ntk wrote:Pier u r the man of numbers... and don't understand u is back rm... no sorry... it so simply know who is... look his stat 77% see on list the same rank (same rank, same AI)... it's so easy
Anyway just to give you some numbers that having an high ratio of victories with few games does not mean too much.
Imagine that someone, second account or just quick to sort out problems, arrives in league 4 with 100% ratio of victories and 50 games.
Now, since the guy is strong, his real winning ratio in league 4 is 65% . To accumulate enough games to lower the winning ratio from 100 to 66 he has to play the following amount of games that we are going to compute.
So we have: x / (x+y) = 0.65 , where X are the victories and (x+y) the total number of game played in league 4 . The ratio is 65%.
So we want : (50 + x) / (50 + x + y ) = 0.66 since the player already played 50 games with 50 victories.
Since we know that
x / (x+y) = 0.65
x = 0.65 (x+y)
0.35 x = 0.65 y
x ( 0.35 / 0.65) = y
(50 + x) / (50 + x + y ) = 0.66
50 +x = 0.66 (50 + x + y )
50 + x = 0.66 * 50 + 0.66 x + 0.66 y
50 * 0.34 = -0.34x + 0.66y
50 * 0.34 = -0.34x + 0.66 * ( 0.35 / 0.65) x
now I approximate because it gets ugly
50 * 0.34 = 0.0154 x
50 * 0.34 / 0.0154 = x
x = 1104
y = 595
So it means to "see" a value similar to the real winning ratio in league 4, the player has to play round about:
50 + 1104 + 595 games or 1749 games.
This is valid for every player that reach league4 with good winning ratios but actually the real league4 winning ratio is definitely lower.
So rm-rf or also other players (nullpointer for example) may have way lower real winning ratio but we will see them only after a lot of games. For this the elo is cool if well tuned, because it shows with some dozens of games the approximated "relative" strength of a player.
PS: today met rm-rf / 4 times, 3 won, 1 draw.