Life In 19x19
http://www.lifein19x19.com/

Rank Comparisons
http://www.lifein19x19.com/viewtopic.php?f=10&t=3724		 Page 1 of 2
Author:  TruthTaco [ Sun Apr 24, 2011 1:42 pm ]
Post subject:  Rank Comparisons
I was wondering if there has been any recent rank comparisons done. I am a KGS 8k, and I was interested in finding what my rank might be on DGS, (without having to play month long games). Also, why can't people just use bots for rank comparions? Shouldn't bots play the same strength on all severs?
Author:  amnal [ Sun Apr 24, 2011 1:48 pm ]
Post subject:  Re: Rank Comparisons
TruthTaco wrote:
I was wondering if there has been any recent rank comparisons done. I am a KGS 8k, and I was interested in finding what my rank might be on DGS, (without having to play month long games). Also, why can't people just use bots for rank comparions? Shouldn't bots play the same strength on all severs?


http://senseis.xmp.net/?RankWorldwideComparison is a comparison table such as you request. Bear in mind that this kind of exercise inherently has massive errors, so you should expect large errors on any rank it suggests. Also, it's probably rather out of date.

Using bots for rank comparison is theoretically possible, but comes with a host of problems. An obvious simple one is that you'd need to equalise the time limits, which is not easy between a turn based and real time server. Another is that you'd need to do this in a controlled way (people learn how to beat bots, because bots don't learn, etc.), which would probably need quite a lot of games and quite a lot of someone's time and effort.
Author:  TruthTaco [ Sun Apr 24, 2011 2:05 pm ]
Post subject:  Re: Rank Comparisons
yes I'm aware of that one, seems out of date :(
Author:  karaklis [ Sun Apr 24, 2011 5:08 pm ]
Post subject:  Re: Rank Comparisons
I have created a new one based on 2010 data.
You can find it here
Author:  emeraldemon [ Sun Apr 24, 2011 6:53 pm ]
Post subject:  Re: Rank Comparisons
FWIW I think you can say roughly DGS = KGS - 2, and it will be accurate at most ranks. I'm 2k KGS and 4k DGS, for example. I play with a 3.5d AGA who is 1d DGS.
Author:  brodie [ Tue May 03, 2011 8:43 pm ]
Post subject:  Re: Rank Comparisons
Ive been wondering about this myself lately, as the disparity between my kgs rank and dgs has been making me think ive gone crazy. I play in binges, so my play (and rank) fluctuates, but right now I am an 18 on DGS and a 12(?) on KGS. There are certainly some things that effect this in my case:

- I tend to play better by intuition and instinct, so a fast real time game is good for me. On DGS, however, my opponent is allowed to play as slow as they want, and I know that several of my regular opponents will spend agonizingly long amounts of time studying their moves. I dont have the patience for this, and still play fast, so I end up losing a fair amount of life and death problems to people that are not as strong.

- I tend to play the same people over and over on DGS, so if there numbers are skewed, mine will be too.

I have wondered how much DGS is skewed because of the fact that people can study moves as long as they want, and Im sure there are unscrupulous folks that play out sequences from the game. I would guess that you have to be more honest in a real time game.

So i spose its due to my style, differences in play, and differences in handicapping systems orranking algorithms, but im still surprised to see the difference. I beat a 12k on DGS so badly tonite that I started to wonder if he was artificially padding his rank, but when i checked his graph, it seemed legit. go figure.
Author:  Fedya [ Thu May 05, 2011 5:54 am ]
Post subject:  Re: Rank Comparisons
Quote:
I have wondered how much DGS is skewed because of the fact that people can study moves as long as they want, and Im sure there are unscrupulous folks that play out sequences from the game.


What's unscrupulous about this? Isn't one of the points of turn-based servers (or the old correspondence equivalent) that you can take the time to analyze a whole bunch of sequences?

Unscrupulous would be playing it out with a stronger player.
Author:  brodie [ Thu May 05, 2011 6:24 am ]
Post subject:  Re: Rank Comparisons
True, of course, what I meant to say is folks that may be tempted to load the game into an sgf editor or onto a board and play out all the different scenarios. That, I feel, is dishonest. Staring at the board for 30 minutes and reading for 30 minutes is admirable, something I cant claim to possess the patience for, but certainly a reason I lose.
Author:  EdLee [ Thu May 05, 2011 6:28 am ]
Post subject: 
brodie wrote:
so a fast real time game is good for me.
You may enjoy Zuma Blitz or Bejeweled Blitz. :mrgreen:
Code:
www.facebook.com/zumablitz
www.facebook.com/bejeweledblitz
Author:  BaghwanB [ Thu May 05, 2011 9:25 am ]
Post subject:  Re: Rank Comparisons
That's been the inherent contradiction to me on DGS rankings vs. KGS (or other real-time servers).

You'd hope that you could play the best game of your life on a turn-based server. All the time you want. Most people know (and accept) that players can try out moves or use references to research lines of play. SGF out every path you'd like to your heart's content.

So what happens? I sometimes loose the path of play (even using the notes section) and the game turns into a whole-board reading exercise at some point or another. Or I just go "ooh, ooh" and put a move in like I'm playing blitz and can't even spend 30 seconds making sure it is right.

So I'd say on an individual basis, your DGS and KGS rank can have some wild variations or none at all depending on how you approach your games on BOTH servers. But on average, I'd agree with the -2ish variation most of the time (most...)

Bruce "You are not playing me in this corner; you are playing Cho" Young
Author:  brodie [ Thu May 05, 2011 10:04 am ]
Post subject:  Re: Rank Comparisons
Perhaps Im just the sucker. If I was sgf-ing (i love verbalization) I would most likely be pretty close to my kgs rank, which is perhaps telling me that I am in a minority by not doing so? Has this been discussed before on this board? I thought I maybe remembered one such discussion from godiscussions that frowned on the practice.
Author:  BaghwanB [ Thu May 05, 2011 10:18 am ]
Post subject:  Re: Rank Comparisons
brodie wrote:
Perhaps Im just the sucker. If I was sgf-ing (i love verbalization) I would most likely be pretty close to my kgs rank, which is perhaps telling me that I am in a minority by not doing so? Has this been discussed before on this board? I thought I maybe remembered one such discussion from godiscussions that frowned on the practice.



I think the "average" line for turn-based server play is that books and sfging is fine, but soliciting advice from stronger players isn't. So self-research is OK; asking others isn't. But your mileage may vary. Many people probably have their own lines they aren't willing to cross and would want their opponents to honor that as well. But on the other hand I've never seen anyone put any specs like this in their invites or posted game listing on DGS (I'm picturing something like "No SGF" or "No referencing"). Personally (and this is pretty much a public declaration here for any potential DGS opponents) I do use online databases and book references during DGS play but look at it as almost research in practice as opposed to my real-time, face-to-face skills. And I expect my opponents to do likewise.

Bruce "Verbing weirds language" Young
Author:  hyperpape [ Thu May 05, 2011 11:16 am ]
Post subject:  Re: Rank Comparisons
Huh-I've always tried to take my time on DGS, to the point of sleeping on it and looking again a day or two, but have never used an sgf reader or board. I also try to avoid books or sensei's pages that I think might be relevant.

I guess I was under the mistaken impression that avoiding books was the norm.
Author:  Laman [ Thu May 05, 2011 11:38 am ]
Post subject:  Re: Rank Comparisons
hyperpape wrote:
Huh-I've always tried to take my time on DGS, to the point of sleeping on it and looking again a day or two, ...


i sometimes take even up to week, when i am in a bad mood due to troubles on the board

and i use every possible help, just as BaghwanB wrote. at first i didn't want to play out sequences and look up josekis but how my rank grew i found out i just can't read deep enough and i don't know enough josekis and i can't know what aid my opponent uses, so i changed my mind.

and by the way, my game still sucks ;-)
Author:  Dusk Eagle [ Thu May 05, 2011 12:12 pm ]
Post subject:  Re: Rank Comparisons
Can't speak for DGS, as I've never played there, but I always use a real board to help me analyze variations when I've played Malkovich games. The issue of joseki dictionaries and pro databases is harder though, and so far I've always reached an agreement with my opponent about their usage. I think consulting other players about a current sequence crosses the line however.
Author:  xed_over [ Thu May 05, 2011 12:14 pm ]
Post subject:  Re: Rank Comparisons
brodie wrote:
If I was sgf-ing (i love verbalization) I would most likely be pretty close to my kgs rank,

No, it has very little to do with the "strength" of your opponents. The problem is the length of the time for games to finish and therefore update your rank.

If the average game finish time is 30 days, then compare your current DGS rank with your KGS rank from 30 days ago -- they'll likely be closer. (and this is not entirely accurate either, but should be close enough)
Author:  CnP [ Thu May 05, 2011 1:08 pm ]
Post subject:  Re: Rank Comparisons
As long as I thought I was playing a human (ie had the illusion of a proper opponent) I wouldn't care if they were part computer, part hamster.. If someone gets a 5kyu dgs account using a host of artificial aids, how do I lose out? I get to play a 5kyu, maybe we both improve (whether they're a 'real' 5kyu is basically unknowable and unimportant to me). Btw it almost seems rude to me not looking up the 'correct' joseki sequence!
Author:  amnal [ Thu May 05, 2011 1:29 pm ]
Post subject:  Re: Rank Comparisons
CnP wrote:
As long as I thought I was playing a human (ie had the illusion of a proper opponent) I wouldn't care if they were part computer, part hamster.. If someone gets a 5kyu dgs account using a host of artificial aids, how do I lose out? I get to play a 5kyu, maybe we both improve (whether they're a 'real' 5kyu is basically unknowable and unimportant to me). Btw it almost seems rude to me not looking up the 'correct' joseki sequence!


The bolded part can go horribly wrong ;)

But I agree with the rest of it.
Author:  brodie [ Thu May 05, 2011 5:15 pm ]
Post subject:  Re: Rank Comparisons
xed_over wrote:
brodie wrote:
If I was sgf-ing (i love verbalization) I would most likely be pretty close to my kgs rank,

No, it has very little to do with the "strength" of your opponents. The problem is the length of the time for games to finish and therefore update your rank.

If the average game finish time is 30 days, then compare your current DGS rank with your KGS rank from 30 days ago -- they'll likely be closer. (and this is not entirely accurate either, but should be close enough)


right, i understand the lag time, but theres more at work than just that. Say Im even with a 13k on KGS, where both of us are playing without aids. That same player on kgs using aids is going to win enough life and death scenarios to be a few kyu stronger. I know that there have been several big battles Ive lost in recent KGS games that wouldve turned out differently had I played them out with an empty board.

and dont get me wrong, im not throwing stones. as i said, im relatively new to all of this, and so if theres a norm im not aware of, thats my issue, but it certainly does affect rank. i think i'll start a poll, now im really curious...
Author:  daal [ Fri May 06, 2011 12:56 am ]
Post subject:  Re: Rank Comparisons
brodie wrote:
If I was sgf-ing (i love verbalization)...


I prefer verbing.
Page 1 of 2 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/