Ratings...who cares anymore?

mepbmlist
#1

> How? A newbie with 0 games under his belt has a 1500 ranking. An 8
> year veteran that is 2 of 4 vs. other 8 year veterans has a 1500
> ranking. Stomping a newbie in his first game gets the same "boost" to
> your ranking as narrowly beating the vet that hasn't gone out of his
> way to boost his ranking.

Except games are not played against one player. Straight win% is
meaningless because they quality of your opponents isn't included. We
might want to tone down the starting rating, but it seems that you are
more interested in getting a high rating for never playing than you are
for getting a more accurate rating. A player who is 3/4 is a better
player than one who is 1/1, if only because he's proven himself more.

    jason

Incorrect assumption. What if the 3 of 4 guy always playes Harad or Corsairs, and jumps on the same team that the majority of the other neutrals joins. Now take the 1 of 1 guy that played a grudge game against 8 year veterans. Instead of 5 hours per game per week, he spent 20 hours a week on one game. And his team won.

There are far too many variables in this game for any ranking system to be an accurate reflection of the player's skill. As a result, any system will have flaws that can be exploited by "cunning" players that want to have a high ranking to rub in the face of those that play for fun.

Rankings are a bad idea. The idea needs to be laid to rest.

Darrell Shimel

···

_________________________________________________________________
Surf the Web without missing calls!�Get MSN Broadband. http://resourcecenter.msn.com/access/plans/freeactivation.asp

#2

corsairs game 101 wrote:

Incorrect assumption. What if the 3 of 4 guy always playes Harad or Corsairs, and jumps on the same team that the majority of the other neutrals joins. Now take the 1 of 1 guy that played a grudge game
against 8 year veterans. Instead of 5 hours per game per week, he
spent 20 hours a week on one game. And his team won.

First, your ranking doesn't address these issues either.

Second, just because someone spent 20h/wk on one game, does not make him a superior player.

Third, no ranking can possibly be perfectly predictive, if only because the past does not force the future.

Once again, your only point seems to be that rating are imperfect, and therefore you want no risk of your 1 game/year being possibly marred. That seems very short-sighted.

There are far too many variables in this game for any ranking system to be an accurate reflection of the player's skill. As a result, any
system will have flaws that can be exploited by "cunning" players
that want to have a high ranking to rub in the face of those that
play for fun.

Which is not a reason to completely abandon them.
If players abuse the ratings, they can be changed. Throwing the baby out
with the bathwater is not a solution.

      jason