ELO balances at the end of an LTD game?
Posted: Fri Apr 19, 2013 2:47 pm
To the best of my knowledge, the ELO w/l of a game is re balanced at the end of the game, not at the beginning. Is this true? And if so, can it be changed?
Here is my argument:
*These numbers are mathematically correct for my arguement*. West team has 1600,1000, 1000, 400 ELO players for an average of 1k ELO. East team has 1000, 1000, 1000, 1000 ELO players for an average of 1k ELO. This game starts at +15/-15 ELO w/l. The "common" argument is that the 1600 ELO player is taking a risk with the 400 player on his/her team but is rewarded with a higher ELO gain if his/her team wins.
HOWEVER:
Now, lets say that the 400 elo player quits on wave 2. Now the West team has an ELO average of 1200 and the East team remains at 1000. The West teams ELO w/l numbers are now readjusted to +13.64 / -16.36. Now the game is a 3v4 disadvantage for the West team AND with a loss, the West team will lose MORE ELO than originally stated on the website before the game starts. If the West team pulls of a very difficult 3v4 victory, the team members are rewarded with LESS ELO than originally stated on the website.
(I have talked to many high ELO players where the "leaver situation" results in losing 28 ELO vs the proposed 18 ELO that the website said was on the line. These numbers get substantially more skewed with players of 1800+ ELO rating.)
Does this make sense that a 3v4 game results in a HIGHER risk to high ELO players? It is because of this flaw in the ELO system that you see a lot of players dodging games and trying to !votekick low ELO players before a game starts. There is actually far more risk involved in playing with a potential leaver and the reward for winning such games are far lower than most people realize.
Do the admins have any thoughts on this system and possible changes to it?
Here is my argument:
*These numbers are mathematically correct for my arguement*. West team has 1600,1000, 1000, 400 ELO players for an average of 1k ELO. East team has 1000, 1000, 1000, 1000 ELO players for an average of 1k ELO. This game starts at +15/-15 ELO w/l. The "common" argument is that the 1600 ELO player is taking a risk with the 400 player on his/her team but is rewarded with a higher ELO gain if his/her team wins.
HOWEVER:
Now, lets say that the 400 elo player quits on wave 2. Now the West team has an ELO average of 1200 and the East team remains at 1000. The West teams ELO w/l numbers are now readjusted to +13.64 / -16.36. Now the game is a 3v4 disadvantage for the West team AND with a loss, the West team will lose MORE ELO than originally stated on the website before the game starts. If the West team pulls of a very difficult 3v4 victory, the team members are rewarded with LESS ELO than originally stated on the website.
(I have talked to many high ELO players where the "leaver situation" results in losing 28 ELO vs the proposed 18 ELO that the website said was on the line. These numbers get substantially more skewed with players of 1800+ ELO rating.)
Does this make sense that a 3v4 game results in a HIGHER risk to high ELO players? It is because of this flaw in the ELO system that you see a lot of players dodging games and trying to !votekick low ELO players before a game starts. There is actually far more risk involved in playing with a potential leaver and the reward for winning such games are far lower than most people realize.
Do the admins have any thoughts on this system and possible changes to it?