If specificity is what you are looking for; unoffically post 12, officially 15
Next, I *did* post my thoughts on the first couple of pages. They just happen to be brief thoughts for the moment.
Ok, because that vote looked very RVSy. Im assuming it was actually serious then for the reason that ML said?
Thinking I may have mixed up my calculation too when I started trying to figure out the inverse of it. Not even sure its really useful information though in the end without a bigger data set. Meta is always a little touchy either way but I guess behavioral is better than how its applied most of the time, although I still say meta basically taps out at being good for determining player strength.
Yeah I messed up probability when thinking about it since I should be taking 10 game sample at .25 chance of being scum in each game, and then calculating chance of five (or more for completion sake) scum flips which gets dropped to 8%. Might actually be viable but I would want to see counterexamples for games he is town and does/doesnt RVS first. Think this is just going to start getting in the way of discussion though so im doing to drop it at my math correction.