[Retros] Solidarity Chess (=SC)
Kevin Begley
kevinjbegley at gmail.com
Fri Dec 21 03:08:59 EST 2012
Andrey,
Actually, after further consideration, the logical algorithm is very
simple, and apparent.
Win > Stalemate > Draw (which you might stipulate as: "# > = > ½").
This might even have practical value, in the orthodox game!
This avoids what may be an unfortunate (and multifaceted) compulsion to
ditch your own units (as might be necessary in my previous suggestion
"ideal-# > model-# > mate > ½").
In fact, it might even be useful to stipulate some "#\=" problems this way
-- because the weaker side would almost certainly prefer to settle only for
the stalemate outcome.
In this light, the "#\=" goal would seem the lesser form (easier to realize
some theme, this way, but not the form you would prefer, if possible).
It is worth noting that you likely want all such problems to end in
stalemate (if a win by checkmate were possible, better to reduce the
stipulation to simply "Win"); however, you do not want any problem to solve
as direct-stalemate (you want the threat of checkmate to enter into the
play, and compel some concession, which allows stalemate).
This requires, of the composer, a rather skillful mechanism!
No* #*n (for any value of n), No *=*n (for any value of n), but *#\=n* (for
some value of n).
Presuming you do not want to disclose what n is (essentially, you're
interested only in studies, yes?).
If this is the case, it would seem wise to quickly survey theoretical
positions where stalemate can be forced.
Upon reaching such a theoretical position, quite naturally, I presume the
study would terminate, since the alternative would be to indefinitely chase
the opponent's lone King around the board, with your King and your pair of
Knights (not the worst idea in a game, perhaps, but far from valuable in a
study)!
Best,
Kevin
On Thu, Dec 20, 2012 at 11:09 PM, Kevin Begley <kevinjbegley at gmail.com>wrote:
>
> On Thu, Dec 20, 2012 at 1:01 PM, afretro <afretro at yandex.ua> wrote:
>
>> Naturally, in orthodox chess there can be only three outcomes: White
>> wins; Black wins; drawn. But in fairy chess – why not have a fourth, fifth,
>> etc, type of outcome?
>
>
> Theoretically, fairy chess should easily allow for the construction of new
> goals, based upon logically combinations of existing aims...
>
> For example,
> h#\=2 : help [checkmate-or-stalemate] in 2.
> (This is Win Chloe Notation, the popeye equivalent is: h#=2.
> I prefer something akin to Win Chloe's logical notation here -- seems more
> logical to use, | for OR, & for AND -- but, Win Chloe's notation is not
> always consistent.
> The inconsistency across software is also troubling. In Win Chloe, h#=2
> signifies: help [checkmate-or-possibility to capture of the enemy King] in
> 2.)
>
> h+(3)#2 : help [triple-check-and-mate] in 2.
> (Again, Win Chloe notation, I find no popeye equivalent -- and, I'd prefer
> "+(3)&#")
>
> It is, of course, also possible to invent new aims; but, this requires
> delicate consideration.
>
> Things get fuzzy when you introduce alternate strategies of play, such as
> reflexmates, and semi-reflexmates.
> There are powerful arguments, made by some, that these are actually fairy
> conditions (which restrict the opponent's movement, forcing the play of a
> checkmate in 1, whenever possible).
>
> In the latter case, I do not agree.
> I prefer to view stipulations as recursive, which typically (and by
> default) wind all the way down to completion of the ultimate aim (in this
> case, white King stands mated).
> [aside: it should also be possible to notate a halt in the recursion.]
>
> I don't look at a semi-reflexmate as a fairy condition, but as a modified
> goal, which is intended to be solved recursively (by default, down to the
> ultimate aim).
> The difference is, white's goal is not a traditional aim -- like mate "#",
> or like capture "x" -- nor is white's goal a logically combinatorial aim --
> like "#\=", or "+(3)#" (discussed above).
> Instead, the goal is a layered stipulation unto itself.
> In essence, no fairy condition is needed to describe the "semi-r#n" --
> white is simply in pursuit of forcing a position in which black can satisfy
> a "#1" stipulation, while black is simply resisting the goal (realization
> of the layered stipulation).
>
> However, the same can not be said of the full reflexmate.
> Here, there is no denying that a simplistic recursive model fails to
> describe the strategy; and, it becomes difficult to argue with those who
> would characterize this stipulation as merely a selfmate, under some
> implicit fairy condition (with a reflexive #1 property).
>
> The truth is, a reflexive fairy condition does a poor job of describing
> what is really intended to be a description of a considerably more complex
> motivation (for both players).
>
> And, this does pertain to your consideration of the possibility of a
> multi-layered (read: non-tertiary) game outcome?
>
> Consider the case that neither you, nor I, can force selfmate upon the
> another.
> If that were the intended objective of our game, we might consider this a
> draw.
> Or, as you have eluded, we might sub-parse the outcome, based upon a more
> complex algorithm.
> For example, we might define a weighted set of possible outcomes...
> s#1 : 1 full point,
> ideal stalemate : 0.875 points,
> stalemate : 0.75 points,
> none of the above: 0.5 points.
>
> In such cases, the play may become extremely complex -- and the strategy
> may become exceedingly subtle (to the point of mystifying)!
> Sometimes, you may be able to jockey threats of a higher outcome, to
> secure the best outcome. Other times, you may need to settle for the direct
> route (in which case, it may beg the question: why define a complicated
> array of weighted outcomes?).
>
> The complexity of playing for weighted stipulation, I dare say, might (at
> times) rival even that of the 3D chessboard!
> A game for the mentats.
>
> Beyond that, it requires some complicated notation, to describe the
> objective.
>
> If you really want to consider such possibilities, I would recommend
> trying to first establish some highly logical weighted set of outcomes.
> Maybe based upon varying types of mates (ideal, model, just mate, draw).
> I'd give it some careful thought.
> Once this is well established, try to stick with reuse of the known
> algorithm (for simplicity).
>
> I'd be very interested to see where this goes!
>
> Kevin.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pairlist.net/pipermail/retros/attachments/20121221/a8a3cdf3/attachment-0001.htm>
More information about the Retros
mailing list