Better go players simply make any move which is "good enough" when the flag is 
about to drop. When they play in that manner, their clock never runs out. This 
is quite frustrating for someone who hopes to catch others in time trouble. ( 
speaking from personal experience - I just lost a game this past weekend to a 
methodical player who used 29 seconds of every byo-yomi period while I still 
had a good bit of time on my clock. Give me sudden death :D )  Generally, by 
the time one is playing byo-yomi, the game is settled, and the clock time has 
been used to accumulate sufficient advantage that one need not sweat each and 
every end-game move.
 
Terry McIntyre <[EMAIL PROTECTED]>
They mean to govern well; but they mean to govern. They promise to be kind 
masters; but they mean to be masters. -- Daniel Webster

----- Original Message ----
From: Don Dailey <[EMAIL PROTECTED]>
To: computer-go <computer-go@computer-go.org>
Sent: Tuesday, June 19, 2007 8:11:15 AM
Subject: Re: [computer-go] Congratulations to GNU and to MoGoBot19!

On Tue, 2007-06-19 at 07:14 -0700, steve uurtamo wrote:
> Don, I like you very much, but when you say that byo-yomi
> is unfriendly to humans, I have to say that you clearly haven't
> played enough go.  Byo-yomi is incredibly friendly to humans.

You are not thinking clearly here because I am comparing 3 different
styles of time control and I ranked byo-yomi as friendlier than sudden
death by far.

Compared to sudden death, byo-yomi is very friendly - however, it still
puts you in sudden death like situations - racing the clock.   It's
clearly more unfriendly than Fischer time because you have little
control over your own time-allocation - you are always losing (or
maintaining) time, never accumulating.   

You perception of how friendly any time-control is has much to do with
the parameters.   Sudden death at 8 hours per side is far friendlier
than byo-yomi with really short time limits.  For example:

          Main time : 5 minutes.
      Byo-Yomi time : 5 seconds
   Byo-Yomi Periods : 3

isn't particularly friendly.   Any time-control style can be manipulated
to be comfortable or uncomfortable for humans.   If you try to compare
these systems of time control you must have settings that approximately
equal each other in terms of how long an expected match will last.

I would also point out that by adjusting parameters byo-yomi OR Fischer
becomes (or approaches) sudden death.  Just set time and period to zero
for byo-yomi.   So the practical difference is how much control you have
over your own time-allocation and this is clearly greatest with Fischer
time. 

Also, byo-yomi has the bizarre and illogical characteristic that the
player who used the least amount of time could be the one to LOSE on
time.   That's a side-effect of the characteristic that byo-yomi gives
you less control over your time allocation.   If you believe that makes
it friendlier than Fischer time,  then I believe you are not thinking
very clearly about this.   You are probably just making a judgment based
on what you are personally most familiar with, not what is objectively
best.   

Playing actual games with byo-yomi time is totally useless for judging
how friendly it is - your perception will be colored if those matches
are played with liberal byo-yomi time and liberal number of periods.  It
will be comfortable and you will be happy.   So you can only talk about
the characteristics of each type of time period and you have to reason
it out.   With byo-yomi you can get into time-crisis situation that
never go away.

Here is a table:

  1.  sudden death - very unfriendly to humans.

  2.  byo-yomi - you can get into time-crises situations that you can
never 
      recover from - but if the byo-yomi time is liberal, at least you
can 
      never be forced to move instantly.   

      byo-yomi is easy to manage for computers.  A computer can simply 
      always spend almost all of the byo-yomi time no matter how obvious
the 
      move which also will have the benefit of annoying the human
opponent. 

  3.  Fischer - Best.  If you are really short on time, you can gain
time on
      your clock by playing easy moves more quickly.

Sudden death is best for computers when playing humans but it's not the
easiest to manage for computers.  It's just that humans aren't good at
it.

So the ordering above is in worst to best order for humans but just the
opposite for computers.   

Fischer is also hardest to manage for computers.  None of these are
difficult to implement for computers,  but computers are very poor
judges of how to allocate time wisely - they pretty much have simplistic
algorithms for time management that doesn't consider (or does a poor job
at) how difficult or critical the decision happens to be.  Fischer is
wonderful for letting humans exploit this skill.   byo-yomi tries to
make you play at a steady rate - not friendly.

So if I wanted to play a very important computer/human match (and I'm a
computer) and I couldn't play sudden death,  I would clearly prefer
byo-yomi over Fischer, given the same approximate expected match length.

- Don
 


 

> If you don't like it, try canadian timing, which is also very
> friendly to humans.
> 
> Please, for the love of god, do not now make a chess analogy.
> Simply play a few hundred games of go with canadian, byo-yomi
> and fixed time to compare.
> 
> s.
> 
> 
> 
> 
>        
> ____________________________________________________________________________________
> Need a vacation? Get great deals
> to amazing places on Yahoo! Travel.
> http://travel.yahoo.com/
> _______________________________________________
> computer-go mailing list
> computer-go@computer-go.org
> http://www.computer-go.org/mailman/listinfo/computer-go/

_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/







 
____________________________________________________________________________________
We won't tell. Get more on shows you hate to love 
(and love to hate): Yahoo! TV's Guilty Pleasures list.
http://tv.yahoo.com/collections/265 
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/

Reply via email to