The "b" and "6" in my original FEN mean black to move and make his 6th move, i.e. PG 5.5. So yes, I intended the impostor knight in 5.0; no, I missed the cook. Back to thinking.
A proof game
Sort:
JoachimJo wrote:
Hmm actually without the knights, in 9 moves, it could have been a very good SPG with two distincts solutions (which is rare but autorized).
The 1st one being the obvious one and the second one being the beautiful :
1. Nf3 a5 2. Ne5 Ra6 3. Nc4 Rh6 4. Nxa5 Nc6 5. Nxc6 Rxh2 6. Na7 Rxh1 7. Nxc8
Rxf1+ 8. Kxf1 Qxc8 9. Kg1 Kd8
Unfortunatly there is a dual in the second one as the N can go to a5 via d4 and b3...
So it's worthless, but there might be an idea.
Congrats on having time to waste on positions I spent mere seconds thinking about [literally] just before posting. Needless to say, I don't take your particular form of composing very seriously. However, for the sake of completeness (after this I am OUTTA HERE), here are two SPGs that I've revised to meet your [ridiculous] standards. Don't bother trying to break them, btw - this time I've actually analyzed them myself (something I'm rather good at doing when I feel like it), and they're both airtight.
Remellion wrote:
As an amateur solver, I do like unique solutions (unless the duals add value, like an allumwandlung SPG). It's not so much to add/reduce the difficulty of the problem so much as to have a sense of tidyness about things, like a mathematician cleaning up a proof. And once you try your hand at making unique PGs you realise how magical it is that we have so many of them today, with all kinds of themes to boot. And once you succeed in making one, well, that feeling is gooood - I've made a couple cute ones, but they can be cute in an interesting way.
I "like" unique solutions, too. My question was more like, "From a puzzle-solver's perspective, do tidyness and technicalities automatically override ALL other factors combined, such as difficulty, entertainment value, and soundness of content? And does this apply even when the entire list/set of moves is provably unique?"
The SPGs in the above post are too trivial to solve to even be considered legitimate puzzles, in my book. I freely admit to this in spite of the fact that I myself composed them (fortunately they weren't especially hard to come up with 
). The same is not true of the puzzle from post #46, though, and that one needed a lot more work to get right.
cobra91 wrote:
Remellion wrote:
As an amateur solver, I do like unique solutions (unless the duals add value, like an allumwandlung SPG). It's not so much to add/reduce the difficulty of the problem so much as to have a sense of tidyness about things, like a mathematician cleaning up a proof. And once you try your hand at making unique PGs you realise how magical it is that we have so many of them today, with all kinds of themes to boot. And once you succeed in making one, well, that feeling is gooood - I've made a couple cute ones, but they can be cute in an interesting way.
I "like" unique solutions, too. My question was more like, "From a puzzle-solver's perspective, do tidyness and technicalities automatically override ALL other factors combined, such as difficulty, entertainment value, and soundness of content? And does this apply even when the entire list/set of moves is provably unique?"
The SPGs in the above post are too trivial to solve to even be considered legitimate puzzles, in my book. I freely admit to this in spite of the fact that I myself composed them (fortunately they weren't especially hard to come up with ). The same is not true of the puzzle from post #46, though, and that one needed a lot more work to get right.
They don't. There are myriad PG examples showing that we can have it all.
cobra91 wrote:
However, for the sake of completeness (after this I am OUTTA HERE), here are two SPGs that I've revised to meet your [ridiculous] standards. Don't bother trying to break them, btw - this time I've actually analyzed them myself (something I'm rather good at doing when I feel like it), and they're both airtight.
Cobra, (if you're still here) the second SPG does have fairly mundane duals. The solution starts with moving the knights to g6 and g3 in 4 moves. But there are 2 routes for that, one starting with 1. Na3 and one with 1. Nc3.
JoachimJo reported there exists only one "home" SPG and I assume that was computer-tested. If so, then your position must be either identical to his, or it must be cooked.
@cobra91 : congrats, the firt SPG of your last post is sound !
For some reason you seem rather angry, but you should enjoy you just managed to compose one dual free SPG !
BTW, having a unique solution is not "my form of composition", it's just called "composition". In studies, in orthodox problems, in heterodox problems, in SPGs, or even in a multiple choice questionnary at school there must be only one solution.
When an idea is good but not good enough for a dual free SPG, the composer has to ask questions like "history of the captures ?r or "first move of wK ?" in order to have only one answer.
Yep ! Bravo Ariskotle !
When I discovered this SPG I was pretty sure it was the only at home with 16 pawns. I asked it as a composition problem on a French forum and some found this solution but no other so I was 99%+ confident in the unicity. Years later, the informatician and SPG fan François Labelle proved it is the only one.
It's a nice one! Knight interfering knight and forbidden king check in the tries. I count myself lucky finding this one relatively fast since SPGs with invisible moves and "free" captures are always amongst the most difficult, also for the composer!
I think I've both the Donati and the Pronkin.
I've normaly avoided the many cooks found by cobra, but there might more. As this thread has shown it's always easier to cook another's SPG that our own because we are biaised by knowing the intended solutionand the hope it's the only one.
I give the PGN for both of them, the Pronkin as a variant as it's not as beautiful.
That looks like a very promising scheme! And very solid when it comes to counting the black moves, so critical to the solution. Identifiable rooks and the magnificent disappearance of the ghost rook with the original rooks still on the board. The ultimate invisibility cloak!
However, there are some glitches in the initial moves which are hopefully repairable. Starting with 4. .. Rh5 5. Ba6 Ra5 etc. You sure have to work for your record!
oh ! I've missed it. It's so obvious. I'm sorry.
Arisktotle wrote:
cobra91 wrote:
However, for the sake of completeness (after this I am OUTTA HERE), here are two SPGs that I've revised to meet your [ridiculous] standards. Don't bother trying to break them, btw - this time I've actually analyzed them myself (something I'm rather good at doing when I feel like it), and they're both airtight.
Cobra, (if you're still here) the second SPG does have fairly mundane duals. The solution starts with moving the knights to g6 and g3 in 4 moves. But there are 2 routes for that, one starting with 1. Na3 and one with 1. Nc3.
JoachimJo reported there exists only one "home" SPG and I assume that was computer-tested. If so, then your position must be either identical to his, or it must be cooked.
I did abandon the thread, and had no intention of returning. However, two days later a realization suddenly hit me that a knight can reach the opposite-side corner in 5 moves even if it initially moves to the rim... which of course meant there would be minor duals in the at-home SPG I came up with. So, I returned to find you'd posted that right away, followed by the correct position.
It seems I'd spent too long analyzing too many positions too quickly (around 50-60 in ~90 minutes, which is tougher than it sounds, as cooks only become trivial to show once a position's "SPG number" has been proven), and was starting to see mirages. I observed [the illusion] that a knight's route to the opposite-side corner was unique, and then within minutes had the position from post #63 on the board in front of me. Despite my exhaustive analysis, it all just amounted to the equivalent of an airtight mathematical proof that still gets overturned because the axioms used were inconsistent (I assumed the initial idea which led me to the position was true, without any serious thought).
As for the composition challenge itself: obviously it turned out to be pretty difficult, but I honestly don't see how it had anything whatsoever to do with real composing (other than testing whether someone already happened to be familiar with the idea used in the only position that was entirely dual-free). Real composing is about one's skill and effectiveness at making subtle changes to a position until an idea works - yet such changes were all blocked off by the "16 pawns at home" restriction. To scrap 99.999...% of interesting positions with a "no transpositions" rule is bad enough, but to reduce the set of allowable problems to one just seems completely antithetical to what composers are supposed to care about.
Now, with all that being said, I guess my failure to correctly meet the demands of the challenge causes me to lose all credibility until I can produce a sound at-home SPG without duals. So here is another try, although (for obvious reasons) it has fewer than 16 pawns this time:
I'll present another composition of mine. This one took at least 5-10 hours over the past week to fully refine into something I was satisfied with. The intention is to show a deep idea which transcends the limits of what a "single move-order" SPG can contain, thus demonstrating how costly it really is to unnecessarily ban the vast majority of interesting problems.
cobra91 wrote:
As for the composition challenge itself: obviously it turned out to be pretty difficult, but I honestly don't see how it had anything whatsoever to do with real composing (other than testing whether someone already happened to be familiar with the idea used in the only position that was entirely dual-free). Real composing is about one's skill and effectiveness at making subtle changes to a position until an idea works - yet such changes were all blocked off by the "16 pawns at home" restriction. To scrap 99.999...% of interesting positions with a "no transpositions" rule is bad enough, but to reduce the set of allowable problems to one just seems completely antithetical to what composers are supposed to care about.
Well, nothing wrong with this SPG, though nothing spectacular either. No need to convince me since I am quite perceptive to your composing potential.
Note that Joachim had no idea whether there would exist 0, 1 or 10 different home SPGs when he started this journey. One of the most important composer talents is to pick the right objective! Not impossible, and not too easy since such wouldn't beat the competition. I'd say he picked the perfect bear for this hunt.
"Subtle changes" are quite a different thing for a solver than for a composer. Faced with a fixed position the solver is bound to pick his choices from a set of legal chess moves under rigid game length conditions. For a composer "subtle changes" take place in his mental SPG space and include dimensions like "swapping colors", "swapping king and queen side", "using a different shielding or blocking unit", "using a different cage entrance or exit", "changing the objective or game length" which have dramatic effects in terms of concrete chess move translations. This inventory is not much different in the home SPG from composing SPGs in general. Note for instance that I had to rotate the board 90 degrees when constructing my high road solution. Just a "subtle change" for a composer but quite a different outlook on the diagram!
No retro-problem-type is ruled out a priori, but they all start their lives in the "Smullyan" category. Off-beat and hopefully lots of fun but mostly unfit for serious competition. The reason is that there is nothing to compare them to. Only after composers have adjusted to creating this type en masse and solvers to solving them, will it be possible to get a feel for what is good or bad, hard or easy. It requires that the type is first liked by the retro-community. I for one have little affinity with the "free order" SPG with the "free format" stipulations. Though the occasional one may be lots of fun!
Arisktotle wrote:
Well, nothing wrong with this SPG, though nothing spectacular either.
Thank goodness... the curse has been broken! 
It wasn't supposed to be spectacular, needless to say. Here are two more - would like to know what people think of them:
The 1st one is just another "at home" position, but is a bit better than my previous one. The 2nd should be relatively difficult for a dual-free SPG. Both of the above problems were assembled rather quickly, so I'll just have to hope my analysis holds up.
Note that Joachim had no idea whether there would exist 0, 1 or 10 different home SPGs when he started this journey. One of the most important composer talents is to pick the right objective! Not impossible, and not too easy since such wouldn't beat the competition. I'd say he picked the perfect bear for this hunt.
Pfff... I'd be willing to bet he just stumbled upon the position, and was aware of how poorly such positions lend themselves to having completely dual-free SPGs. So it was reasonable to assume there could not be very many, and that it might even be the only one considering there are a mere 16384 (2^14) positions within the restricted set. The real achievement was proving that only one position could meet all assigned criteria... and that was done by someone of far greater intelligence than either myself or the OP.
"Subtle changes" are quite a different thing for a solver than for a composer. Faced with a fixed position the solver is bound to pick his choices from a set of legal chess moves under rigid game length conditions. For a composer "subtle changes" take place in his mental SPG space and include dimensions like "swapping colors", "swapping king and queen side", "using a different shielding or blocking unit", "using a different cage entrance or exit", "changing the objective or game length" which have dramatic effects in terms of concrete chess move translations. This inventory is not much different in the home SPG from composing SPGs in general.
Then how come all of the changes I wanted to make were resulting in positions that were no longer "at home"? 
I've developed many composing techniques that usually work for me, but none of those are useful if I'm not allowed to rearrange the pieces as needed. 'Twas a very self-contained assignment, if you ask me...
I for one have little affinity with the "free order" SPG with the "free format" stipulations. Though the occasional one may be lots of fun!
I'll resurrect the problem from post #46. Its diagram is below. Anyone following the thread should give it a try, as it's quite a fun one to solve, but not too easy. The puzzle in post #75 is a lot tougher - won't shock me if nobody solves it, but the solution is profound and extremely rewarding for those who manage to find it! 
cobra91 wrote:
Thank goodness... the curse has been broken! It wasn't supposed to be spectacular, needless to say. Here are two more - would like to know what people think of them:
The 1st one is just another "at home" position, but is a bit better than my previous one. The 2nd should be relatively difficult for a dual-free SPG. Both of the above problems were assembled rather quickly, so I'll just have to hope my analysis holds up.
The first SPG is indeed a more interesting home-SPG than your previous product, if only because it is asymmetrical and offers more solving options. A disadvantage is the mass of captures which leave the trail of an elephant.
The second SPG is a good one. I like the question surrounding the wQ identity and how it is to be solved. The capture on h6 is a nice touch!
Thank you very much for the review! 
I won't attempt to deny my relief that [it looks like] you were unable to break the solutions. In the 1st, the "mass of captures" was needed to stop me from feeling genuinely terrified of potential cooks I may have missed. The 2nd initially included 17. b4 f5, but I managed to bust that one (I leave it as an exercise for solvers to find the unfortunate glitch there ). That's why the queen is on a4, as opposed to the more aesthetic location on c4 - it's an "artifact" of the original version.
If no one gets around to solving my other two compositions in this thread within the next couple of days, I intend to post the solutions, so that it will be easier for people to comment on them. The one from post #75 is a monster, of course - the full analysis would be beyond the scope of these forums, so (if/when the time comes) I shall merely try to give a "flavor" for the depth of the coolest alternative tries.
Having been humiliated to the point of permanently giving up on all forms of composing, I now fully understand the fact that both of my above puzzles would have easily been solved, had even a single person cared enough to spend so much as 10 minutes thinking about either of them. So I won't insult anyone's intelligence by bothering with things like proof games or explanations; I'll merely give the answers, and hope for a comment or two on the problems' publication potential (my assumption is "none", of course, but perhaps with a major overhaul, something useful could be done with one of the traffic-related ideas).
Answer: g5, e2, e3, c3, b2, a2, h7, e6
Answer: 28...Ra8-a4
It's also possible to have a) SPG in N moves and b) PG in N+n moves.
For instance the most simple exemple I could think of.
a) SPG in 3.5
b) PG in 4