It is currently Sat Aug 19, 2017 6:11 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 95 posts ]  Go to page Previous  1, 2, 3, 4, 5
Author Message
Offline
 Post subject: Re: 18th Nongshim cup
Post #81 Posted: Mon Nov 28, 2016 4:38 am 
Lives with ko

Posts: 147
Liked others: 12
Was liked: 167
Rank: UK 2d Dec15
KGS: mathmo 4d
IGS: mathmo 4d
Chang Hao was saying how different this is to the disaster of 1993 in the related 5 player 3 country, team win and continue tournament, where China had no wins, Nie was knocked out by Yoda. From then, this was so disappointing in contrast to the victories in the China-Japan supergo, with progress of the youngsters too slow, that this shock led to a big push in go.

even so, Korea has dominated for nearly 30 years until about 2013

Tang Weixing was expressing worry before the LG semi-final that in each game the Korean seems to have a stronger chance on paper, but with China dominating 6-2 in the quarter final, having an all Korean final would be hard to accept: normal 3 years ago, but not these days. Then he says that maybe it would be a good thing though, let the Koreans win some. ... but Park and Shin were both knocked out.

Uberdude: how do you calculate winning probability for go elo ratings?

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #82 Posted: Mon Nov 28, 2016 4:52 am 
Lives in gote

Posts: 498
Liked others: 30
Was liked: 163
GD Posts: 10
If Japan win, nobody will question about Iyama Yuta's world ranking for at least ten years.

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #83 Posted: Mon Nov 28, 2016 5:03 am 
Tengen

Posts: 4068
Location: Cambridge, UK
Liked others: 145
Was liked: 2000
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
dhu163 wrote:
Uberdude: how do you calculate winning probability for go elo ratings?

For Remi's WHR ratings on goratings.org, it's the standard Elo formula e.g. http://www.bobnewell.net/nucleus/bnewell.php?itemid=279. I don't know if the Chinese pro ratings or the Korean Park Taeil or mamumamu's follow that too.

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #84 Posted: Mon Nov 28, 2016 12:46 pm 
Lives with ko

Posts: 147
Liked others: 12
Was liked: 167
Rank: UK 2d Dec15
KGS: mathmo 4d
IGS: mathmo 4d
Uberdude wrote:
dhu163 wrote:
Uberdude: how do you calculate winning probability for go elo ratings?

For Remi's WHR ratings on goratings.org, it's the standard Elo formula e.g. http://www.bobnewell.net/nucleus/bnewell.php?itemid=279. I don't know if the Chinese pro ratings or the Korean Park Taeil or mamumamu's follow that too.


but the chess and go ratings have different numbers for their parameters.
http://www.europeangodatabase.eu/EGD/EG ... system.php

I couldn't find a reliable source for the chess a factor, but Wikipedia suggests 400. But EGD uses around 70 at a high level. So a 400 point rating difference in chess has equivalent win probability as 70 in go (pretend no draws are possible). So perhaps you probabilities should be much more extreme. e.g. rating gap of 70 to imply 25:75 chances

I couldn't find a explicit reference to any value of a factor in the Remi Coulom paper, unless the rescaled "natural rating" is relevant.

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #85 Posted: Mon Nov 28, 2016 2:01 pm 
Gosei
User avatar

Posts: 1952
Location: Tokyo, Japan
Liked others: 1690
Was liked: 1073
Rank: Jp 6 dan
KGS: ez4u
Uberdude wrote:
dhu163 wrote:
Uberdude: how do you calculate winning probability for go elo ratings?

For Remi's WHR ratings on goratings.org, it's the standard Elo formula e.g. http://www.bobnewell.net/nucleus/bnewell.php?itemid=279. I don't know if the Chinese pro ratings or the Korean Park Taeil or mamumamu's follow that too.

The Chinese rating system is described here http://baike.baidu.com/view/8060410.htm(in Chinese) with standard Elo probabilities shown in the table included. Mamumamu uses glicko-2 instead of elo so has a completely different scale.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #86 Posted: Mon Nov 28, 2016 2:33 pm 
Tengen

Posts: 4002
Location: North Carolina
Liked others: 399
Was liked: 658
Rank: AGA 3k
GD Posts: 65
OGS: Hyperpape 4k
dhu163 wrote:
I couldn't find a reliable source for the chess a factor, but Wikipedia suggests 400. But EGD uses around 70 at a high level. So a 400 point rating difference in chess has equivalent win probability as 70 in go (pretend no draws are possible). So perhaps you probabilities should be much more extreme. e.g. rating gap of 70 to imply 25:75 chances
400 points in chess is 9-1 odds according to Newell, which would imply that Ilja has 9-1 chances of beating Artem (if 70 points EGD = 400 points in Chess). Am I misreading? Because that seems high.

_________________
Occupy Babel!

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #87 Posted: Mon Nov 28, 2016 3:48 pm 
Gosei
User avatar

Posts: 1898
Location: Groningen, NL
Liked others: 189
Was liked: 1015
Rank: Dutch 4D
GD Posts: 645
Universal go server handle: herminator
dhu163 wrote:
but the chess and go ratings have different numbers for their parameters.
http://www.europeangodatabase.eu/EGD/EG ... system.php

I couldn't find a reliable source for the chess a factor, but Wikipedia suggests 400. But EGD uses around 70 at a high level. So a 400 point rating difference in chess has equivalent win probability as 70 in go (pretend no draws are possible). So perhaps you probabilities should be much more extreme. e.g. rating gap of 70 to imply 25:75 chances

I couldn't find a explicit reference to any value of a factor in the Remi Coulom paper, unless the rescaled "natural rating" is relevant.


The GoR system (EGD) is not universal to go, it is just one possible implementation of the system. Original Elo uses 1 / 1 + 10(D/400) whereas GoR in its higher ranks uses 1 / 1 + e(D/70) (but uses higher numbers than 70 at lower ranks). The fact that Elo uses 10 as a base while Gor uses e means that in you cannot compare the 400 and 70 factors directly. In practice, the difference is smaller, and using e as a base, chess rating effectively use 173.3 (i.e. 10(D/400) = e(D/173.3). Remi's implementation, AFAIK, uses the original simpler Elo rating without the extra variable that GoR introduces.


This post by HermanHiddema was liked by: dhu163
Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #88 Posted: Mon Nov 28, 2016 5:10 pm 
Lives with ko

Posts: 147
Liked others: 12
Was liked: 167
Rank: UK 2d Dec15
KGS: mathmo 4d
IGS: mathmo 4d
ah I see thanks, that makes sense, I'm a mathematician, but didn't notice the 10

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #89 Posted: Tue Nov 29, 2016 12:44 am 
Lives in sente

Posts: 1106
Liked others: 0
Was liked: 432
9th game;

Park Junghwan defeated Fan Tingyu by resign.

Korea: Park Junghwan, Kim Jiseok, Kang Dongyoon(seeded), Lee Sedol, Lee Donghoon
China: Ke Jie, Tuo Jiaxi, Lian Xiao, Fan Yunrou, Fan Tingyu
Japan: Iyama Yuuta, Murakawa Daisuke, Kono Rin, Ichiriki Ryo, Cho U

Bold case - alive

2nd round ended.

See you in February!!!!!

3rd round,
2/21 - 10th game
2/22 - 11th game
2/23 - 12th game
2/24 - 13th game
2/25 - 14th game


This post by trout was liked by: emeraldemon
Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #90 Posted: Tue Nov 29, 2016 1:20 am 
Lives in gote

Posts: 498
Liked others: 30
Was liked: 163
GD Posts: 10
Is the next match Park Junghwan vs. Iyama Yuta?

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #91 Posted: Tue Nov 29, 2016 1:25 am 
Dies with sente

Posts: 80
Liked others: 23
Was liked: 4
Rank: KGS 3d
Yes, for sure.

Now the #1 from both countries have equal chance to win the tournament and make history. :lol:


This post by kimidori was liked by: ez4u
Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #92 Posted: Sun Feb 19, 2017 4:39 pm 
Lives in gote

Posts: 590
Location: Austin, Texas, USA
Liked others: 41
Was liked: 192
trout wrote:
9th game;

Park Junghwan defeated Fan Tingyu by resign.

Korea: Park Junghwan, Kim Jiseok, Kang Dongyoon(seeded), Lee Sedol, Lee Donghoon
China: Ke Jie, Tuo Jiaxi, Lian Xiao, Fan Yunrou, Fan Tingyu
Japan: Iyama Yuuta, Murakawa Daisuke, Kono Rin, Ichiriki Ryo, Cho U

Bold case - alive

2nd round ended.

See you in February!!!!!

3rd round,
2/21 - 10th game
2/22 - 11th game
2/23 - 12th game
2/24 - 13th game
2/25 - 14th game


Reminder!
Park Junghwan vs. Iyama Yuta 2/21
https://www.wolframalpha.com/input/?i=Feb+21+15:00+kst

The winner will try to beat all 4 remaining Chinese players: Ke Jie, Tuo Jiaxi, Lian Xiao, Fan Yunrou


This post by yoyoma was liked by: swannod
Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #93 Posted: Sun Feb 19, 2017 8:23 pm 
Tengen

Posts: 4002
Location: North Carolina
Liked others: 399
Was liked: 658
Rank: AGA 3k
GD Posts: 65
OGS: Hyperpape 4k
I've created a git repo with code to calculate winning chances in win and continue tournaments (https://github.com/hyperpape/win-and-continue/). Prior to the start of this year's Nongshim, Korea was the slight favorite. But after Fan Tingyu's 7 game winning streak, the odds are now
Japan: 2.5%
Korea: 6%
China: 91.5%

One question I'm curious about: are there any four team win and continue tournaments? I can't find any linked from Sensei's or Mr Kin's site.

(For anyone who follows the group on Facebook, sorry for the duplicate post).

_________________
Occupy Babel!

Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #94 Posted: Tue Feb 21, 2017 11:20 am 
Dies in gote

Posts: 45
Liked others: 3
Was liked: 7
Iyama lost to Park by resignation. Japan is now eliminated.


This post by ewan1971 was liked by: emeraldemon
Top
 Profile  
 
Offline
 Post subject: Re: 18th Nongshim cup
Post #95 Posted: Wed Feb 22, 2017 2:56 am 
Tengen

Posts: 4068
Location: Cambridge, UK
Liked others: 145
Was liked: 2000
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
Today Fan Yunruo 5d beat Park Junghwan by 1.5 points, winning the Nongshim cup for China with 3 players spare. China has such a depth of world-class players: when the two Fan's were selected for the Chinese Nongshim Cup team apparently there negatives mutterings in Chinese forums about if they were good enough (leaving behind higher rated players like Zhou Ruiyang, Shi Yue, Mi Yuting, Chen Yaoye) but they won the whole tournament without Tuo Jiaxi, Ke Jie or Lian Xiao needing to turn up!


This post by Uberdude was liked by: emeraldemon
Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 95 posts ]  Go to page Previous  1, 2, 3, 4, 5

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group