Bugcat/World Championship Format Idea

Sub-page of Bugcat

This is my idea for the format of a rigorous and modern World Championship for Go.

The tournament is held every two years. In the first year is the league tournament, and in the second year is the title match.

The league tournament is a ten-player single round robin. The players are selected as so:

  • The 4 best-placing players from last league tournament (including the beaten champion if necessary)
  • 1 Chinese player
  • 1 Japanese player
  • 1 South Korean player
  • 1 Taiwanese player
  • 1 American player
  • 1 European player

The title match is held between the challenger (the winner of the league) and the defending champion, and is a best-of-ten.

Uberdude: What properties of this format make this rigorous or modern, or indeed more worthy of the title of World Championship over the several existing similar events?


Why it is rigorous:

  • 1. The league format is a round robin.

In a knockout format (typical of a lot of professional Go tournaments), a player can be knocked out in one game. So for instance, in a knockout system, a favourite to win the league (such as, for instance, Lee Sedol) could lose a game in an upset against a weaker player and be knocked out of the tournament. In a round robin format, on the other hand, he would have another eight games in which to prove his playing strength. Even in a double elimination tournament, a player only needs to lose two games to be knocked out.

Furthermore, _every_ player has to play nine games. This gives us a much better indication of their relative playing strengths than, say, a four-round knockout (where the winner would only have to win four games). The highest-level use of the round robin format is in the World Chess Championship (where they actually play a double round robin - two games with each player).

Another benefit is that the strongest players are guaranteed to play against one another in the tournament. This is a real boon for advertising, but it's also really good for us fans.

  • 2. The best-placing players of the league stay on.

This means that players who proved their strength the previous year can't be knocked out by an upset in preliminary qualifying. As they don't have to play prelimaries, they can also spend more of their time preparing for their league games.

  • 3. There is a relatively long title match.

The usual final match for a big title Is a best of five. In my format, we double this amount. This forces the winner of the match to prove their superior playing strength over a larger amount of games. A longer match also allows the sponsors to get more publicity and so increase their profit (and put some of that back into the tournament funds).

Why it is modern:

  • 1. The players are international, and include the West.

Being sure that there will be at least one player from the six major federations of the world listed means firstly that most players in the world can theoretically qualify. Secondly, it means that almost all Go fans have a player to root for. This means more fun for us funs, more publicity for the sponsors, and probably more funds for the tournament. Looking at the next fifty years, the West is probably going to be the major area of growth of the game, so it's a modern, forward-looking idea to include them (as the AGA and the EGF, the two major Western federations).

  • 2. The round robin league system means that every _gets_ to play everyone else

Low-ranking pros from the West (or even high-ranking amateurs) can qualify and get to play a serious game with every other player in the tournament. So for instance, Ilya Shiksin might qualify and get to play Lee Sedol, Lee Changho, Gu Li etc. Then he can come back to Europe with that experience, improve his play, play other Europeans and raise everybody's game in the upper tier of European Go a little bit. You don't usually get a game with top ten players if you get knocked out in the first or second round of a round robin. The same goes for young professionals in the Orient, trying to get some experience to improve their game.

In summary:

  • More rigorous
  • More fun for fans
  • More publicity for sponsors
  • More experience for improving players

Uberdude: Rigorous-1: I agree that a round-robin is better at finding the best player from the 8 who made the league than an elimination, but the big downside is it is a much smaller pool of players vying for the title than existing international tournaments (typically 32 or 64 for main rounds, and much more in prelims). So there is more chance that some strong player who 'should' be the true winner isn't even in the event. How do you choose the 1 Chinese/Korean representative who wasn't in the 4 best last year? A knockout tournament in that country is just pushing the deficiencies of knockout from this event out to the selection tournament. Also I believe most tournaments pay game fees, so lots of games in a league means more cost.

Rigorous-2: True, but also means some old dinosaur who got weaker sticks around and takes a spot from a young improver. You won't get loads weaker in 2 years, but from top 4 in world to not in top 10 and barely top 50 is very plausible. Take a look at Xie He for example from around 2011, he dropped pretty quickly.

Rigorous-3: Nice indeed, but also more cost and tiring for players. Something you didn't mention for why this event could be _the_ World Championship was prize money: I think you need it to be significantly more than existing ones so about $ 1 million. (Ing is $400k, Mlily/Samsung about $270k, Mlily Lee vs Gu jubango was $750k. AlphaGo vs Lee was $1m). You need a rich/generous sponsor.

Modern-1: Including Western players is nice for the audiences there, but what are comparative audience sizes and interest to sponsors? By denying 2 strong Chinese/Korean players to give 2 seats to Westerners who are certain to not win you reduce the ability of the tournament to find the best player in the world due to the small entry.

Modern-2: Whilst it's fun to see Ilya getting smashed by a top pro like Lee Sedol or Ke Jie, he can learn plenty from playing some no-name Chinese 1p who is world #534. I think the EGF team playing in the Chinese C league (as they did this year) is a far more efficient way for the Western pros to get valuable pro game experience without wasting 2 of 10 spots in a world championship with top 10 pros.

Anonymous: Definitely a round robin is better than any knock-out method. Remember back in the 20th century when Fujisawa Hideyuki (aka Shuko) held the Kisei title for six straight years though his play elsewhere was not on the same level. He also made a well-known comment to the effect that he only had to win four games a year to stay at the top. I would expect a real world champion to win a long match from a challenger from a round robin selection tournament and also do well in other top level tournaments.

I've always disliked the calling an international tournament "a" world championship. Determining the title by ratings is also not good.

bugcat: I've read Uberdude's comments and I've revised my idea to make it less stodgy, tweaking the format of the league to encourage the qualification of strong new players. The round robin format and best-of-ten match stays the same, though.

The tweaked league make-up (in order of qualification precedence):

  • Runner-up of last tournament.
  • 2 Chinese players (the finalists of the Tianyuan)
  • 2 Japanese players (the finalists of the Honinbo)
  • 2 South Korean players (the finalists of the Kuksu)
  • One Western player (either US Cup winner or European Championship winner (rotates each tournament)
  • One Taiwanese player (winner of the Taiwanese Tianyuan)
  • Winner of the Samsung Cup. If already qualified, then the other finalist. If other finalist already qualified, then the organisers' choice.
  • Highest ranking player on Goratings not included

Here, we have 11 players (and thus ten rounds). From these, we can currently expect there to be about eight players with a good chance of winning (the 2 Chinese, the 2 Koreans, 1 Japanese (Iyama), the Samsung finalist / organisers' choice, the Gorankings player, and the runner-up from last cycle. Then we have the other three players (the other Japanese, the Western player, and the Taiwanese player) to keep the tournament international. Taiwan get 1 space as opposed to the AGA / EGF's half-space each because their pros are currently stronger.

See that there is now only one player remaining in the league from tournament to tournament, and only two players (including the champion) remaining in the entire cycle. This is more in line with the structure of the World Chess Championship. See also that the qualification to the league is more defined, though I chose those big titles fairly arbitrarily based on seniority.

Now all we need is pros to play in it ;)

Herman: I think "only two players (including the champion) remaining in the entire cycle" is still two to many. Especially the champion qualifying directly for the final is wildly unfair, IMO. Qualifying the reigning champion for a final league is as far as I would go.

bugcat I think we'll have to agree to disagree there. From where I stand, a player shouldn't call himself the World Champion unless he can prove that he's stronger than the current Champion. The league tournament is really just about selecting the challenger.

Herman: IMO, if the reigning champion participates in the league, then whoever else wins the league has proven themselves stronger than that champion. Challenge matches are fun and are good PR, but they are objectively terrible at finding the best player in the world.

Uberdude FWIW, in the British championship we used to have a round-robin league (selected as top 8 from a swiss prelim with minimum rank requirement) to select the challenger for the title match against the incumbent champion. This was changed to top 2 from the league for the title match, with current champion qualifying direct to the league. This was more because although he (Matthew Macfadyen 6d) was going to win anyway, by not playing in the league other people didn't get many chances to play him. I think I have the dubious honour of being the first incumbent champion to fail to finish in the top 2 of the league ;-) (we have a more even field now Matthew has retired).

Slarty: Glancing at this, I don't understand why the assumption that the world is made of countries carries so much weight. Some Chinese provinces, for example, have more talent than the United States, while there is some Go talent in other regions.

Uberdude Also choosing the country players as finalists of Tianyuan/Honinbo/Kuksu seems somewhat arbitrary, ditto Samsung, why not LG? (and could be not distinct from other categories), plus the Kuksu no longer exists.

bugcat: Fair points from Slarty and Uberdude. As a side note, I didn't realise that the Kuksu was dead because the Korea information on the BigTitles page is three years out of date. Someone updating that would be much appreciated :) I guess we could use the Myeongin instead, but even that page hasn't been updated since 2015 so idk whether or not it's alive.

In response to both of your comments, at the end of the day there has to be a qualification procedure or no players will be there to play. I think using tournaments to qualify is better in general than using mathematical rating from Goratings or just picking players that "look strong" or are "big names" with no formal criteria. The tournament choices are going to be semi-arbitrary because of the nature of the professional Go scene: China, Korea and Japan each have multiple "big titles" but no single "national championship". Since relative prize purses can increase and decrease, that isn't a sound basis for choosing a qualification tournament in the long term, so seniority is the next best criteria (as it indicates stability). By seniority, the LG Cup would actually be the proper choice (it's a few months older), so you're right there. As to why I think it would be reasonable to include it as a qualification path (as with the one Goratings qualifier), is that it's another type of qualifier and it lets a player in who is, for whatever reason, doing well internationally but not nationally.

Uberdude Myeongin is dead too I think, Korean pro Go is in dire straits. I certainly agree using a qualifying tournament is better than picking "big names", my implication was comparing using existing tournaments versus organising your own large and open preliminary (as is normal, e.g. for Samsung). Using an existing event saves time/money, but means you get their disadvantages, e.g. imagine you picked Meijin instead (I know Kisei is richest but newest of big 3 Honinbo/Meijin/Kisei, but don't know which is highest prestige) and Iyama didn't manage to be challenger vs Takao, then you'd have Takao plus not Iyama as the Japanese players whilst Iyama held the other 6 big titles, which most would agree was a selection failure. Japanese titles also have quite a lengthy tier of preliminary leagues, so if a new young superstar appeared (e.g. Shibano) it can take them several years to advance to the final so would miss out on this event. I could well imagine a lot of moaning and "what ifs" if the qualifier by your chosen tournament happened to be someone generally regarded as a weaker player who was lucky to win that arbitrarily chosen title and the stronger players of that country got left out (Lian Xiao is no weaky, but Ke Jie didn't even bother to play in the last Tianyuan for example, though perhaps he would if this event existed with a $1m prize and it was publicised the Tianyuan was the qualifier).

DudleyMoore: It isn't such a bad suggestion as formats go. The only problem is that there is no desire from the professional world to band together into one association, or to create one (mal)governing body, such as the game of chess has in FIDE.

Bugcat/World Championship Format Idea last edited by DudleyMoore on May 17, 2019 - 14:55
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Login / Prefs
Sensei's Library