EGF and Fischer

[Liisa]

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

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.

[zinger]

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.

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.

[Liisa]

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.

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.

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.

[topazg]

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.

[Liisa]

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.

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.

[kokomi]

I still think 60 + n x 20' is better than 75. I do not like single byo-yomi time. With 75m sudden death, I do not set an alarm at 60m, telling myself i need to hurry up. At 60m, what i think is I still have 15m, not I'm in a status of byo-yomi and may run into lack of time.

Sudden death is just horrible, not for serious game. I would play fun game with sudden death setting.

[willemien]

More from a theoretical background

I would prefer [sl=BronsteinTiming]Brondstein or delay timing[/sl]

As with standard Byoyomi and canadian byoyomi is is spilling and spilling systems are supposed to lead to a more relaxed game (If by quick playing you lose the bonustime anyway, it is hoped that players therefore will play more relaxed)

maybe a job for the rules comittee to give guidlines for minimal timing requirements using the different time systems

[karaklis]

I am not sure whether I got it right what you wrote, but I think you can play in a more relaxed way if time is not spilt (because the player can accumulate an appropriate amount of time for difficult situations). Also it is fairer if it is guaranteed by the system that both players have the same amount of time available regardless of how much they have already used up. This only works if time is not spilt.

[Harleqin]

I think it needs to be stressed that time is not accumulated with Bonus time. This may seem contradictory, but I urge you to think about how the number of moves remaining and the time reserve are correlated. The accumulation is really an illusion.

Anyway, I believe that the idea to force the players to use more time through spilling is fundamentally bad. Even if one accepts the basic premise of this idea, there are many moves that are more or less automatic and do not warrant thinking for even 10 seconds. Players should generally be allowed and encouraged to use their time as they see fit.

[willemien]

I think it needs to be stressed that time is not accumulated with Bonus time. This may seem contradictory, but I urge you to think about how the number of moves remaining and the time reserve are correlated. The accumulation is really an illusion.

Anyway, I believe that the idea to force the players to use more time through spilling is fundamentally bad. Even if one accepts the basic premise of this idea, there are many moves that are more or less automatic and do not warrant thinking for even 10 seconds. Players should generally be allowed and encouraged to use their time as they see fit.

Am i forcing players to use more time?
I hope i did not do that, but the result of delay timing is that you don't get any "profit" from it.

Are the moves that do not warrant thinking for even 10 seconds sometimes the losing move?

I urge you to think about how the number of moves remaining and the time reserve are correlated

What do you mean?

I am not sure whether I got it right what you wrote, but I think you can play in a more relaxed way if time is not spilt (because the player can accumulate an appropriate amount of time for difficult situations). Also it is fairer if it is guaranteed by the system that both players have the same amount of time available regardless of how much they have already used up. This only works if time is not spilt.

I don't understand what you mean here.
Delay time garantees that players have the same amount of time available regardless of how much they have already used up. (they always get the bonus time for every move) the difference is that they cannot transfer unused bonus time to a later period.

Maybe i was not clear the idea is that with delay timing the bonus time is higher than with fisher timing,because it is spilling. (Did i forget to mention this )

I think that one of te reasons for the introduction of delay timing in chess was that players always had time to write down the move on their gamesheet.


The background to my reply was the [sl=ShodanGoBet]Shodan Go Bet[/sl] [http://dcook.org/gobet/details.html] where the timing system that is used is Fischer time control. 40 mins + 20 secs/move
I think this is quite in favour of the computer, in the first F moves the computer will hardly use any time (the moves are just picked out of the opening or joseki book) and so the computer has F x 19 seconds extra time for when the middle game starts.

I think for the human ([sl=JohnTromp]John Tromp[/sl]) a delay time control of 40 mins + 30 secs/move would give a beter game.

[topazg]

... in the first F moves the computer will hardly use any time (the moves are just picked out of the opening or joseki book) and so the computer has F x 19 seconds extra time for when the middle game starts.

This is logically false.

The computer will have F x X seconds extra, where X is the average time per move of John Tromp minus the average time per move of the computer.

[HermanHiddema]

I think it needs to be stressed that time is not accumulated with Bonus time. This may seem contradictory, but I urge you to think about how the number of moves remaining and the time reserve are correlated. The accumulation is really an illusion.

Anyway, I believe that the idea to force the players to use more time through spilling is fundamentally bad. Even if one accepts the basic premise of this idea, there are many moves that are more or less automatic and do not warrant thinking for even 10 seconds. Players should generally be allowed and encouraged to use their time as they see fit.

Am i forcing players to use more time?
I hope i did not do that, but the result of delay timing is that you don't get any "profit" from it.

Are the moves that do not warrant thinking for even 10 seconds sometimes the losing move?

Moves that do not warrant thinking time are, for example:

• Taking back a ko when your opponent has responded to your ko threat.
• Responding to ko threats that are obviously large enough.
• Playing out the moves of a squeeze that you have decided to play
• Playing out standard sequences (e.g. hane and connect endgame)

With such moves, if they are playing with 30 seconds delay, then you're effectively forcing players to use 28 seconds on reading other things if they want to use their time optimally. Which is annoying, IMO.

[willemien]

... in the first F moves the computer will hardly use any time (the moves are just picked out of the opening or joseki book) and so the computer has F x 19 seconds extra time for when the middle game starts.

This is logically false.

The computer will have F x X seconds extra, where X is the average time per move of John Tromp minus the average time per move of the computer.

Did i made an error again?

Suppose

Opening is 20 moves (40 stones)

The Computer is using his opening book uses only 1 second per move

Using Fischer time control. 40 mins + 20 secs/move

The computer has after the opening 40mins + 20 x 19 secs(unused fisher time) = 40 mins + 380 secs = 46 mins 20 secs remaining.

Using Bronstein time control. 40 mins + 30 secs/move

The computer has after the opening 40mins + nothing = 40 mins remaining

(And I was only refering to this 6mins 20 secs difference)

For John Tromp to have the same time under Fisher time control he also has to play 1 move per second.
While under Bronstein time control he won't get any difference as long as he is under 30 sec per move.

[topazg]

(And I was only refering to this 6mins 20 secs difference)

Ah, I was assuming you were comparing as advantage the computer gains over the human due to thinking times. Sure, if the human is slower but within 30 seconds there would be no difference between them at the end of the opening, but I'm not sure how much difference this makes or relevance this has. With 30 second delay timing, there is never any benefit in making a single move in less than 30 seconds, even if it is protecting against an atari of a 45 stone group, and I think this unnecessarily drags out the game. It is possible for the two sides to end up with very unbalanced overall times, and I don't consider this to be much of a benefit in a system.

Herman has raised some excellent points IMO.

[karaklis]

The accumulation is really an illusion.

Of course. With accumulation I mean the time you can use for a single move until you run out of time (e.g. because there's a complicated situation on the board and you need a lot of time to ponder on this move).

I don't understand what you mean here.
Delay time garantees that players have the same amount of time available regardless of how much they have already used up.

There are situations where a player can (and wants to) play a sequence of moves rather quickly. In the delay time system this time is spilt whereas in bonus time this time can be saved for later. It is obvious that spilling time is disadvantageous.