﻿﻿ MatPlus.Net

Website founded by
Milan Velimirović
in 2006

4:18 UTC
 ISC 2024
 Headlines Forum* Fellows Members DL Archive Links

Remember me

 CHESS SOLVINGTournamentsRating lists1-Jul-2024
 B P C F

MatPlus.Net Forum General "Who is mated?"

### "Who is mated?"

In an orthodox problem that's no problem at all, just check the check.
But there are so many rules possible in fairy, with the effect that a check is only apparent
(maybe even only after retro considerations) - has anyone ever pulled this stunt?

Alternatively, this joke surely has been done x times...
(= 5+3 )

Black is not mate (illegal check), rotate by 180° and White is mate

EDIT: And the next obvious question, do you know fairy conditions
where Black can be mated without being in check? (Surely there are...)

And the next obvious question: Why is white mate when the board is rotated?

About the bonus question: Mate is defined by a check that cannot be parried, so in order to be mated one must be in check. The definition of check is of course configurable (e.g. in Vogtländer chess, a side is in check if it threatens the opponent king, or SAT where a side is in check when the king has flight squares).

Ask the stupid pawn! :-) (Diagram edited)

@Hauke,

QUOTE
do you know fairy conditions where Black can be mated without being in check?

Republican Chess - there are no Kings; if the side which has played can put the opposite King on a square where it would legally fulfil the stipulated aim/goal (e.g., checkmate, stalemate, etc), then the stipulated goal is met.

Thus (as I understand this condition), if the stipulated aim/goal is checkmate, a King may never actually be seen in check, even when checkmated.
If your stipulation asks, "who is checkmated?", it is safe for solvers to assume that checkmate is the stipulated goal.
If your stipulation asks, "who is stalemated?", the solver should assume stalemate is the stipulated goal.
If your stipulation asks, "who is in check?", a King might be both checked and checkmated.

If asked, "who is not in check?", the solver should call a philosopher (or an exorcist).

Note: Studies are best treated as a conditional aim/goal, where a King may be dropped to achieve either checkmate or stalemate (when both drops are possible, consult a magic 8-ball).

It's arguably bad practice to draft the rules of a fairy condition such that they are dependent upon an aspect of the stipulation (or worse, an aspect of an implied stipulation), but this condition may be considered an interesting exception.

ps: Is there an Exorcist Chess? It makes me think of a funny chess comic ("Take Me!").

In republican chess, the king is added on the board as part of the move, so in the final position there actually is a king on the board.
There's also republican II chess, where the side whose king is added first can reply with a move and adding the opponent king if that move/king addition breaks the own mate and leaves the opponent king mate.

Stephen Emmerson
Variant chess 1998
(= 3+3 )

#2 Republican II
1. Rd3 [2. Rb3 (Ka4/a5)#]
1... Rg1 (Kd1) 2. Se3 (Kf1)#
1... Rf7 (Kd8) 2. Se6 (Kf8)#

@Joost,

If the stipulation asks "who is checkmated?", the diagram shows no King (thus no check is seen, despite the fact that a King is essentially checkmated).

If you doubt that the King is essentially checkmated, even before the King is dropped on the board, I would remind you that addition of this King is mandatory.
To prove this fact, consider a helpmate in this condition. The black player can not disregard the possibility of checkmating the white King, en route to helping checkmate the black King. The King drop is not optional, thus a King may be considered checkmated even before the drop (the drop is automatic and implied) -- which may provide a sufficient answer to Hauke's question (he would need to elaborate, and he might not wish to elaborate, as this might reveal his idea).

note: I would presume solvers may disregard the possibility of adding a stalemated King, in a helpmate problem (just as the possibility of a checkmated King drop may be disregarded when the stipulation is help-stalemate), but I suspect different interpretations would be desirable here (and our present nomenclature does a poor job of expressing these possibilities, which is why I would reiterate that fairy conditions are best rendered independent of any aspect found within the stipulation).

You make a good point: I neglected to mention the possibility of intercepting a checkmate with a counter-checkmate (adding the second King), but it's important to note this analysis presumes checkmate is the aim/goal.
If the problem asks "who is checked," the same logic should apply (since the rules of the condition are, for economical purposes, modified by the aim/goal stipulated) -- you may intercept a check by dropping a second King (providing this results in cross-check).

It's worth noting this condition may break down when you introduce a stipulation with an aim/goal which does not directly impact Kings.
For example, what if your stipulation is help-en passant? Need the solver add a King to prove en passant legal? But, the King is supposedly dropped after the move (thus the stipulation should have been met prior to the drop).

This may be completely irrelevant to what Hauke is asking, and it's not my intent to pursue a tangent to the detriment of his thread. I only mention this is a strange feature of this condition (which should be considered), if it does indeed satisfy his question.

QUOTE

EDIT: And the next obvious question, do you know fairy conditions
where Black can be mated without being in check? (Surely there are...)

And if you interpret this literally: In checkless check there are no checks, only mates. So a black king will never be checked, only be mated.

Also note that some people argue that a mate isn't a check, so if the goal is +, it cannot end with a mate.

@Joost,

QUOTE
Also note that some people argue that a mate isn't a check, so if the goal is +, it cannot end with a mate.

First time I have heard about this. Very interesting! Thanks.

Christian Poisson
Problemkiste 1989
(= 2+1 )

ser-+6 Circe

5. h8=Q 6. Qg8

This is given as C+ in Winchloe.

This offers very interesting possibilities -- especially with the right fairy condition(s) -- for a series-mover based entirely on checkmate avoidance, which could otherwise be solved in 1 ply.

(11) Posted by shankar ram [Friday, May 10, 2024 14:25]

One more difference between Popeye and WinChloe:
WKc5/WPa7/BKa6 +1
WinChloe: 1.a8=R+
Popeye: 1.a8=Q/R+

And for the Poisson Ser-+6, Popeye gives 6.Qh8-a1 +, 6.Qh8-b2 +, 6.Qh8-a8 +, 6.Qh8-g8 +, 6.Rh8-a8 +

There are two issues here:
1. Is a checkmate a check?
2. Where a problem stipulates its goal as "+", does a line ending in checkmate (and fulfilling all other aspects of the stipulation) solve the problem?

I see 2. as a matter of convention as to what problem stipulations mean. By contrast, 1. is true by the laws of chess.

In Checkless Chess a checkmate is legal, but a check is not.

If this is indeed to be the accepted convention, somebody really should explain the reasoning behind this nonsensical convention.

If checkmate (literally defined to a King in check with no legal option to escape check) does not satisfy a check stipulation, why should double-check (or triple-check, or n-tuple-check) satisfy a check stipulation?

You can not argue that the achievement a higher order stipulation (#) invalidates the lower level stipulation (+) if double-check (++) satisfies the check (+) stipulation.

In these cases, the better nomenclature should win out, but our nomenclature is flawed to the core, and we have no intelligent authority to arbitrate bad ideas.

We might have used formal logical symbols, to represent "help-(+ and not #)-2" with something like h+&!#2, but these symbols are presently corrupted by meanings created by careless/goofball inventors.
Beyond that, the experts in problem chess have a very poor understanding of formal stipulation theory (most still can not distinguish between a fairy condition and a stipulation).

With any luck, some outsider will one day revolutionize the problem chess community's understanding of their own artform, and some future generation will settle the matter (discarding every bad idea that has been passed down to them).

At the moment, bad standards serve as the best defense against domination from an AI problem composer.

Below is ChatGPT's take on "Formal Stipulation Theory":

Formal stipulation theory is a concept within philosophy and logic that deals with the nature and structure of stipulations, particularly within formal systems such as mathematical logic or legal frameworks. Stipulations are essentially rules, conditions, or agreements that are established by convention or authority, rather than being inherently true or false.

In formal stipulation theory, the focus is on understanding how these stipulations are formulated, how they interact with each other, and how they affect reasoning and inference within a formal system. This theory explores questions such as:

1. How are stipulations expressed within a formal language or system?
2. What are the criteria for a stipulation to be considered valid or acceptable within a particular context?
3. How do stipulations influence the interpretation and application of formal rules and principles?
4. What role do stipulations play in determining the validity or soundness of arguments within a formal system?
5. How do different interpretations of stipulations affect the outcomes of logical reasoning processes?

Formal stipulation theory is closely related to other areas of study such as formal logic, philosophy of language, and legal theory, as it deals with the foundational principles underlying the construction and evaluation of formal systems and their applications in various domains.

Seems sound enough and not an "AI hallucination"!

Does a double-stalemate satisfy a stalemate aim?

@Shankar: Never believe ChatGPT a single word. "Formal Stipulation Theory" gives no Google hits.
Nevertheless, if you replace "stipulation" with "axiom" the stuff sounds legit.
Remember, an axiom doesn't need any evidence. Defining "mate" as "check plus something"
would be my preferred axiom (consequently, "both sides stalemated" is "stalemated
plus something" for me), but nevertheless, chess is a game, and the rules are convention.
For example, in kiddie chess training, it is a common task asking them to sort positions into
"Play/Stalemate/Check/Mate" - and sorting a mate to check would be an error.

QUOTE
Never believe ChatGPT a single word.

That statement is not likely to hold up over time, but it does seem fair at the moment, particularly in light of the fact that ChatGPT fails to address the two real issues:
1) the value of a formal stipulation system depends upon the capacity to minimize a fully coherent stipulation nomenclature, and
2) any formalized stipulation should be independent from the possibilities of the system (the laws of physics, the rules of the game, the video game environment, the mathematical construct, etc).

For example, consider a selfmate in 3.
There are at least eight key factors in this stipulation (I'm likely missing a few):
1) What is the AIM (or the ultimate goal) -- in this case, the AIM is to checkmate the white King.
2) To which player(s) does this goal apply -- in this case, the white player is tasked to achieve the stipulation (but a composer needs an option to express if it is for black).
3) What is the motivation for any player not tasked by the stipulation -- in this case, the black player will do anything to avoid the AIM.
4) What is the immediate GOAL -- in this case, the white player wants to force black into a "compel mate" position (where black has no legal move other than checkmating the wK).
5) Who starts -- in this case, we assume white starts, unless retrograde analysis proves that impossible.
6) What is the move limit to achieve this goal -- in this case, white (the player tasked) has the 3 moves stipulated, and black must have 3 moves (unless white's not on the move the diagram).
7) Where does the problem end -- in this case, the selfmate is assume to end with checkmate, but how does a composer end with the compelled mate position (if this is desired)?
8) What is considered a cook (a faster selfmate, an alternative selfmate in 3, a position where black has multiple choices when reaching the compelled mate, etc)?

Note that where the problem ends (7) might impact what is considered a cook (8). If a forced-compel-mate problem ends prior to checkmate, it may be deemed no cook if black has multiple checkmating moves in the final (compel mate) position (for black).

The whole point of "formalized stipulation" in chess (a key concept which ChatGPT also fails to address) is to express the stipulation without the language dependency (and the heft) that had existed in the 1800s (when problems were commonly stipulated "White to play and force checkmate in seven", rather than as "#7").

There are things found within the key components of a stipulation which the stipulation can not itself define:
What is the nature of checkmate?
What is the nature of a move?
What is the move sequence?
etc.

These questions (and many more like them) should be defined by the rules of the game (or by some set of fairy conditions -- which must be defined to be entirely independent from the stipulation).
A formalized stipulation should only lay out the nature of the solver's task.
So, presently ChatGPT is useless in addressing this matter.

ChatGPT can't tell us why the reflex-mate stipulation is merely a selfmate with an additional (reflexive) fairy condition, nor can it tell us why a semi-reflexmate can be expressed without any fairy condition.

If you want to get formalized stipulations right, you need to set aside the failures of our stipulation system today (just as we set aside the wordy stipulations of the 1800s), and develop a formal classification system which doesn't just identify categorization by our failed precedent, but provides a logical distinction between fairy conditions and stipulations.

Problem chess will never get there, so long it is controlled by humans with an undeniable self-interest to maintain bad precedents (which distort everything from titles on down).

An outsider will eventually need to formalize our problem stipulations in a coherent manner, and this will only catch on when a younger generation adopts it (especially database developers, solving tool developers, problem journal editors, etc -- the FIDE Album will be the last to overcome its deeply flawed categorization, because that's where the titles are decided).

I suspect an AI will need to do this work, because any AI composer will benefit from a more coherent, more formalized system.
When AI composers are leading the way, our flawed human system will be discarded (at which point, our entire edifice will reveal a plague of bad categorizations).
Some outsider will eventually completely rewrite problem chess, because we failed to provide any fundamental definition for the meaning of our terms.

And, when AI is done with formal stipulation, it's going to come back and address fairy conventions (read: it's going to rewrite the rules for fairy conditions).

What happens to the PWC problem you published when the AI decides pawns on the 1st should default to a single move forward, and checks?
Maybe your PWC will remain sound with the new default. Maybe it will require a second fairy condition.

Humans blocked every attempt to resolve these matters (and refused to establish logical conventions), but in the end our flawed system has no chance to prevail.
We are only surrendering any capacity to influence these conventions for ourselves -- we're going to let some outsider (bot?) from the future sort things out for us.
What do we care? We retain our titles, right? Ha, I wouldn't bet on that either!
Eventually, dominant AI composers are going to provide a better rating/title system for composers (which will apply retroactively).
Who will care if we ancients (who didn't know a stipulation from a fairy condition) awarded titles differently (to our pals)?

How soon will we arrive at this AI dominant future?
It may take an eternity if left to ChatGPT.
If this was a serious goal of AI developers, significant changes would be seen in a very short span.
Within twenty years, I would expect to see remarkable achievements in stipulations humans barely imagined possible (at which point people will begin to appreciate how our human failures have hindered the full flowering of this artform).