reform of Japanese Kisei tournament

[macelee]

Some notes on the new Kisei tournament format, I will write something formal for Go4Go later.

Since term 40, the Kisei tournament will have the following stages:

preliminary: open to all professionals and even some top amateurs. 16 players will be promoted to upper level,

C stage: 32 players, playing 5 rounds of Swiss, selecting top 6 to promote the upper level.

B leagues: 16 players divided in 2 8-player leagues B1 and B2, top two of each league are promoted to the upper level.

A league: 8 players, top 2 of league promoted to the upper level.

S league: 6 players.

playoff stage: the top 2 of S league (S1 and S2), the winner of A league (A), the winner of B1 vs B2 (B), the top player of C stage (C), are qualified for the playoff. They then play in the following order, C plays B, the winner plays A, the winner then plays S2, the winner then plays with S1 (S1 only needs one win to become Kisei challenger, his opponents needs to beat him twice to be challenger).

Titlematch: best-of-7 between the challenger and title holder.

For this year only, the various leagues will be populated based on players' past performance.

These major changes means:

- good players can play more game against strong opponents because of the existence of multiple leagues
- young talents can make more impact. In the past, assume there is a very strong player just turns professional, he needs to win all the preliminary games to make any impact. If for example, he loses one game at the final preliminary (which is a great achievement for a new player to get that far), he needs to start over and pass all the preliminary stages again next year. Under the new rule, he might be able to secure a position at a higher level league so his job next year is much easier.
- more competitive games. More stages mean more games, and less game fees given a fixed budget. So strong players get richer and weak poorer.

[EdLee]

- young talents can make more impact. In the past, assume there is a very strong player just turns professional, he needs to win all the preliminary games to make any impact. If for example, he loses one game at the final preliminary (which is a great achievement for a new player to get that far), he needs to start over and pass all the preliminary stages again next year. Under the new rule, he might be able to secure a position at a higher level league so his job next year is much easier.

About damn time.

Other than the Kisei, are all the other major Japanese tourneys (still) so blatantly protective of old(er) pros and so rigged against new(er) pros ?

Compare and contrast this aspect with the major pro tourneys in China and Korea.
Discuss among yourselves.

[oren]

Other than the Kisei, are all the other major Japanese tourneys (still) so blatantly protective of old(er) pros and so rigged against new(er) pros ?

No, Kisei, Meijin, and Honinbo have leagues where it takes effort to get into. I wouldn't say it's rigged against newer pros when 17 year olds have been making it in. 20 year old Ida won his way into the Honinbo league and then won it outright to challenge for the title.

The rest of the tournaments use single elimination preliminary tournaments from larger pools.

[EdLee]

when 17 year olds have been making it

Perhaps they made it despite the system.

[John Fairbairn]

About damn time.

Other than the Kisei, are all the other major Japanese tourneys (still) so blatantly protective of old(er) pros and so rigged against new(er) pros ?

Compare and contrast this aspect with the major pro tourneys in China and Korea.

Seems unnecessarily snide (rigged? - really? as well as ill informed.

The Japanese preliminaries system was reformed as long ago as 2003. The basis of it now is that if you lose in Round 1 you drop down a section the following year, irrespective of grade. Game fees were also revised so that players get a fee appropriate to how far they get in the tournament. In the past fees varied by grade. But the system was never entirely biased against young players anyway, if you recall the success of the likes of Rin, Ishida, Cho, Kato and others in their early twenties.

But the more important point is that each country's system reflects other country-specific cultural or social factors. In the case of Japan, the apparent bias against young players has been compensated for by longevity of one's career. In contrast, just recall how many young shooting stars there have been in Korea and Chinese (who also have seeding arrangements anyway) who have disappeared from the firmament already. You can even argue the oldies are discriminated against in China and Korea. In China you are put out to pasture as a coach. In Korea you get a few scraps from veteran tournaments. In Japan you can keep paying your mortgage.

Of course it may be that the Japanese system has worked against their success at international level, and my reading of the Kisei reforms is that they are more to do with that than with helping ALL young players.

[macelee]

But the system was never entirely biased against young players anyway, if you recall the success of the likes of Rin, Ishida, Cho, Kato and others in their early twenties.

Even after the reform of 2003, there was several transitional years when older/senior players got significant advantages. For example, have a look at http://www.nihonkiin.or.jp/match/kisei/028.htm showing the preliminary groups set up based on ranks during the transitional period. While 64 9-dans got 8 positions in the final preliminary, 72 1d-dan to 4-dan players got only 2. Wasn't that significantly biased?

[John Fairbairn]

Even after the reform of 2003, there was several transitional years when older/senior players got significant advantages. For example, have a look at http://www.nihonkiin.or.jp/match/kisei/028.htm showing the preliminary groups set up based on ranks during the transitional period. While 64 9-dans got 8 positions in the final preliminary, 72 1d-dan to 4-dan players got only 2. Wasn't that significantly biased?

That was only temporary and because the system was being phased in, in the sense that it all began at current settings. Phasing-in is perfectly normal in any organisation. It was especially sensible to do it this way here because the oldies at the same time (and this was not phased in) lost pension and game fee perks and had their promotion tracks slowed down considerably (effectively cut off in many cases), whereas for ALL the youngsters promotions were speeded up - though it has to be said promotions are no longer worth what they used to be, and I've considered dropping grades in sgf files for modern games.

[EdLee]

ill informed.

Yes, guilty as charged. It's only my gut feeling. Based on anecdotal evidence , (quoted in post 2).

if you recall the success of the likes of Rin, Ishida, Cho, Kato and others in their early twenties.

Which is more anecdotal evidence. Not rigorous analysis, one way or the other.

If I recall correctly, in HNG, they showed the pro qualifying exam process as every candidate playing every other candidate exactly once.
In other words, if there were N pro candidates that year, then they played exactly N(N - 1)/2 games.
The top X candidates with the highest scores would become the new pros that year -- I think X = 3 in HNG ?

I don't know if they still have the same process in Japan today, for the pro qualification.

But I'm pretty sure they don't do it that way in China.
In recent years, the number of candidates taking the pro qualifying tourney in China is something like 400 each year.
If they were to follow the same process in HNG, that would mean 400 x 399 / 2 = 79,800 games,
which is clearly not practical. So they have some kind of lucky draw to determine the pairings.
And the actual number of games played is much less than 79 thousand.
Again, the top Y candidates with the highest scores make pro for the year -- Y is something like 20, currently, in China ?

So, luck is involved -- not every candidate plays every other candidate.
Even if N(N-1)/2 happens, luck is still there: between any two particular candidates, A and B,
maybe in 100 games, A can beat B 75 times. But in the pro qualifying tourney, A and B only play each other once,
and B could get "lucky" -- the 25% chance -- and B beats A in this tourney.
Food poisoning. The common cold. Traffic congestion so you miss a tourney game (auto forfeit ?).
Hey, life is tough.

Do we know for sure the 20 new pros in China each year are the "best" among the ~400 candidates ?
We don't. But we recognize them as the new pros from the current system.

Luck is always an element. That's one thing.

But we also figure that in the long run, on average,
the people who are really "better" at this will eventually "move ahead". That's another.

For the same practical reasons, I don't think any pro tourney does the N(N-1)/2 -- do they ?
So they must come up with some kind of lucky draw system, seeding system, preliminary rounds, etc.
I'm guessing this also happens in pro chess tourneys, and in other pro sports, like tennis, etc.

To really demonstrate whether a particular tourney system favors or hinders certain people,
we need rigorous analysis.⁽¹⁾ And even if such analysis is published, I'm probably not qualified
to read and understand it in its original form -- I'll probably need someone else who can digest it
and explain it in a way more accessible to the general public.

But the more important point is that each country's system reflects other country-specific cultural or social factors. ...
In China.... In Korea.... In Japan you can keep paying your mortgage.

Indeed it would be informative to see some census data of the income distributions of the pros --
from tourney prizes, from teaching games, from publishing, etc., by age, by country (Japan, China, Korea), over the years.
Has anyone actually collected such data ? The Chinese Go magazine WeiQiTianDi used to publish some annual charts about certain pros in China,
including their tourney prize winnings (maybe they still do) -- a partial picture of their finances.

_________
⁽¹⁾ Example: statistical evaluation of basketball .

[ez4u]

- young talents can make more impact. In the past, assume there is a very strong player just turns professional, he needs to win all the preliminary games to make any impact. If for example, he loses one game at the final preliminary (which is a great achievement for a new player to get that far), he needs to start over and pass all the preliminary stages again next year. Under the new rule, he might be able to secure a position at a higher level league so his job next year is much easier.

About damn time.

Other than the Kisei, are all the other major Japanese tourneys (still) so blatantly protective of old(er) pros and so rigged against new(er) pros ?

Compare and contrast this aspect with the major pro tourneys in China and Korea.
Discuss among yourselves.

Actually the comment by Mace that a player who loses in the final preliminary has to start over at the bottom (which if I recall is taken from the announcement of the new structure) is an exaggeration anyway. ALL the preliminaries have a seed structure based on where people lost in the previous tournament. This can be easily seen in the preliminary tournament charts on the Nihon Kiin web site. A person who loses in the final preliminary will be seeded highest in the ABC preliminary and only has to play and win two games to re-enter the final preliminary the following year. This compares to the six games that someone starting at the bottom requires (in the case of the former Kisei structure).

[EdLee]

A person who loses in the final preliminary will be seeded highest in the ABC preliminary and only has to play and win two games to re-enter the final preliminary the following year. This compares to the six games that someone starting at the bottom requires (in the case of the former Kisei structure).

Thanks, Dave.

[ez4u]

Some notes on the new Kisei tournament format, I will write something formal for Go4Go later.
...
C stage: 32 players, playing 5 rounds of Swiss, selecting top 6 to promote the upper level.
...

This statement that the C tournament would be a Swiss also appeared on SL, but I do not see it in the announcement in Japanese. I expect that it will be something different. As far as I can tell, it will be a 5-round, triple-knockout tournament. This will result in one player with five straight wins, who will enter the playoff stage. In addition there will be five players with 4-1 records. They will go up to the B leagues next term together with the champ. The ten players with 3-2 records will remain in the C tournament. The sixteen players with 3 losses will drop back into the preliminary tournament. The players will drop out as soon as they record 3 losses. Should the C winner go all the way up to challenge the title holder, then one additional person will fall out of the S and A leagues to make room for the new star (or the deposed title holder if Mr./Ms. C goes all the way!).

This structure will require a total of 73 games per year.

[macelee]

ez4u thanks for your explanation. At the begging of my original post I say this is just some notes taken from various sources. The C stage arrangement is indeed not very clear, and it is a bit strange that they call it a 'league'. We only need to be patient - when the games are actually played we can surely have more information.

[oren]

Shukan Go called C league a Swiss-like system.

[ez4u]

Shukan Go called C league a Swiss-like system.

They may consider it 'Swiss-like' in that players with the same record will be paired in each round. This is straight forward in a 32-player 5-round event since there are always sufficient players with equal records in each round.

[macelee]

Wrote a formal article to record information available here

http://www.go4go.net/go/tournaments/kisei