Posts:
5
Registered:
Jul 1, 2008
From:
California
Age:
20
Ratings Change in a Draw
Posted:
Nov 10, 2009 1:16 PM
I have seen a few forums where the effect of a draw on turn based ratings was mentioned but no specific forum about it.
I feel that a tb game resulting in a draw should still affect the ratings of the two players. To me, a draw between two players shows that they are of similar ability and skill. Players with similar skill levels should have similar ratings, right? That is the purpose of ratings in my opinion. So when two players tie, i think their ratings should move closer together. For example, if a player with 1800 and a player with 1700 play a rated tb set and the set results in draw, their ratings should change to something like 1780 and 1720 respectively.
I know some people are opposed to this and I would like to know why. Thanks.
Posts:
474
Registered:
May 9, 2002
From:
Northeast USA
Age:
31
Re: Ratings Change in a Draw
Posted:
Nov 11, 2009 11:15 AM
I don't like this idea. The reason is that an individual game of pente intrinsically leans very far in Player One's favor, regardless of the skills of the players. The expected result is a win for Player One. By playing a set, the "match" is evened out so that if one player or the other wins the match then ratings should clearly be adjusted. But when a match ends in a draw, that's just to be expected and neither player likely did anything special for this to happen. In your example, perhaps the 1800 player completely crushed the 1700 player as Player One. But as Player Two, the 1800 player put up a very good fight and after a long and complex game he still could not overcome the inherant Player One advantage and still lost. This player should not be punished for a result like this.
Here's another way to look at it. When agreeing to the match, the 1700 player knows he only has to win one out of two, which is to be expected, to gain points. So, he'll always request matches like this. But, WHY would the 1800 player ever agree to this match? He must win both games, including one where he has a large disadvantage to gain points -- if he wins one out of two he loses. So, he might never agree to this match. The way it is now, after the first game is over, there is still something to play for. The loser of the first match is playing for either a loss in rating, or a no change -- the winner of the first match is playing for either a no change or a gain in rating.
What you are suggesting is much more like how live play is already handled at DSG, which is bad. It is the reason players rated 1900 and over rarely play any more -- they simply do not want to play a set against anyone rated lower since they know that most of the time they will split the set and they will lose points. This should be fixed in the live play section immediately and it should be left the way it is in the turn based section.
Re: Ratings Change in a Draw
Posted:
Mar 8, 2010 7:28 PM
I do like this idea. I definitely disagree with the statement that "it is to be expected (that white wins)". For someone, like me, who hasn't realized yet how white has an advantage and how to (fully) exploit that, it is obviously _not_ to be expected that white wins, the difference in rating reflect that imho. And the very frequent and sustained whoop-*** I get when playing stronger players and losing as a white player support that.
I can follow your reasoning in the example about the 1700 and 1800 players, but that does not hold when the white players lose.
boobies, math, pente, climbing and digital wireless things, let's fire up those neurons
Posts:
1,193
Registered:
Mar 4, 2007
From:
Age:
31
Re: Ratings Change in a Draw
Posted:
Mar 10, 2010 5:37 PM
the little fluffy white cloud does not know why it moves in such a way in such a direction, but the sky beyond the cloud knows why.
we do not require you to know what we know in order for us to expect Player one to not only start with the winning advantage, but also to maintain it to the end contingent upon white not making a single error along the total path. ask any red player and they will tell you the same, that white has sure win from K10 moving forward in perfect play. whites ability to move first is a advantage by simple math of having a higher number of stones to work with on the board. nosovs performs perfect play the most of us humans. try looking at his P1 loses vs his P2 loses. P2 can only win when the human Player 1 makes a mistake in understanding the laws of initiative and position which are actually easily both just summed up as initiative in the end.
