Mate in two (634)

The third curse of chess problems is that experts and novices face the same stipulations. If you tell an expert the theme, then that gives half the game away. (This is a skill inflation problem, and in this case, the curse corresponds exactly to one identified in broader gaming in the paradigm-shifting GDC video. Figuring out the point is essential.) Knowing the composer is sometimes a bit of a hint in figuring out the theme. The novice may solve a problem, but won't have the context to get half the enjoyment from it. This is a huge barrier to entry. Teachers like Rocky64 here perform a valuable role. Should really out his website, ozproblems.com: it's really informative, and his community of solvers is very friendly. My favourite group at the moment is Facebook's Chess Endgame Studies and Compositions, run by GM Max Illingworth and IM Cyrus Lakdawala, with Prof Stephen Dowd and author Satanick Mukhuty. Some famous names also participate, and it's very friendly and supportive. Other top sites with great communities are pdb.dieschwalbe.de, matplus.net, france-echecs.com and (for fairies) juliasfairies.com.

@anselan Ah, but you didn't try my suggestion, hxgxf and gxf.

OK so the other way is hxgxf, gxf, fxPe, cxd
White does axPb=X, b=X, c=X, h=X. There's even one Bl piece spare.

I allege that I am trying to get too many things done at the moment. Please check out the stupendous GDC video that I linked to.  There is also a rather sprawling disquisition about curses, and two curses in particular: (1) change (2) jokes in https://pdb.dieschwalbe.de/search.jsp

You have also been known to wield a dummy pawn or two in your time. So would also appreciate your response on that

@anselan

It took my all of 10 seconds to find the key move. I just seemed obvious a knight had to deliver the mate from a specific square, so I immediatly looked at 1. Rxe7.

I solved it by looking at the problem from the position of the king, rather than from the outside in, which apparently was in opposition to the theme of the puzzle.

@anselan That's an interesting video on gaming "cursed problems", and it kept reminding me of a PC game I play casually, rather than chess! Other than the "skill inflation problem", I had no idea how these ideas could apply to chess compositions, so it was enlightening to read your discussion with James. First, I think this type of important discussion should go on the Mat Plus forum. There should be a big thread about the PDB where people like yourself can announce new keywords and features, where people can ask questions on how to use the database, etc. 

Great job in identifying two "curses" in chess problems. Your answer to one of them, the Golden Age Principle, sounds good to me. It certainly makes sense to apply it when the outdated rule wasn't in dispute, but I wonder about cases like promotion-to-P, which AFAIK was never broadly accepted. Hence even the very early examples of this idea could have been meant to be jokey, perhaps?

Re the second curse, joke problem stipulation, I lean on the side of "spoiling" the joke, because I worry about solvers who don't want their time wasted. The thing is, even if a problem has the "joke" stipulation, that doesn't give away the unorthodox rule being used, so the solver still has plenty to think about.

Hi Rocky thanks for your kind and intelligent words. Let me respond.

• First, I think this type of important discussion should go on the Mat Plus forum. There should be a big thread about the PDB where people like yourself can announce new keywords and features, where people can ask questions on how to use the database, etc.

I have documented the Dummy Pawn & Golden Age keywords in PDB. I think the Golden Age one in particular is a bit of a PDB experiment. The documentation is really a manifesto (or short novel!) for a broader way forward, inspired by the Tanzanian folklorist & Fulbright scholar, Joseph Mbele. Although Mat Plus is a great place, it is not entirely troll-free, and I would anticipate tedious repetitive interventions from someone who apparently can't speak without using LOTS OF CAPITALS. Sigh. I don't want to go there until it's all a lot more solid. (I feel Dummy Pawn & at least one more issue should be bedded in happily first.) I would sooner discuss in the Facebook group I linked to, where there is rigorous prevention of naughtiness, and people are generally well behaved, although the attention span is low, given that it's Facebook - everyone is just writing on sand. Or here, or publish in a magazine.

• Golden Age [...] certainly makes sense to apply it when the outdated rule wasn't in dispute, but I wonder about cases like promotion-to-P, which AFAIK was never broadly accepted. Hence even the very early examples of this idea could have been meant to be jokey, perhaps?

History is written by the winners. I think Steinitz and others were sincere enough. These problems will still be tagged with Golden Age & Joke(not!) keywords, which of course the overwhelming majority of problems from that time do not. If the context tells us it was a joke (e.g. one I saw was published as April Fool) then certainly it's a real Joke not a Joke:not!. So I hope I have the balance right.

• Re the second curse, joke problem stipulation, I lean on the side of "spoiling" the joke, because I worry about solvers who don't want their time wasted. The thing is, even if a problem has the "joke" stipulation, that doesn't give away the unorthodox rule being used, so the solver still has plenty to think about.

I "lean that way too" in fact further, because I think if you say that you're only going to give a smilie (: as a clue, that puts a tight lid on the complexity of problem that you can demonstrate. Have you seen some of James Malcom's "joke" promotions? They are incredibly complicated. They weren't presented for solving in their superproblem originals, but for walk-through. I think that the chances of someone solving them with just a (: are very low indeed. The interesting solving only begins when someone knows what the joke family is. Joke problems are not even tested for soundness against other jokes, and creative jokey individuals will come up with off-the-wall explanations in any case.

This is the curse. I think the composer must ask (at the time of re-publication, if not before): what do I want someone to know in order to have a decent chance of solving it themselves? And I should put that in the stipulation. There is a complexity point at which (: is simply not enough. The problem remains a joke but a serious joke |: (: and a big indication of the mechanisms involved must be given.

Some solvers no doubt would thrive with just a (: like some people enjoy cipher-cracking. But my guess is that the majority of solvers would appreciate more structure. Another curse is that one can only have 1 stipulation.

Do these not become fairy? Somehow, even though the tricks are being tamed, I am reluctant to disavow my debt to the trickster god of jokes. Therefore the stipulation must be of a form in which it cannot be misinterpreted as a fairy convention. I don't like the words "hint" or "clue" - they are too patronizing. I would prefer a word like "assertion".

For example, Rocky, how would you want to stipulate your own dummy pawn problems? Assuming that they weren't being viewed in a Mat Plus thread about dummy pawns, but one was just sitting in a magazine with only the diagram and the stipulation? What do you want every solver coming fresh to the problem to see?

I guarantee you that he does. This all depends on the personal and cultural background of the individuals involved. Recently I was faced with a joke problem requiring the super long castling move. This was a closed joke chapter to me, really, really old school. It's quite likely that I would never come up with the solution even though I might have explored every other innovative angle imaginable. I'd probably had a better chance never having heard of the joke castling in the first place! 

Every subject has its joke shell because it is the result of reflective thought. It's one of the things that makes humans human.

@anselan Okay, that's fine by me re the tagging of most of these promotion-to-P problems as Golden Age. So I just looked up your documentation for these keywords on the PDB, and they're a good read too. Your mention of the fairy piece "dummy" as being distinct from "dummy pawn", though, highlights a small issue of terminology that I disagree with. I think "dummy pawn" as a fairy unit with no moving power – unrelated to promotion – is a well-established term (see Dickins' Guide to Fairy Chess). Even I have composed a couple of problems using them a long time ago, though surprisingly I have trouble finding examples of these fairy units on the PDB. Since the idea of promoting-to-P is a special rule rather than a special unit, it may be better to call this rule something other than "Dummy pawn", to keep the ideas distinct. (Indeed you might have mentioned that an unpromoted P on the 8th rank, if shifted due to another fairy rule, would move as an orthodox P.) Just a thought!

When I republished my promotion-to-P problems in my blog, I gave them the "Legal P=P" stipulation, which kind of turned them into fairy problems, besides giving away the solution. If they were to be published for proper solving, I suppose I'd just label them as joke problems. :-) Notwithstanding your "joke(not!)" tag!

Hi Rocky,

Wikipedia & Popeye talk of Dummy as the inert fairy unit, and Henrik Juel argued for this, and I think he's right. This object has no Pawn characteristics, so should not be called Pawn. I am surprised if Dickins said something else. Hope you can agree?

The Dummy Pawn is the unpromoted Pawn on the 8th rank, and absolutely if it gets shifted somehow e.g. captured in Circe then it can just keep moving forwards. I will make that clear in the documentation.

Both these units are supported by Popeye. You can have promonly Q R B S P DU. The pawn promotion is interpreted as just e.g. c2-c1 rather than c2-c1=P. This brings up a third possibility which is to actually have a *promotion* to P, which is what you mentioned in your stipulation. Here as you say it shifts from joke to fairy.

Maybe there are 3 levels of joke difficulty:

1) no clue given

2) (-: in stipulation, but no clue as to kind of joke

3) (-: would not be enough - composer needs to give more. Maybe it becomes fairy? Hard not to give the game away. But the focus must shift somehow to the chess at this point: figuring out which joke theme it is would be distracting.

You said you might shift from 3 to 2. But I think that's because you are familiar with the problem now. I think it would be hard for any solver to crack this with just a (-: to direct them.

@anselan It may be a bit more logical to use "dummy" for the fairy unit, but I'm just saying there's already a tradition of using "dummy pawn". Not only Dickins but when my two examples appeared in The Problemist and Phenix, their editors called them dummy pawns/powerless pawns. 

I didn't realise that you can use the PromOnly condition in Popeye to allow promotions to Ps! Naturally "PromOnly" is stored as a fairy condition (= rule) in Popeye (not as a fairy piece, unlike Dummy), so this distinction is backing up what I'm saying! But really, it's just names and not a big deal to me.

It's cool that you can test promotion-to-P problems in Popeye, and here's a screenshot for the Loyd mate-in-3, for the position after the key.

I don't get what you mean by "a third possibility" of promotion to a P. Although Popeye writes the solution in a funny way and leaves out the P, such as 1.c7xd8 in the Loyd, surely there's now a WP on d8 (as shown in the diagram) and it would be the same as 1.c7xd8=P, isn't it?

On the naming. Dickins' idea of "Dummy Pawn" for the inert fairy unit is confusing in several ways, and even if some people are used to it and are no longer misled, there is clearly another, better, tradition which is to call it simply "Dummy": PDB, Wikipedia, Popeye, Henrik Juel all follow this. Who am I to go against them when I agree with them. Traditions are fine and all, but if there are two traditions which conflict, as here, then let us bury the inferior one immediately. Please.

Clearly under the "Dummy Pawn rule" what you end up with on the 8th rank is a pawn, which happens not to be able to move for the moment. While it occupies this state, we call it a Dummy Pawn, but the pawn itself is not a fairy piece. It remains a pawn: c2-c1. Popeye promonly to P does this. As you know, I would say this is not a fairy rule or even a joke rule: it's just the historical rule. I think the fact that Popeye classifies this as (fairy) promonly is not significant - the fact that it appears in the score as c2-c1 is what matters.

It is also possible to *promote* to a dummy c2-c1=DU. That's just a standard case of promoting to a fairy piece. If there's another dummy on the diagram, this is fairy-legal, in the usual way.

So we have c2-c1 & c2-c1=DU. The third corner case is that one might in theory want to have c2-c1=P. But I can't think of a fairy condition where this would differ from c2-c1, so let's stick with that.

I can see you both have some history on this material. Let me suggest a simple criterion for separating fairy rules from jokes (and from orthodox rules).

A rule is only a joke rule when you cannot identify it without knowing the goal of a problem and knowing that the goal is achievable. The redundancy of information resulting from it can be used to find a matching joke rule. For instance, you can find the joke rule of an unjustified e.p. move by insisting that a checkmate in 2 moves must exist - even though it is impossible by the given orthodox rule set.

A fairy rule is no different from an orthodox rule in that one has a priori knowledge of it but it's a bit different from what we do most of the time.

IMO, this resolves 99% of all discussions about separating these categories. Of course, inside the joke domain there are still flavors about the set of available joke rules and whether or not these rules are strictly defined at all. But they will still fall within the goal-driven description.

@anselan So Popeye writes 1.c7xd8 to indicate a "normal" P move that doesn't change the P to anything, hence it left out the "=P" part. I see, it wasn't being funny but super logical!

@Arisktotle The simple difference between fairies and jokes is that fairy rules are disclosed to the solver while joke rules are not. A typical fairy rule, like Circe, can be turned into a joke if you don't mention it's in effect; a typical joke rule, like promotion-to-kings, can be turned into a fairy if you specify it as a condition. Which unorthodox rules work better as fairies and which work better as jokes is a separate issue.

Yes, I agree but often there appears to be set of known joke rules (in a certain context) from which to select the applicable one(s). Such is only possible when you relate them to the stipulated goal. So it's a bit of both.

Really unknown joke rules have no future. For instance, who will ever guess my favorite but private joke rule "do_not_visit_a7_d4_and_g6"?

There are a small number of original chess jokes, most of them endlessly recycled and recast. It reminds me of an old story my father used to tell, which has the distinction of having multiple alternative punchlines. Pick whichever you like. Suggest an alternative, do.

There was a new convict in a prison, it seems. After lights out, he was surprised that the other convicts in their beds started calling out numbers: e.g. "33!" followed by massive laughter, then "26!" with even greater merriment, and so on.

The new convict whispered to his neighbour: "What's happening?"

"Ah well, we've all heard all our jokes so many times, that now to save effort we've just given each joke a number, and someone calls that out."

At that point, a convict says: "86!" and there was huge, uproarious, gales of laughter. Grown men wiping tears of hilarity from their eyes, etc.

"What was so funny about that one?" asked the newbie.

Alternative punchlines:
(a) "Well, it's the way he told it."
(b) "We hadn't heard that one before."
(c) "Umm... actually I didn't get that one. I just laughed because everyone else did."
(d) "That's the one that begins 'There was a new convict in a prison'."

I like Ari's definition of jokes, as things where the solver has to guess the rules, based on assumptions like the problem must be legal. But what do you do for a joke with difficult chess in it? Even a (-: is not enough. Must it become a fairy? Nothing against fairies, but it seems sad somehow to lose the wackiness

You finally found me out. 86 = "do_not_visit_a7_d4_and_g6". By admitting this I violated probation and they will lock me up again. When I think of it, that can't be worse than where I'm now

Jokes are like the atmosphere of a planet. It's always a thin layer and it's always nearby or it wouldn't be seen as belonging to the planet. In this metaphor, fairy types are different planets belonging to the same solar system though some are far away from its central sun.

I agree with Rocky that the fundamental difference between joke and fairy is disclosure. Any rule becomes a joke when applied to a system (planet in my metaphor) where it is not a disclosed part (atmosphere). For instance, not having dead positions in a retrograde problem would be a joke rule though not a very funny one. "Joke" is not an attribute of a rule but of the {system,rule} relationship.