| |
BGonline.org Forums
Why create a system we know is flawed?
Posted By: Phil Simborg In Response To: Proposed World Ranking System (Ken Larsen)
Date: Sunday, 17 November 2013, at 4:51 p.m.
If you took the top 20 players and put them in a room for a month and had them play each other and recorded all their matches and compared w/l records and PR you probably would not have enough data to accurately determine the order of the players. So how could we possibly do that with the very spotty and limited information we can gather from live play? I would have little faith in the accuracy of any system that says that Player A is better than Player B if the two players are reasonably close in skill.
I would put faith in a system that delineates player by "skill category," however. Using w/l records and PR, I think it is possible to be reasonably accurate to say that a Player A is a Level 1 player, and maybe the 20 players in the world with the best statistics are Level 1, and the next 40 are level 2, and the next 60 are level 3 and so on.
This would be fairer, kinder, and more accurate. Of course there will be some at the bottom of Level 1 that maybe should be Level 2, and at the top of Level 2 that should be 1, but if these categories are updated, say, monthly, no one will be “unjustly” rated for very long, and those who are truly on the edge will probably shift back and forth over time.
My point is, why bother to institute a ranking system we know is flawed and probably inaccurate when we can do a more general one that is far less flawed and far more reflective of a player’s ability? Yes, this method will not tell us that Mochy is better than Falafel, but any method that tries to do this is not very accurate, and to say they are both Level 1 players is quite accurate and fair (if the data supports this).
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.