LMI July Puzzle Contest (30-31 July 2011)

[PuzzleScot]

Logic Masters India announces July 2011 Puzzle Test (#2) — Magic Cube
Authors : puzzle-makers of the 20th WPC
Dates : 30th and 31st July
Length : 120 minutes
IB and Submission Link : here

This will be the fourth and final contest in the set of 4 LMI monthly Puzzle contests (April - July 2011) that will determine 2 of the places in the UK team for the 2011 World Puzzle Championship.

Therefore, if you are seriously considering competing for a place in the UK team this year, it is strongly recommended that you download the puzzle instructions when they become available, and ensure you are registered on the LMI site. This is especially important if you have already missed an eligible puzzle contest. The latest UK Puzzle ratings table is here.

You should ensure that you are familiar with the puzzles being used before the contest starts, as it will be a race against the clock!
Remember you can choose the time to compete within this window that most suits yourself.

An overview of UK WPC/WSC team qualification is available here, with full details here

(Note: Although this is the 2nd LMI Puzzle contest this month, the first did not count towards 2011 UK ratings as it was composed by a UK contender)

[drsteve]

Having had a go at the sample puzzles, I would strongly recommend others do the same - some of the samples aren't exactly easy...

[PuzzleScot]

Having had a go at the sample puzzles, I would strongly recommend others do the same - some of the samples aren't exactly easy...

Agreed. I haven't been able to do the Heyawake, Fillomino or Star Battle one yet...
The 'Special Suduku' example is particularly interesting as it reveals a short-cut for this style (That's for you to work out )

[GarethMoore]

I'm glad it's not just me! The ones you mention seem ridiculously hard, and since the instruction booklet says they are typical of the standard in the competition itself and of the forthcoming WPC I didn't find that very reassuring. If this will truly be the standard of puzzles at the WPC this year (as the instruction book claims) then it puts me off wanting to go, at least just a little bit.

The Heyawake is in my view the worst offender in the instruction book since it is essentially luck - there is very little you can place via small logical steps, and after that there are multiple guesses needed to see it to completion no matter where you start (I spent about an hour trying every possible start point since I was sure I was missing something. Now I'm sure I wasn't). On the other hand if you guess quickly and carefully and erase quickly you can solve it in a few minutes, but where's the satisfaction in that?

The Fillomino also requires far too much guesswork. It can be massively simplified by careful counting to start with but unfortunately guesses are still required and there are several ways to fill the entire puzzle but for a single contradiction, so guesswork is potentially very slow if you're unlucky. There is no way I can see to solve the whole puzzle other than using some trial and error. Perhaps I missed something but in any case I think it's fair to say that sufficiently obscure logic in competitions, even if it exists, is often resolved via guesswork by fast solvers so I think this puzzle fails.

The Star Battle you can solve a chunk of the lower-right and work out that two of the stars from the bottom-left area must end up in the centre, but past that I don't see how to proceed logically. On the other hand I always suck at Star Battle so I assumed I was missing something.

I also didn't enjoy the Capsules or Dutch Loop. I can't solve the Capsules without guessing, and if I'm missing something it must be a bit of if-then logic I think since I stared at it for ages. I can only place 14 digits. For me at least I would be far faster solving this via guesswork than logic. Dutch Loop I found I needed one single guess after which the whole thing fell immediately, but maybe I missed a step.

I could do the rest without guessing, although I discovered I was very slow at the Sudoku one. What's this trick then? [edit: I'm pretty sure I can see what you mean]

[motris]

I've been developing new approaches to getting some of these solved without guesswork. I can say there is a very satisfying global constraint on the Star Battle which really helps its solution. It may be the puzzle where the cube shape mattered most and introduced new logic to discover. Not sure on all the others talked about up-thread. I will say that the Easy/Medium/Hard approach the Hungarian WPC authors have used on their puzzles has worked rather well in the past two WPCs where I've seen the puzzles. I'm hoping these are very representative of the top end of the spectrum (or also that they are simply less elegant puzzles than those eventually chosen for the competition, which is often the case).

[drsteve]

I've been developing new approaches to getting some of these solved without guesswork. I can say there is a very satisfying global constraint on the Star Battle which really helps its solution. It may be the puzzle where the cube shape mattered most and introduced new logic to discover. Not sure on all the others talked about up-thread. I will say that the Easy/Medium/Hard approach the Hungarian WPC authors have used on their puzzles has worked rather well in the past two WPCs where I've seen the puzzles. I'm hoping these are very representative of the top end of the spectrum (or also that they are simply less elegant puzzles than those eventually chosen for the competition, which is often the case).

If it's the same piece of logic that I'm thinking of - probably not, as mine's pretty simple - then it can also help with the Magic Snail - not that the example for that needed much help though...

[drsteve]

Having had a go at the sample puzzles, I would strongly recommend others do the same - some of the samples aren't exactly easy...

Agreed. I haven't been able to do the Heyawake, Fillomino or Star Battle one yet...
The 'Special Suduku' example is particularly interesting as it reveals a short-cut for this style (That's for you to work out )

I wonder - is the Sudoku solvable without assuming/proving that the shortcut works?

[sknight]

Well, I solved the Sudoku in ways that didn't feel like I was using any special gimmicks, so without knowing specifically what y'all are talking about, then yes I think it is.

One thing I will also point out about the sudoku -- this is the booklet's example of an exploded version of the cube diagram. There are only three "horizontal" rows of 8, but they definitely count as part of the restriction, and are used in the puzzle. Since there will be other puzzles with this layout in the booklet, this is probably a particularly important example to look at even if you hate sudoku and are planning on spending your time on other types.

And yeah, the Heyawake example was a pain. Definitely took some trial and error for me.

[motris]

If it's the same piece of logic that I'm thinking of - probably not, as mine's pretty simple - then it can also help with the Magic Snail - not that the example for that needed much help though...

It's quite likely the same bit of logic, but it is real overkill for the Magic Snail example. Still, that logic should convince you that you can only get a 2N-length cube for a magic snail puzzle or for a Star Battle if it's the same thing.

Regarding "exploded" cubes, I've been playing with this kind of design for a long time, so pull out Mutant Sudoku to see other options of what they will possibly look like. The wording points to it just being the sudoku form (which is from the "less sane" part of our progression there) but it is possible some of the many other possible grids with smaller stacked cubes could appear, as in the "more sane" part of that chapter. I'd naturally think a puzzle like Star Battle or certainly the Magic Snail needs a different design than that particular exploded cube.

[PuzzleScot]

I'm interested in finding good fast notation. But I'm worried that greying in cells is time-consuming.

For what it's worth, I use O/X notation most of the time.
I shade in more fully (out of the corner of my eye) if I'm stuck, whilst looking for other logic, and to clarify the position.

When it's getting tricky, and almost always for 'eminent domain', I draw lines through connected cells, and terminate completed lines with a dash.

It's easy to assume everyone uses the same notation, but this is clearly far from being the case. Interesting to see what other people do too...

[drsteve]

If it's the same piece of logic that I'm thinking of - probably not, as mine's pretty simple - then it can also help with the Magic Snail - not that the example for that needed much help though...

It's quite likely the same bit of logic, but it is real overkill for the Magic Snail example. Still, that logic should convince you that you can only get a 2N-length cube for a magic snail puzzle or for a Star Battle if it's the same thing.

Regarding "exploded" cubes, I've been playing with this kind of design for a long time, so pull out Mutant Sudoku to see other options of what they will possibly look like. The wording points to it just being the sudoku form (which is from the "less sane" part of our progression there) but it is possible some of the many other possible grids with smaller stacked cubes could appear, as in the "more sane" part of that chapter. I'd naturally think a puzzle like Star Battle or certainly the Magic Snail needs a different design than that particular exploded cube.

I was presuming that the majority of the puzzles would be of the stack or "cube with bits pushed inwards" style, if you know what I mean. I suppose it's feasible that Capsules might work on a weird layout (like the Sudoku example), as might Fillomino or Crossword.

[drsteve]

I'm interested in finding good fast notation. But I'm worried that greying in cells is time-consuming.

For what it's worth, I use O/X notation most of the time.
I shade in more fully (out of the corner of my eye) if I'm stuck, whilst looking for other logic, and to clarify the position.

When it's getting tricky, and almost always for 'eminent domain', I draw lines through connected cells, and terminate completed lines with a dash.

It's easy to assume everyone uses the same notation, but this is clearly far from being the case. Interesting to see what other people do too...

When I say shading, it's not a neat process - just a scribble more than anything else. Actually quicker if you need to mark a number of cells in a row.
I tend to use O and X for trial and error if I get stuck.

[Valezius]

I'm glad it's not just me! The ones you mention seem ridiculously hard, and since the instruction booklet says they are typical of the standard in the competition itself and of the forthcoming WPC I didn't find that very reassuring. If this will truly be the standard of puzzles at the WPC this year (as the instruction book claims) then it puts me off wanting to go, at least just a little bit.

Hi!

No. The samples are medium and hard puzzles. Sometimes too hard. I chose this option because many puzzlers didnt solve too many cube form puzzle. And I could provide other practice puzzles.

Don't forget, there are 33 puzzles and 120minutes. The average difficulty cannot be too high.
The first puzzle from every type is really easy, I promise . And more easier than the sample. In many cases the medium puzzle is easier than the sample. And you can see from the points in most cases the 3rd puzzle has same difficulty than the 2nd.

I hope this wont be a terrible test And I could calm you down

Zoltan

[drsteve]

Can I just say that I really appreciate a challenging instruction booklet, given mostly new puzzle forms, and also having the solutions at the back, rather than next to the puzzle?

Much appreciated and really looking forward to the contest.

[GarethMoore]

I hope this wont be a terrible test And I could calm you down

You have - thanks! As instruction books go as Steve points out it is great to have real puzzles you can practice on. I probably took the introduction a little too literally in terms of what the puzzles represent, so thank you for taking the time to clarify.

[david mcneill]

I'm confused by some of the discussion here. In the Sudoku puzzle, all I can see is that the oddments make up a 1-8 set. In the Star Battle, all I can see is that there must be an equal number of stars on each face. I don't see how this leads to a 2N constraint on the length of the cube. Is N the number of stars in each row/column/region? Or does 2N simply mean that the cube length must be an even number? Either way, I don't get it.

Please forgive my lethargy. But I would be grateful for a little prod in the right direction.

[drsteve]

I'm confused by some of the discussion here. In the Sudoku puzzle, all I can see is that the oddments make up a 1-8 set. In the Star Battle, all I can see is that there must be an equal number of stars on each face. I don't see how this leads to a 2N constraint on the length of the cube. Is N the number of stars in each row/column/region? Or does 2N simply mean that the cube length must be an even number? Either way, I don't get it.

Please forgive my lethargy. But I would be grateful for a little prod in the right direction.

That's all I came up with, but was discussing them obliquely due to the fact that we are using this as a qualifier via the league for the WPC. And I don't get the 2N bit either.

[drsteve]

I'm confused by some of the discussion here. In the Sudoku puzzle, all I can see is that the oddments make up a 1-8 set. In the Star Battle, all I can see is that there must be an equal number of stars on each face. I don't see how this leads to a 2N constraint on the length of the cube. Is N the number of stars in each row/column/region? Or does 2N simply mean that the cube length must be an even number? Either way, I don't get it.

Please forgive my lethargy. But I would be grateful for a little prod in the right direction.

That's all I came up with, but was discussing them obliquely due to the fact that we are using this as a qualifier via the league for the WPC. And I don't get the 2N bit either.

Scratch that, yes I do.

For a cube of side length n, there are 3n stars. For those stars to be paired up, n has to be even, hence 2N.

[david mcneill]

Thanks Steve. I'm getting it now. No matter what the number of stars is in each row/column/region, the cube side length must be even.

[motris]

I was trying to be oblique too so apologies if I confused you initially, but it looks like you've caught up to the major constraint(s) placed on the construction.

[emma]

Can I just say that I really appreciate a challenging instruction booklet, given mostly new puzzle forms, and also having the solutions at the back, rather than next to the puzzle?

Much appreciated and really looking forward to the contest.

I agree. Also glad to hear that these are not the easy ones: I tried timing myself on each, and it wasn't looking like I'd get much done.

[detuned]

Tom Collyer in not making an answer entry error shock!

This was quite a fun set - sometimes with "mutant" puzzle types I get the feeling that the twists are a little too gratuitous, and a lot of the value in the puzzles comes from how well you were prepared for these twists. With this one, the twist being the same for all the puzzles managed to pull the set together as a whole in quite a satisfying way. it was nice to see a couple of "word" puzzles in the set too.

Anyhow, that's probably vague enough a comment whilst the test is still going...

[drsteve]

All done from me - wasted too much time on one particular puzzle that should not have caused me a problem.

More after the competition ends

[nickdeller]

Of course, it *would* be the highest value puzzle I solved that had my only answer key FAIL in it.

Which in turn serves me right for solving it in a way that was almost guaranteed to throw me off-stride when extracting the answer key! Nice set, though!

[dickoon]

Enjoyed the puzzles, which scored quite well on the "I'll never solve this; oh, wait, I have amazed myself by actually solving it" front. That said, I don't really know how to solve the "Cave" puzzles and could only barely get started with three of the other types.

I thoroughly approve of having a variety of difficulties within each puzzle style, to provide some relatively accessible starts for us more modest solvers, noting that several other tests have done similar things in the past.