Perpetual check

14 views
Skip to first unread message

iken...@cix.compulink.co.uk

unread,
May 19, 1997, 3:00:00 AM5/19/97
to

I thought I understood what a perpetual check is until I read the recent
discussions about game 2 of Deep Blue-Kasparov. I had always assumed
that a perpetual check was *not* a special kind of draw by repetition
(otherwise everyone would just call it a draw by repetition) and that
therefore a computer could not (strictly speaking) ever find it since it
has no repeating pattern.

So my question is, how is a perpetual check actually adjudicated over
the board? And is it true that perpetual check is precisely a special
kind of draw by repetition but just with one side playing a check on
every move?

Most games I have seen which have ended in a perpetual check have not
been of the repeating position type but rather an endless stream of
checks which could be continued randomly ie in a fashion that would never
repeat (assuming the human could maintain their randomness!). In this
case it would never be found by a computer - although presumably after 50
consecutive checks you would draw by the 50-move rule and feasibly you
could search to the end of that.

Ian Kennedy

-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet

Komputer Korner

unread,
May 19, 1997, 3:00:00 AM5/19/97
to

Perpetual check is in fact a draw by repetition. Eventually the
position will be repeated 3 times.
--
Komputer Korner

The inkompetent komputer.

Siva Chander

unread,
May 20, 1997, 3:00:00 AM5/20/97
to

Komputer Korner <kor...@netcom.ca> wrote:

: Perpetual check is in fact a draw by repetition. Eventually the

: position will be repeated 3 times.

Yes, but it could require a very deep search to find that even fairly
simple perpetual checks are drawn by repetition. If the moves on one
side (say the King moves) are forced, then it helps by pruning the
tree a lot and the program may find it (especially if it looks at
checks first).

However, if there are 'quiet' moves and interposing moves (ie a piece
blocks the line of check) which are unforced, then the tree grows much
faster. In game 2, we had all of these factors which made the perpetual
impossible to find.

1) The checks begin several moves after Qe3
2) There are many lines with quiet moves for black like Rd8, Rf8 as well
as for white like Bf3, h4 etc.
3) There are interposing moves by the white bishop which are unforced.
4) There are white king moves that are unforced.
5) There are a fair number of basic repeating positions (even with just
5 checking positions, it takes (5 x 2) x 2 + 1 = 21 ply to get to the
first three-fold repetition.

si...@emi.net

Kevin James Begley

unread,
May 20, 1997, 3:00:00 AM5/20/97
to

Komputer Korner (kor...@netcom.ca) wrote:
: Perpetual check is in fact a draw by repetition. Eventually the
: position will be repeated 3 times.

Not entirely ture. In the case of DB-GK perhaps, but...

Perpetual check either results in a three-fold repetition,
or it may result in draw by the fifty-move rule.
In either case, one cannot claim draw by "perpetual check."

Kevin.


Komputer Korner

unread,
May 20, 1997, 3:00:00 AM5/20/97
to

Siva Chander wrote:

>
> Komputer Korner <kor...@netcom.ca> wrote:
>
> : Perpetual check is in fact a draw by repetition. Eventually the
> : position will be repeated 3 times.
>
> Yes, but it could require a very deep search to find that even fairly
> simple perpetual checks are drawn by repetition. If the moves on one
> side (say the King moves) are forced, then it helps by pruning the
> tree a lot and the program may find it (especially if it looks at
> checks first).
>
> However, if there are 'quiet' moves and interposing moves (ie a piece
> blocks the line of check) which are unforced, then the tree grows much
> faster. In game 2, we had all of these factors which made the perpetual
> impossible to find.
>
> 1) The checks begin several moves after Qe3
> 2) There are many lines with quiet moves for black like Rd8, Rf8 as well
> as for white like Bf3, h4 etc.
> 3) There are interposing moves by the white bishop which are unforced.
> 4) There are white king moves that are unforced.
> 5) There are a fair number of basic repeating positions (even with just
> 5 checking positions, it takes (5 x 2) x 2 + 1 = 21 ply to get to the
> first three-fold repetition.
>
> si...@emi.net

Agreed, but the original poster asked what was the difference. The
answer is as you have stated, they are the same concept differing only
in complexity.

Reply all
Reply to author
Forward
0 new messages