I find that completely ridiculous. Are you seriously telling me that the EGF committee and Rules Commission are so inept, that they could not (for example) agree to let somebody as trustworthy as their own secretary do something as simple as to correct the English in the document?

PS: As you will see, rewriting the document in plain English actually reduced its length.

1) The language has already been corrected by, among others, Toni Atkins.

2) That you don't like the result of the language correction does not mean that by all means another correction fitting your language feeling would be necessary.

3) 90% of ordinary rules' text discussion tends to be wasted with different preferences for different language styles. You may like that but it does not mean that everybody has to like it.

4) As said before, I did not find a majority for letting the public proofread the language before the text was adopted.

5) You hope language could be changed quickly. By experience, I know that the EGF bodies do not work as fast as we might wish. It is realistic to assume that the EGF Committee does not want to bothered with endless language corrections.

6) Nevertheless, as already indicated, you (or everybody) can send in proposals and we will store them for probably much later usage. If the proposals are excellent, maybe even soon usage. But I cannot promise anything like this because the EGF does not work like a Wiki.

The following was sent to the EGF by the British Go Association today:

The English is pretty difficult to understand, so we think that a rewrite by a native English speaker is advisable.

The document responds to much of our proposal that was approved at the AGM in 2010, but this is the first time we have had an opportunity to discuss the Rules Commission proposals and we believe that the following changes are necessary for the 2011 event:

1) drop the Relegation system

We strongly disapprove of this as a) it is possible that a large number of people may be involved with this, b) it takes up one of the allegedly free days of the tournament and c) we think that using tie-breaks is going to be good enough for the minor places in this play-off system.

2) drop the rule that the European Open Champion must be a non-European

We think very strongly that the Open Champion should be the person who's got the most wins, allowing for tie-breaks, in the Main Event. Games in the Closed play-off section should count towards this event also. It's possible (likely?) that this may be the same person as the Closed Champion, but not necessarily.

A side point to this argument is that it is possible that with the current proposal one of the European players not involved in the play-offs might score more wins than any non-European, but would be prevented from being Champion! Bizarre.

It should be pointed out that neither of these parts of the system have, as far as we're aware, been discussed outside the Rules Commission, which is obviously regrettable as they're clearly controversial.

3) Eliminate the play-off for 3rd place

This contradicts the basic simplicity of the KO system as proposed which says that losers rejoin the main tournament, excludes from the possibility of any final game against an Asian and risks a repeat game. It also contributes nothing to the overall tournament.

4) We have some technical issues with the section on Final Results about forced pairings. Geoff Kaniuk will communicate directly with the Rules Commission about these.

Jon Diamond
President

This is quite an evasive reply Robert I would dearly love to see the document at the point where Tony Atkins corrected it, I wonder if that could be provided (perhaps privately)? For points 2&3 - I hope you would agree that explaining the presence of phrases in a document such as "5+4=9" by saying that 5 is smaller than usual is not acceptable in a document of this stature.

As evasive as you not contacting the Rules Commission as a commission yet and asking us whether your complete language rewriting would then be discussed in the Rules Commission and then in the EGF Committee for possible adoption.

Recall that we are not discussing only my personal opinion but you want to change (rather than only comment on) an official EGF rules text. This means a) the text must be opened again for change by the EGF Committee or AGM and b) there must be a majority in the Rules Commission for accepting external input and for this, since I am open to such input, Matti Siivola needs to be convinced. Last time I tried to convince him to allow external input before submitting to the EGF Committee for adoption, I failed to convince him.

Toni Atkins corrected almost the same as the final document; only a few changes he considered necessary.

As a member of the Rules Commission, I have not received it yet but it is still early in the morning:)

What follows is my personal opinion:



If the EGF Committee doen not mind extra correction cycles, such is possible but might waste much time for little benefit. Rewriting a rules text is never just a language issue but the contents must remain unchanged when language improvement is the only concern. Therefore any language correction must then be checked again as to accidental contents changes.



Every year starting from 2011 is important, not just 2011.



(a)

The number of people involved in relegation games is between 0 and 16. The number of people otherwise involved in seeding by tiebreakers in about the same range. The expected average number is somewhere in the middle, around 8 players. Calling this large as an absolute number or in relation to tiebreaking is an exaggeration.

Apart from the size itself, the question is whether a slightly smaller or slightly larger percentage of relegation players as a fraction of players with the smallest eligible MMS is better. If one considers relegation games to be not meaningful, then one would conclude that a smaller fraction is better. If one considers relegation games to be meaningful, then one would conclude that a greater fraction is better. Therefore the argument is not an argument itself but merely repeats the question whether relegation games are good at all.

(b)

For the 0 to 16 relegation players, the otherwise free day is not free indeed. The question is: Is having relegation games worth this? My opinion: Absolutely yes! (Because of the reasons that make relegation and seeding by greater number of wins a much better seeding than by tiebreaking.)

(c)

Why? The opinion lacks reasoning.

"Minor places" is a rhetorical trick. It might affect all seeding places 1 to 8! We are not talking about necessarily only the last places 8 or 7.

These are reasons why relegation is much better than tiebreaking:

- Qualifying (among high MMS players) by greater number of Wins is a much more meaningful seeding criterion than SOS - rating - lottery tiebreaking because Wins is caused by each player's own responsible achievement while SOS is caused by opponents' achievements, which the player cannot significantly influence.
- Unsportsmanlike SOS manipulation (like reported from Grenoble EGC 1987) becomes essentially impossible.
- If there is a high percentage of strong non-Europeans, then after 7 rounds McMahon the lower end seeding candidates might have so few Wins (like only 4) that qualification for some might be too easy; e.g., 4 wins in 7 rounds is not a convincing seeding. By having to play a relegation game, the most doubtful candidates will have to show that they can make 1 more win before they will actually qualify.
- Similarly, too few wins might be too few especially when compared with the top European(s) after 7 rounds, who might have 6 or 7 wins. E.g., if a 4 or 5 wins player has to play a relegation game to then have 5 or 6 wins, then the difference of wins among the qualified players will be much more acceptable and a Eu Champion with fewer wins than the finally 2nd European will be much less likely.

The BGA reasoning does not even discuss these important reasons at all.



There is much freedom of how to define the European Open Champion. I do not not have a strong preference here, except for the tiebreaking aspect. However,...



...this means that SOS tiebreaking for determining the EOpenC would become more doubtful than ever because players pools are separated by the McMahon - KO split. E.g., if there are many very strong non-Europeans, then the strongest Europeans will get less chance to collect high scoring opponents while they are separated in the KO and may not get a chance to play stronger non-Europeans than strongest Europeans.



Emotions are not good reasons. Can this be expressed by reasoning, please?



Agree.



I do not have a strong preference. There are advantages and disadvantages for either solution.



As always, the Rules Commission is happy to receive technical tournament system suggestions.

But if there is only one place left to be decided and there are, for example, 9 players who are tied such that they would play in the Relegation, then I'd call that very large, not just large. It would also require more than one game to decide the place - how many would be required Robert?

'better' is an opinion, 'would' should be 'could' with a different interpretation and your logic is faulty too as the therefore doesn't follow.

Our opinion differs.

Why do we need to provide reasoning?

In general we believe that people who argue about the minute details about tie-breaking are wasting their time. If you're hard done by in losing out in a tie-breaker the remedy was always in your own hands - just win more games to avoid the tie-breaker.

If everybody's got the same number of wins then we shouldn't expect seeding to be very accurate, so shouldn't spend too much time and effort over it.

Sure, it's better in this case, but not an important enough reason in our view.

Is once in 25 years (at least) a good reason, when the Appeals Committee could (in principle) have authority to rule in these exceptional cases?

Is this an argument?

Actually, we don't care about this point. If the Champion has fewer wins than another then that's just life. The Champion is the winner of the Knock-Out - end of story.

We don't think these are important.



I don't think we need to - it's an opinion and summary of how strong our view of the argument immediately preceding is. If you want to take this as emotion then please in your responses only use objective terms and logic. For example, terms you've used such as 'better' and 'should' should be avoided, except when they refer to facts.

And in the end this kind of decision is not a totally objective one as there are, as I've expressed here, some arguments that you've provided that we can accept. They just don't balance the reasons we have for an opposite view - a subjective judgement but equally valid I believe.

Apparently you have not understood the relegation system yet.

If only one place is left to be decided, it means that we have 7 players who qualify automatically (+) like in the following example, where "x" means "not in top 16 and "-" means "not qualified automatically and "R" means "enters relegation":



Exactly one relegation game will be played.

Now you call this a "very large" number, LOL.

Here are some more examples from the years 2000 and 2002-2006, which I have already studied:


Abbreviations:

x y := #wins (supergroup players; non-supergroup players with more
wins included) | #Europeans playing all rounds thus far

n := smaller MMS / greater D
W := number of wins (top Europeans only)
D := MMS difference to top MMS Europeans
Q := qualified
+ := qualified automatically
- := not qualified automatically
R := relegation
#R := number of players in relegation
#S := number of players in SOS - rating - lottery tiebreaking








1) To convince more others than you might convince by not providing reasons.
2) To characterize yourself as open to objectivity.

Not providing reasoning is as weak as trying to convince others of "Earth is flat." by "Look around and see, it is flat!".



Then why do you support a system with heavy tiebreaker seeding?



If you believed in what you saying, then you would appreciate the "just win more games" by using this as the only essential seeding criterion.



If everybody's got the same number of wins, then better seeding is gotten from playing relegation games so that players with the ability to win more are distinguished from players without that ability.



It is very unfortunate that you value a free day for the few decisive players higher than their playing of a decisive game, which furthermore by its nature fits very well into the KO nature: Win and you stay in - lose and you are out.



Which 25 years do you count? Did they all have a KO? How can you be sure that in other years it did not happen? Just because nobody noticed? It is extraordinarily easy to lose won games intentionally but usually extraordinarily hard to prove this. Also unintentional losses can be decisive for SOS and deciding the winners / qualified players. Stefan Kaitschik, strong 4d can tell you how his last round loss against a 2d determined the 2007 winner.



It does have this power but it is useless in almost all practical cases.



If you cannot understand why 4 wins in 7 rounds is too weak a criterion, I suggest that you take your time for studying fundamentals of what makes a good versus bad tournament system.



You don't care about too many things. With this your argument, you might as well omit the McMahon and replace it by a pure lottery. IOW, what happens until the KO is very important and not just fun of doing careless, rough approximation.



You show that you don't want to design good systems. Example: You want KO because you believe in greater number of wins being the most meaningful. At the same time, you don't want relegation instead of a free day for also the few relegation players because... you believe in greater number of wins not being important. Maybe such contradictions are good enough to convince the BGA itself but they are not attractive for convincing enough others.

Here are more complete statistics:

Previous EGCs statistics for combined McMahon 7 rounds - KO 3 rounds system:


Abbreviations:

x y := #wins (supergroup players; non-supergroup players with more
wins included) | #Europeans playing all rounds thus far

EDIT: In the years 2007-2010, only European supergroup players were counted.

n := smaller MMS / greater D
W := number of wins (top Europeans only)
D := MMS difference to top MMS Europeans
Q := qualified
+ := qualified automatically
- := not qualified automatically
R := relegation
#R := number of players in relegation
#S := number of players in SOS - rating - lottery tiebreaking








Averages

#R = 5.6 // This is the average number of relegation players.
#S = 10.6

Players with X wins after 7 rounds qualified (all 10 years)

Considering that the average number of relegation players is only 5.6, the BGA claim of a (very) large number of players losing their free day is totally wrong.

I'm not familiar with the prior history of the debate on either side, but on the evidence presented here this "totally wrong" is totally wrong. The BGA has listed three separate reasons. Each reason stands on its own. You have conflated two and here ignored the other. The BGA quote is below.



I'm guessing, but on past experience I'd expect (b) to be the most important reason for many people. The EGF rules committee appears to have taken the stance that the system is the most important thing and the players are there to serve it, even if it means losing a day off. I imagine most people attending feel rather that the system should be there to serve them. Since attendance at a congress is basically a holiday and a strenuous diet of long games for many days is not normal for most amateurs, I'm sure that a rest day is seen as an important, and maybe vital, part of the experience. Spoiling a holiday even for just one person is surely not what the EGF wants.

If you read the discussion again, then you will notice that it refers to the number of players in relegation. Since the evidence is 5.6 on average, "very large number" is totally wrong.



I do not participate in unnecessary meta-discussion.



Here you write as if relegation games affected all congress participants. They affect 5.6 players on average.



Presumably there are congress participants for whom that is so. There are also other participants for whom it is a combination of holiday and tournament play. Yet other participants consider the tournament play to be the major aspect of their congress holiday. Top players have yet another consideration: Participating to strive for winning the title.

It is only for the latter for whom relegation games can become relevant at all. That such participants belong to the latter group they express by signing to play in all rounds and wishing to play in the supergroup. Strong players not striving for winning the title do not need to enter the supergroup, and this has sometimes happened. Almost all are very eager to enter the supergroup though.

For players striving to win the title, you put up the consideration whether the players serve the tournament system or whether the system serves the players. Since those players are striving to win the title, a system that determines the title well is important to them. At the same time, the other direction also applies: Players do not want to play 18 hours per day in the championship. So surely there must be some balance. It is different for every player. Some strong players play in every tournament available - others play in only the main tournament. Some strong players play in Wednesday side tournaments - others don't. Therefore a general impression, which you try to paint, that strong players would per se not want to play in tournaments on Wednesdays is wrong.

There is good experience for how popular relegation games are among strong players: The Toyota-Oza-Denzo-Cup (or WTH was the name?) in Amstderdam was extraordinarily popular among strong players, although it had a dense schedule with relegation games after a heavy day during the evening before the first KO round. The relegation games were hard fought. From that, the most likely conclusion is: A tournament system with a profound seeding is very attractive for strong players.

From the statistics, there is another curious practical observation:

Relegation games affect only players with 4 wins in 7 rounds.

7, 6 or 5 wins are enough to qualify but 4 is still too weak a result. I think this is a very nice feature because players know when they have qualified for sure versus when they still need to show greater ability. Over 60% won games is the border.

Especially bad would be years in which most KO players would be qualified on 4 wins and SOS like in the years 2005 or 2009 with 5 such players.

Since you are the one who conflated two independent things to create an unintended further and thus "meta" topic, I'd say you do participate, with relish. Unless by meta-discussion you mean anything you don't like.

I can't say whether the majority of stronger players want to play every day if necessary, but it seems no-one else can either, short of a survey. However, they are a small sub-group relative to the attendance as a whole, and 5 or 6 of that small group may be a large proportion. Whether it is or it isn't, I can accept that a balance has to be struck and your solution may well be the best balance of all the requirements. But the real point here, surely, is that, on the face of it, the rules committee is unilaterally and even contemptuously dismissing the concerns of a national association. If the rules committee really wants to be respected for the merits of its proposed solution, at the very least a bit of diplomacy is necessary. The tail shouldn't be seen to wag the dog.

The brilliant relegation was invented by Matti.

The rules commission has not officially received the BGA proposal yet. Therefore we cannot meet its concerns as an NA yet.

I actually made the exact same proposal for the 2008 AGM. I don't think it's particularly brilliant, it's rather obvious actually.

I don't consider it especially bad. To be honest, I don't even consider it bad at all.

What is objectively bad about a system that says "get 3 wins or less and you don't qualify, get 5 wins or more and you do, get 4 and the required names will be drawn at random from a hat - you may or you may not"?

No-one can complain about not qualifying unless they won every game they played (technically the most they could have achieved) and fail on a random selection, or even on a tiebreak. Anything else to me leaves no cause for complaint. You want to qualify, make sure you win at least 5 games next time, otherwise you may not qualify. As long as this clear, I see no failing in the system.

Still: what happens if, let's say, eight players are tied for one place? Surely you cannot force them to play more than a single relegation game during the rest day. Is something like SOS, SOS-1, SOS-2 or some such system then used to determine who gets the right to play relegation and who doesn't?

_________________
My name is Gijs, from Utrecht, NL.

When in doubt, play the most aggressive move

This is what Robert explained above. In this specific case, the best two based on SOS (or whatever the current tiebreak criteria are) will play relegation, the rest is out.

In all cases, relegation for X places will involve 2*X players exactly, based on tie-breakers.

Ok, then the honour goes to you! Sorry for having missed the fact!

It is brilliant because
- exactly 8 players are seeded as desired and in a natural manner
- the top MMS players are seeded as desired
- there are not too many relegation games as one could get if necessarily all "lowest" MMS players had to play relegation games
- number of wins gets a decisive meaning
- SOS is not overinterpreted in as great a detail as SOS numbers tempt
- the rough essence of SOS that there is some relation between greater SOS and greater performance is kept to a reasonable extent