Can I become a titled player soon? • page 5/5 • General Chess Discussion • lichess.org

@Anant06 said in #38:

@Sleepy_Gary You say I don't know what a mathematical proof is? I came 3rd in my math Olympiad in my entire state (I'm in 10th grade)

Congratulations!

@Anant06 said in #39:

I'm not saying it is a mathematical proof. But if the RD value is better in Glicko-2, it's quite logical that Glicko-2 scores better. And anyways, I wrote many other things in my comment #18. But you guys only seem to hear one kinda sus thing in the whole comment. Not like I provided detailed info about Glicko-2 and Elo rating system, which apparently many people posting in this thread didn't understand.

The challenging point is "the RD value is better in Glicko-2". Elo does not have the concept. So it isn't a case of whether or not it is "better", it is a concept that one system has, and one system doesn't have. It may or may not be good.

Anyway, I have done some more digging, and:

It seems that Glickman's intuition is actually correct and mine was wrong, FIDE did some analysis to demonstrate this.
A Glicko variation failed in "real world" tests for various reasons; a relatively inactive player returned to rated chess, and due to winning his first few games back a player ~ 100th in the world suddenly found himself in the top 10. It was one unfortunate case, which doesn't render the system as a whole redundant, but if it were to happen soon after adoption, the system would quickly fall into disrepute even if the underlying methods are statistically valid.
In England, Glicko was rejected on the basis that the added complexity was unnecessary for the minimal benefit that was gained. (They recently moved from a 3-figure rating to a 4-figure rating, so were actively investigating it.)
The assumption of the additional Glicko variable is that time/activity has a linear relationship. It seems more likely that it isn't a linear relationship, but it is worthy of further investigation.

@Anant06 said in #38: > @Sleepy_Gary You say I don't know what a mathematical proof is? I came 3rd in my math Olympiad in my entire state (I'm in 10th grade) Congratulations! @Anant06 said in #39: > I'm not saying it is a mathematical proof. But if the RD value is better in Glicko-2, it's quite logical that Glicko-2 scores better. And anyways, I wrote many other things in my comment #18. But you guys only seem to hear one kinda sus thing in the whole comment. Not like I provided detailed info about Glicko-2 and Elo rating system, which apparently many people posting in this thread didn't understand. The challenging point is "the RD value is better in Glicko-2". Elo does not have the concept. So it isn't a case of whether or not it is "better", it is a concept that one system has, and one system doesn't have. It may or may not be good. Anyway, I have done some more digging, and: - It seems that Glickman's intuition is actually correct and mine was wrong, FIDE did some analysis to demonstrate this. - A Glicko variation failed in "real world" tests for various reasons; a relatively inactive player returned to rated chess, and due to winning his first few games back a player ~ 100th in the world suddenly found himself in the top 10. It was one unfortunate case, which doesn't render the system as a whole redundant, but if it were to happen soon after adoption, the system would quickly fall into disrepute even if the underlying methods are statistically valid. - In England, Glicko was rejected on the basis that the added complexity was unnecessary for the minimal benefit that was gained. (They recently moved from a 3-figure rating to a 4-figure rating, so were actively investigating it.) - The assumption of the additional Glicko variable is that time/activity has a linear relationship. It seems more likely that it isn't a linear relationship, but it is worthy of further investigation.

Thank you! @Anant06

I just love it when people treat 2000 lichess as if it's a great milestone or something. Go play on ICC and I bet you can't beat 1600s there. You have no idea how far you are from an actual titled player in terms of positional understanding and other aspects of the game. I'm not writing it to discourage you but the first step to improve is to know your weaknesses and shortcomings and set sensible and realistic goals for yourself.