Hi,
This glicko2 rating system I do not like.
According to fide system people would gain more points.
In my personal games I win very few points and suddenly I lose quickly more.
According to fide calculator:
http://ratings.fide.com/calculator_rtd.phtmlI would gain +19 points with a victory but with glicko 2 I hardly win 9 points.
(2223 my opp 2175).
Very annoying. So nobody wins with this system.
Well, that's not true, actually.
As they point out, in QUICK and BLITZ, the K factor is 20.
A 2223 that wins against a 2175 with a K factor of 20 gains 8.6 points, so quite in line with what happened here.
Now, that they lined up so nicely is a bit of a coincidence, but there's nothing wrong with glicko.
If anything, glicko makes more sense than the FIDE implementation, actually. Most notably, the FIDE implementation doesn't consider the certainty of the rating; you're either provisional or not. However, if I play and win against a 2100 who has played 15 games over the last year, my rating should change less than if I play and win against a 2100 who has played 100 games in the last week, because we have a lot less information about the rating of the first player.
Glicko considers the uncertainty of the rating, while FIDE's implementation doesn't, which is a big advantage to glicko.
Theoretical considerations aside, there's a simple solution: don't obsess so much over your rating, and play just for fun. Chess is much more enjoyable that way :)
Glicko is an improvement on Elo and Glicko 2 is an improvement on Glicko. The fact that you would gain fewer points in one situation than you would in a more simplistic system is really not a good reason to dislike the system.
Basically what #2 and #3 have said.
We used to use Elo, like 2-3 years ago. Changing to Glicko-2 has been fantastic as far as I'm concerned. Actually, the fact that you gain/lose less points in a single game is fantastic as it means your rating is far more stable. If you gained/lost close to 20 points in a single game like in your example, in less than a handful of games your rating could fluctuate over 100 points. That would be terrible and is precisely the problem that we are avoiding by using glicko-2 in the first place.
The problem here is it is too stabe. U can hardly make any progress.
U win 5 games and lose 2 . The result is nothing!
Ok. This is not fun.
When we become old players, ratings hardly increase, new they increase by 100s, the by 1,2,3s. That's not fun.
The purpose of having a rating isn't for "fun" it's to determine relative strength so you can match against people that are approximately your own skill level and thereby have more fun. If you need bells and whistles and trophies and points and XP and shiny coins and all that stuff in order for playing chess to be fun, then you are on the wrong site.
Ok. Then make a total different rating system. Like on
www.flyordie.com/chess, where from zero is null point and the absolute max is 1000(only engine can reach it).
Or just use the one that we're using because the Glicko systems have been the standard for online chess. It's been used by FICS for years, and
chess.com as well. Even the Australian Chess Federation has adopted a version of the Glicko2 system. The reason it's used is because of the very stability and confidence it has in the ratings.
Mind my language, but what the fuck?
"The problem here is it is too stabe. U can hardly make any progress.
U win 5 games and lose 2 . The result is nothing!
Ok. This is not fun."
This logic shows absolutely no regard for what a rating system aims to achieve.
It also shows that you're myopic towards your matches against weaker players and ignore the fact that it's the opposite way around when you are playing stronger opponents.
I'm sorry, this thread is getting closed as there is absolutely no credit in its argument.
