time played on offense * global average of scout as offensive class choice + sum of time played as scout) / (global average of time played on offense + total time played on offense)

so C = global average of the "total" and m = global average of the average being calculated.

I've not set it all up quite yet, but I've been testing to see the results, and, as an example, the Average Blue Pipe Kills values currently work out to: global average blue pipe kills (m): 5.3538 (calculated as total blue pipe kills / total rounds of demoman played) and global average of rounds of demoman played (C): 25.6489 (calculated as total rounds of demoman played / number of players)