[Retros] rights & ocassions / not answering Andrew anymore

Guus Rol grol33 at gmail.com
Wed Jun 4 13:58:43 EDT 2014

Hi Andrew,

I get your point but my paradigm is fundamentally different. When I explain
some of the basic components you will see why. I will first skip the matter
on DR and 3R since (a) the DR-rule is special in the sense that it looks
forward instead of backward; you remember we had issues with looking
forward between 3R and "castling right" (b) 3R and 50M are not really retro
conventions. You can see that by applying them outside the retro framework.
Someone solving a problem may encounter automatic 3R draws and 50M draws
even if the full 3R or 50M content starts after the diagram - no
information holes to fill but everything in plain sight. E.g. a helpmate in
55 moves may be aborted by the 50M convention even when neither side is
interested in claiming a draw. Now there are problably a few individuals
who will claim that even under these circumstances the automatic draws
don't apply but I am afraid they simply don't want these conventions at
all. And there are some who will only apply a convention when convenient to
solve a problem. I will come back on that in the "role play" later.

You provided a more useful example for analysis in one of your previous
posts. The absolutely insignificant and therefore magnificent example of a
diagram which is mate unless someone has e.p. right. What better way to
demonstrate the scarcity of retro-active attributes in orthodox chess when
we need to include such outlandish examples? Your view was that mate
required that all exits were closed and such was not the case. On the other
hand, e.p. would not be allowed by convention taking us into a judicial

Whether or not I reflected your viewpoint precisely, the example still sets
the scene for a fundamental understanding of the relationship between rules
and conventions. We begin without convention and then make the observation
that the solver of a retro problem is not playing a game at all, but he is
managing a cluster of games. Every choice he makes is made in millions of
different games simultaneously all of which may materialize as the "actual
game" in some future variation. Of course "millions of games" is an over
the top theoretical number since no more need to be included than to
cover the range of different retro-active states. But the essential point
is that every uncertainty means that you are playing more than one game
simultaneously. There are no FIDE laws related to uncertainties in
positions and to more than 1 simultaneous game. Thus, mate is defined as
the closure of all exits in 1 particular game. But by you own quote: "how
can a FIDE law know how to handle an uncertainty in a retro-active state
when such can never occur in a legal FIDE game?" One can state with FIDE
law in hand that mate implies that a King will be captured after all
possible moves, but one cannot state from FIDE law that such must be the
case in all possible games connected to a retro-active problem position.
This is the point where the retro converntions come in. The total function
of the retro convention is "to reduce the game cluster in such a way that
the FIDE laws can be applied". The FIDE laws can only be applied to known
states. Example: the castling convention means that if you express an
intention to castle, the conventional rule is applied and all games without
castling right are removed from your game cluster, and only games with
castling right remain. Then the FIDE castling law can be applied.

Intermezzo: I am not at all sure that I explained this well and so I will
do it again from a more mathematical angle. The whole thing is abouts "sets
of games" versus "sets of moves". FIDE laws can handle "moves and sets of
moves" but not "sets of games". For this we need the conventions. And there
is of course the middle ground where FIDE laws do handle a "set of games"
but only if intended actions are unambiguous for all games in a set.

What we see here is that conventions do not serve at the tail end of the
FIDE laws to get rid of the garbage but have a complete domain of
themselves separate from the domain of the FIDE laws. They are always
applied before the FIDE laws are permitted to take hold. Some would
claim that the conventions are therefore senior to the laws but I wouldn't
go to war over that. The Pope has expelled people from the church over
lesser crimes.

How does it work for the example we started with "mate with or without e.p.
right"? Well, we have a law about mate in a game but not a convention about
getting rid of unwanted games in a potential mating position. Could we
create one? Yes, example: If there is any game in the proof game cluster
where King capure is guarenteed for all counter moves, then we have a mate.
Note that there is a clear distiction between the conventional part and the
FIDE law in this approach. They do not bite but each handle their own
domain! Do we need a convention for handling mate? No, not at all. The
primary reason is that mate, stalemate and even DR are "terminators" but no
"retro-active attributes". Whatever seems uncertain and retro-active about
them comes from real retro-active attributes, e.p. right in the example.
The terminators only represent "outcomes" and are therefore at the end of
the evaluation chain. In this case, the convention evaluating e.p. right
will temporarily remove all cluster games with e.p. right and find that
there is nothing left to do. Thus, mate will occur by FIDE law. As
explained before, the mating condition is absolutely certain in all games
handed by the conventions to the laws; there is no bargaining with FIDE
laws here.

I chop my post off here since it is getting too long and too late. Much
more to say though

Best wishes, Guus Rol.

On Tue, Jun 3, 2014 at 6:26 AM, Andrew Buchanan <andrew at anselan.com> wrote:

>  Dear Guus,
> I still don’t see that a rule can know that a convention will trigger. For
> example, en passant convention indicates whether a move is *permitted*, but
> there are no retro-analytic implications, except in the branch of moves in
> which the en passant is actually played. A1.3 will not be able to use the
> absence of proof of e.p. legality to terminate the game.
> Similarly here, with 3R convention currently a draw may be automatically
> asserted. But A1.3 will have no visibility that a mere convention is going
> to be triggered. How could it?
> This distinction between rules and conventions may be different to how
> you’ve seen it, but it is clean and consistent. Please think about it. I am
> a bit troubled that you don’t get it.
> Thanks,
> Andrew.
> *From:* Retros [mailto:retros-bounces at janko.at] *On Behalf Of *Guus Rol
> *Sent:* 02 June 2014 21:37
> *To:* The Retrograde Analysis Mailing List
> *Subject:* Re: [Retros] rights & ocassions / not answering Andrew anymore
> Hi Olli,
> Things are not that complicated. You can always construct one proof game
> for the diagram in less than 41.5 moves in which Rg8+ draws because of 3R
> occurring at the next move Bf8. This is sufficient to dual the intended
> solution. Of course, still on the premise that 3R is automatic by
> convention.
> Best wishes, Guus.
> On Wed, May 28, 2014 at 4:16 PM, Olli Heimo <olli.heimo at luukku.com> wrote:
> Hi Joost-
> I can't decide if I agree or disagree. The Codex says: "Article 18 -
> Repetition of Position
> A position is considered as a draw if it can be proved that an identical
> position [21] has occured three times in the proof game combined with the
> solution". So after the forced 32. - Bf8 the game is over and neither
> player can anymore checkmate. It follows that the game is over already
> after 32.Rg8+. But one can't prove that it is an identical position. In
> certain proofgames it is a fact, but if e.g. 19.Na3 Rb8 20.Nb1 Ra8 it is
> not identical. There are several rules competing with other rules. Which
> one is the strongest? Olli.
> Retros Probleemblad kirjoitti 27.05.2014 kello 17:51:
> > On 05/27/2014 11:57 AM, Guus Rol wrote:
> > > Hi Olli,
> > > Yes, you got the idea! I am not sure about the precise position and
> > > timing but basically DR cooks it if you aim for the position after Bf8.
> >
> > I don't think so. DR uses article 5.1b ("The game is drawn when a
> > position has arisen in which neither player can checkmate the
> >  opponent?s
> > king with any series of legal moves."). Since ignoring the 3-fold
> > repetition is legal, there's a legal continuation in which any colour
> > can checkmate.
> >
> > Joost
> > _______________________________________________
> > Retros mailing list
> > Retros at janko.at
> > http://www.pairlist.net/mailman/listinfo/retros
> _______________________________________________
> Retros mailing list
> Retros at janko.at
> http://www.pairlist.net/mailman/listinfo/retros
> _______________________________________________
> Retros mailing list
> Retros at janko.at
> http://www.pairlist.net/mailman/listinfo/retros
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://one.pairlist.net/pipermail/retros/attachments/20140604/29f93c86/attachment-0001.html>

More information about the Retros mailing list