21st Nongshim Cup

[hyperpape]

I've run the probabilities based on the ordering above and the current go ratings:

Korea: 38.4%
Japan: 1.5%
China: 60.1%

Kyo Kagen: 3386
Ichiriki Ryo: 3499,
Yamashita Keigo: 3386
Murakawa Daisuke: 3298
Iyama Yuta: 3439

Mi Yuting: 3634
Xie Erhao: 3600
Fan Tingyu: 3555
Yang Dingxin: 3608
Ke Jie: 3675

Kim Jiseok: 3507
Won Seongjin: 3442
Lee Donghoon: 3532
Park Junghwan: 3670
Shin Jinseo: 3670

It's no surprise that Japan is far behind, but I wondered how much that had to do with the ordering--you'd ideally put Ichiriki last. Ordering all three teams weakest-strongest helped Japan by .2%.

Lastly, China has a 57-41 advantage, even if you put Korea last.

[Uberdude]

That ordering isn't the correct one, I'm not even sure the teams need to declare it in advance. The first game today is Murakawa Daisuke Vs Won Seongjin.

Update Won won.

[hyperpape]

Yes, this is just a placeholder, because the lists are revealed as time goes on.

It turns out that it does make a difference, albeit not necessarily a huge one (viewtopic.php?f=13&t=16027). I ended up not being sure what the best way to handle that. Perhaps I'll just order by strength, and rejigger things as I go. Or, I'll just put the calculator into a website and let the rest of you use it.

[trout]

1st game;
Won Seongjin defeated Murakawa Daisuke by resign.
2nd game;
Won Seongjin vs Yang Dingxin

Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon, Won Seongjin, Kim Jiseok
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta. Yamashita Keigo, Ichiriki Ryo, Kyo(Xu Jiayuan), Murakawa Daisuke
Bold case - Remaining players

[jlt]

@hyperpape (slightly off-topic hence hidden):

it is clear that when there are three teams, winning probabilities depend on the ordering.

Suppose there are three teams with remaining players

Team 1: player A
Team 2: player B
Team 3: players C and D, with D very weak (winning probability against any other player = 0%).

Suppose the next match is Team 1 against Team 3.

If C plays first, then Team 3 wins the tournament if C wins against A and B.

If D plays first, then Team 3 wins the tournament if C wins against the winner of the match of A against B.

We see that in this example, the best strategy is to make weaker players play first. I don't know if there are situations where the optimal strategy is different.

On the other hand, when there are only two teams left, it seems that the order is not important (assuming winning probabilities are computed using Elo's theory), although I don't know how to prove it.

P.S.

Update Won won.

How many wons did Won win?

[trout]

2nd game;
Yang Dingxin defeated Won Seongjin by resign.
3rd game;
Yarg Dingxin vs Yamashita Keigo

Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon, Kim Jiseok , Won Seongjin
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta. Yamashita Keigo, Ichiriki Ryo, Kyo(Xu Jiayuan), Murakawa Daisuke
Bold case - Remaining players

[Uberdude]

Yang Dingxin beat Yamashita Keigo by 3.5. The result wasn't surprising given their difference in rating, but Yamashita was actually decently ahead for a while according to bots but lost it (easier to make territory in edges).

[trout]

3rd game;
Yang Dingxin defeated Yamashita Keigo by 3.5.
4th game;
Yang Dingxin vs Kim Jiseok


Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon, Kim Jiseok , Won Seongjin
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta, Ichiriki Ryo, Kyo(Xu Jiayuan), Murakawa Daisuke. Yamashita Keigo
Bold case - Remaining players

[hyperpape]

jlt

If swapping two adjacent players never changes the probability, then any series of swaps doesn't change the probability, and any reordering of players can be achieved by some series of swaps of adjacent players (that's because permutations are an algebraic group generated by transpositions). So you'd just have to look at the ELO formula and see whether it satisfies the first claim.

[jlt]

hyperpape

I tried that already but it doesn't look easy.

[Uberdude]

Yang beat Kim both times they played recently (Samsung cup and chinese league).

Yang won again, killed a big group.

[trout]

4th game;
Yang Dingxin defeated Kim Jiseok by resign.


Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon , Won Seongjin, Kim Jiseok
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta, Ichiriki Ryo, Kyo(Xu Jiayuan), Murakawa Daisuke. Yamashita Keigo
Bold case - Remaining players

2nd round:
11/22 - 5th game
11/23 - 6th game
11/24 - 7th game
11/25 - 8th game
11/26 - 9th game

3rd round:
2/17 - 10th game
2/18 - 11th game
2/19 - 12th game
2/20 - 13th game
2/21 - 14th game

[trout]

5th game;
Yang Dingxin vs Ichiriki Ryo


Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon , Won Seongjin, Kim Jiseok
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta, Ichiriki Ryo, Kyo(Xu Jiayuan), Murakawa Daisuke. Yamashita Keigo
Bold case - Remaining players

[trout]

5th game;
Yang Dingxin defeated Ichiriki Ryo by resign.


Players:
Korea: Shin Jinseo, Park Junghwan, Lee Donghoon , Won Seongjin, Kim Jiseok
China: Ke Jie, Yang Dingxin, Fan Tingyu, Xie Erhao, Mi Yuting
Japan: Iyama Yuuta, Kyo(Xu Jiayuan), Murakawa Daisuke. Yamashita Keigo, Ichiriki Ryo
Bold case - Remaining players

6th game;
Yang Dingxin vs Lee Donghoon

[Uberdude]

Lee got totally outmanoeuvred by Yang, it almost felt like the sort of game where an amateur dan could get scolded by a teacher for playing a superficial imitation of AI style, while the opponent displayed a much deeper understanding. (I wrote this around move 100 because I'm that confident there won't be a reversal).