Since there is no mentioning in EGF rating page Fischer time, I would like to make examples and suggestions for minimum Fischer settings for different classes.

EGF C-class: 20 10f
EGF B-class: 30 15f
EGF A-class: 40 20f

We should also have AA-class for EGC:

EGF AA-class: 80 40f

Where Fischer adjusted times are 40min, 60min, 80min and 160min respectively. So they will meet minimum EGF requirements by a margin. I have calculated Fischer adjusted time where it is assumed that 2×120 moves are played during the game.

Other, more complicated way to calculate the conversion for EGF tournament classes is, that first the "basic time" is calculated where it is assumed that players do play 60 moves before entering to "overtime". Thus in the case of A-class tournament Fischer time would converge as 40min + 60 stones × 20 sec Fischer increment = 60 minutes. That will meet the minimum criteria for basic time in A class tournament. After that we can assume that Fischer is like Canadian overtime and 20 sec Fischer will be the same as 15 stones in 5 minutes Canadian. That means additional 20 minutes and we will get 80 minutes for adjusted time. 75 min adjusted is minimum for EGF A class tournament

I would propose to calculate adjusted time for Bonus timing as the time for 150 moves, which seems to match the japanese and canadian standards quite well. Class A would be met by Bonus time 38/15, class B by 25/10, and class C by 15/6.

I do not think that it would be useful to increase the requirements. I also do not see a need for an AA class.

_________________
A good system naturally covers all corner cases without further effort.

Thanks for making the translation, it looks ok to me.
I agree with you that Fischer time is a good time system, I believe it may be possible to argue that it is objectively better than any of the overtime systems normally used for go.

Other comments:
* Availability of clocks is and will remain a decisive factor.
* Byo-yomi with several byo-yomi periods is not too bad either.

From the other thread:
* The clocks still need to give some warning that time is running out.
* The fact that time may get short regardless of time system is not the point,
it's undesirable that games end in a silly way and they do more easily with
some time systems than with others.
* Much longer times are used in chess than in go, although chess players have
fewer moves to make. It's fun to play fast games too, but I think it's good
that there are go tournaments with more generous time limits as well. The EGC
is perhaps the best we have in that respect.

regards,
Henric

I would add that Fischer is not just objectively better, but it is also significantly better! And I am sure that most people will agree when they are used to it. Fischer requires some active time management to get most out of it. But requirements for the active time management is in any case far less demanding than with traditional overtime timings.

I think that the most significant difference is that it is not perhaps wise and not even necessary let time run close to zero, but it is good idea to keep decent buffer, e.g. 10 minutes or so, so that if things get difficult there is enough time for reading.


Clocks (e.g. Excalibur2) cost just around $30 so it is not big deal to gather enough clocks to the clubs. In Finland we already have Excaliburs more than enough to meet the demands of any local tournament. Also it is not wrong to have Fischer in first boards and Canadian overtime for the rest where only analog clocks are left. McMahon tournaments are unequal anyway by ordering players by rank. And having digital clocks is anyway advantageous for analog clocks even if with both clocks Canadian timing is used.


Byouyomi with several periods is not that bad for the players. But because some people are chouchikuns, who think that in 2×8 hour game it is wise to enter byouyomi when there is played just 64 moves out of 300. Therefore it is bad for schedule. And those who would like to finish their games during main time had to wait unnecessarily. So if staying in schedule is preferred, severe overtime is required e.g. 1-3×20sec or 20-25/5min Canadian. And that is bad for the players.




With Excalibur it is enough that it beeps 30-60 seconds before time runs out.

We really do not need any more audible features that just a single beep! (referring here to some other approaches taken by other game timer models. =)

Problem is with your calculations that in order to meet with EGF A-class requirements there must be at least 60 minutes time available to be used for the first 60 moves (60 moves is reasonable compromise). This criteria is not met by your calculations. Also when making calculations there must be enough time available even for 230 move game. That means that there must be at least 2×75 minutes time to play 230 move game.

I think 60 moves is too little, 90 is more reasonable.

Currently, A class requires at least one of:

• 75 minutes sudden death
• 60 minutes + 60 moves in 15m Canadian byoyomi
• 60 minutes + 45 moves @ 20s Japanese byoyomi

That would indicate that the rating list considers 45 moves in Japanese byoyomi, or 60 moves in Canadian byoyomi, sufficient to finish the game (since 75 min sudden death is considered equivalent).

Given a reasonable upper bound of ~150 moves per player, that would mean that the EGF considers 90-105 moves in main time reasonable, so I think main time calculations should be based on ~90 moves, not 60.

At 15 seconds Fischer bonus, as proposed by Harleqin, that would add 22.5 minutes to the main time, so 37.5 minutes starting time would suffice.

Just want to say, I love that we are having this discussion and so many bright minds and strong players are actively pushing Fischer timing.

Perhaps your argument is good. But I would still think that proper value is greater than 60 but less than 90. I rounded my figures upwards so that they will surely fit to EGF requirements. So there should be some room to go shorter Fischer times and still give decent games.

But 12 second increment is the same as 25 stones in 5 minutes Canadian. Therefore playing long over 350 move game with 6 seconds increment is just ridiculous and does not meet the spirit of EGF tournament classes. Thus I would set minimum possible increment to 10 seconds or 12 seconds.

The problem is that the EGF still also allows absolute time to be used. As long as that is the case, setting a minimum ratio of bonus to basic time opens an illogical gap: a tournament director could choose, for example, 75 min absolute or Bonus 50/10 to get class A, but 60/6 would not be possible, although it would surely be preferable to 75 min absolute.

You could, of course, also try to get rid of absolute time (I would support that), but I think that this would meet too big political resistance, regrettably. In any case, I would not put these two things into one proposal to the AGM.

_________________
A good system naturally covers all corner cases without further effort.

But it is not actively used! It is there just for reference. But if we set up minimum Fischer people may really try it in practice, although it requires lots of practicing and active time management skill from the players. People are not that particularly good in that. Even 20 10f may be difficult to control for unexperienced players.

But I understand your point and EGF tournament setting certainly are not logical. E.g. EGF considers 60 mins + 1×20 sec to fit to class A, but does not allow 60 min + 25/5min Canadian to fit class A. This just does not make any sense, because probability for time loss is huge in the former (for me probability of timeloss with 1×20sec byouyomi is more than 20% and with close game, probably more than 50%. 25/5min Canadian is still manageable, but 1×20sec is far too difficult for me.)

Perhaps we should rationalize the tournament classes, because they are not well defined. E.g. how many byouyomiperiods are available should be taken in consideration.

I would also set completely new tournament classes where:

class C: 20 10f
class B: 40 20f
class A: 60 30f

and the weights for Gor calculations should be 0.5×, 1.0× and 1.5× respectively.

Actually, it is actively used. In the Netherlands we have the Eindhoven tournament with 75 minutes sudden death (class A), and the Arnhem tournament with 60 minutes sudden death (class B), both of which are large (40-60 participants) serious tournaments.

60/30 has a T300 (time for 300 moves) of 4:30 h, 40/20 3:00 h, and 20/10 1:30 h. Your class B would require significantly more time per game than the current class A.

I think that the current classification sees class A as standard: "this is enough to be included in our rating system". The lower classes (B and C) are "OK too, but we weigh them less".

Your proposed increase of requirements and addition of "more than standard" is a big paradigm shift: I find it questionable whether "even more time" warrants to weigh these games even more.

I believe that it is better to go step by step. First, introduce Bonus timing to the EGF standards, with no other strings attached. Calculating the adjusted time (TA) as the overall time for 150 moves seems like a sensible proposal.

_________________
A good system naturally covers all corner cases without further effort.

The Leuven/Louvain/Löwen-Tournament (Belgium) uses 75 min absolute time without byoyomi.

_________________
SL | EGF | KGS | GMT+1

My assumption was clearly wrong. In that other thread there was not much fans for absolute time without increments and it was considered absolutely worse than overtime or time bonus. I think now as aided by real examples that the difference is far less clear.

Ceratinly for tournament organizers, absolute time is an attractive option, because it is the easiest to schedule for. Whether that reflects any real popularity among the players, I don't know.

C: 10 10f
B: 25 12f
A: 30 20f

After reconsidering, these would be my suggestions for preferred minimum Fischer times. Fischer adjusted times for T270 are 2×32½min, 2×52min and 2×75min.

Of course time controls like 30 2 bronstein are perfectly acceptable. These settings were tried in one rapid tournament in Finland 2008 and in my opinion this should be good time control. But if we use increments that are smaller than 12, we should be very cautious with calculating adjusted times. And Txxx value should be greatly reduced from T270.

Absolute time may be attractive for organizers, but in my opinion it is a horrible way to play Go. I would never, ever enter a tournament with such a time control.

_________________
Main Entry: zing·er Pronunciation: \ˈziŋ-ər\
1 : something causing or meant to cause interest, surprise, or shock
2 : a pointed witty remark or retort

In my opinion 75 minute absolute is far more easy to manage and control than 60min + 1×20sec Japanese byouyomi. I would never participate 60min + 1x20sec tournament. One time in EGC 2010 weekend was perfectly enough. Result was that I lost about 25 gors due to timeloss.

I disagree. Losing on time in this case is just poor time management, and not a fault of the time control. In absolute time, if you get low, your opponent can start playing complete nonsense until your clock falls. With 1 x 0:20, you can at least respond to every nonsense move without risk.

_________________
My Go videos | My Go Teaching Ladder (GTL) reviews
I also offer Go teaching lessons
-- Creator of OGS --
"I played a trick play and my opponent fell for it, unfortunately he didn't know the correct followup, so I died"

With 1×20 sec byouyomi, on average 10 seconds is spilled per move. Therefore you have same amount of time allocated for the game as in 75 min absolute, if you play 90 moves in byouyomi. This is of course ridiculous. And those who organized the EGC 2010 tournament did not show much of common sense but just exploited the illogicalities of EGF tournament classes. Weekend tournament should have been sent in class B.

But of course there are requirements that people behave ethically. And from statistical standpoint I think that at higher level, less than 1 out of 20 games will be decided this way. But this is also because you did not leave big enough buffer (e.g. 5 min) so we can count your hypothetical loss due to time pressure. And it is ok. It just shows your poor time management skill. With absolute time, you have always room for active time management. With byouyomi, you are forced to play more than 50 moves in byouyomi where you do not have any options for time management (expect playing a forcing move to gain extra period).

With Fischer time bonus to absolute time we can reduce the time management skill requirements, because total amount of time allocated for the game is in constant ratio to the number of moves played.