Reply to thread

Message: <blockquote data-quote="vikingerik" data-source="post: 151018" data-attributes="member: 3745">I think something went wrong with the math.  The log method is very heavily weighted to the top of the scale (one #1 is worth two #10) so yeah that doesn't sound right.  #30 is worth only 23 points and #50 only 15, so I can't see how that could add up to over 500 total for Inspector42.I still think the log method is more mathematically natural, because it doesn't have any magic constant to fit some artificial predetermined range.  For negative results, just drop them.  But yeah it is very top-heavy so we could use something else.  The PAPA method does look promising, about exactly what we want, mostly linear plus a power distribution for the top 32 which seems just about right to peg the caliber of where exceptional players start.</blockquote>

No members online now.

Total: 63 (members: 0, guests: 63)