Seven Samurais - LMI June Puzzle Test

Information and discussion on Puzzle Competitions
Post Reply
debmohanty
Posts: 129
Joined: Fri 09 Jul, 2010 12:36 pm

Seven Samurais - LMI June Puzzle Test

Post by debmohanty » Wed 13 Jun, 2012 4:19 am

Seven Samurais - LMI June Puzzle Test
Author : Andrey Bogdanov

Dates : 16th - 18th June

IB and Submission Link : http://logicmastersindia.com/M201206P/

emma
Posts: 314
Joined: Wed 22 Jun, 2011 4:58 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by emma » Mon 18 Jun, 2012 10:56 am

Fun - thank you.

PuzzlerNickG
Posts: 64
Joined: Mon 21 Jun, 2010 6:22 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by PuzzlerNickG » Mon 18 Jun, 2012 1:42 pm

emma wrote:Fun - thank you.
More than fun, judging by your excellent score, Emma.

I enjoyed it too, but struggled on so many puzzles and couldn't get started at all on the Yin Yang, which I thought would be one of the easier ones for me.

emma
Posts: 314
Joined: Wed 22 Jun, 2011 4:58 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by emma » Mon 18 Jun, 2012 10:02 pm

Yin Yang is something I'm usually bad at, so I didn't waste any time on it.

I expected to do better with Fences - I kept coming back to it, but with no success. I'd have done better to have done the rest of the Minesweeper one, but still not well enough to catch Neil and James!

PuzzleScot
Site Admin
Posts: 2745
Joined: Fri 18 Jun, 2010 10:45 pm
Location: Edinburgh, Scotland

Re: Seven Samurais - LMI June Puzzle Test

Post by PuzzleScot » Tue 19 Jun, 2012 9:07 am

Yes, well done Emma - great score!

One of the more infuriating contests I've done - to 7/8 solve a puzzle, but still not be able to deduce an answer key meant there were really just 7 puzzles (not 28) with some bonuses for partial solves. Puzzle choice was critical, as was getting the balance between persist and abandon! I abandoned most, although completed 2 more sets later with no real issues (Star Battle and Hitori), but in the contest, wasted lots of time on Easy As and Fences with no progress. I was pleased to get one set out (Yin Yang) without any typing errors.

How ANYONE managed to correctly solve all 28 in the time is staggering.

Nilz
Posts: 270
Joined: Fri 02 Sep, 2011 5:21 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by Nilz » Tue 19 Jun, 2012 8:42 pm

Yes, really good performance Emma- solving Easy as ABCD and Starbattle within 45 minutes is very impressive, I gave up on the former, and struggled with the latter as well- I'm not sure I deserved 42 points more than you really.

emma
Posts: 314
Joined: Wed 22 Jun, 2011 4:58 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by emma » Tue 19 Jun, 2012 9:42 pm

Him at home wrote an "Easy as" generator many years ago, so I got quite good at those. I don't know if we still have it - it might have been in RiscPC days.

Now, can someone point me at any guidance on Yin Yang, or a source of lots of examples, because I still don't have a clue how to start. It's not the easiest puzzle to google.

kiwijam
Posts: 700
Joined: Wed 01 Jun, 2011 10:03 am

Re: Seven Samurais - LMI June Puzzle Test

Post by kiwijam » Tue 19 Jun, 2012 11:49 pm

Agree with Alan's persistence comment. They were all tough to get started, but after a certain point the multiple grids helped each other a little more. Flitting from one to another looking for easier points didn't work this time.
I was annoyed that I made one small error in one of the Battleship grids, which then caused three of the grids to be 'wrong', although only one looked wrong, the other two just had multiple correct solutions (if they had been non-Samurai). So I spent 10 minutes looking for the error in the wrong place, it turned out to be in one of the correct-looking grids.
Wish I'd spent some time beforehand thinking about them, for example the Fences samurai has rules stopping the lines going in and out of the overlap areas except at 2 of the corners, which makes them a lot faster once you know it.


Regarding Yin Yang, there are two basic deductions which work wonders. You may already know them, but I'll write them down to make sure!
1) You can never have a 2x2 square with different diagonals, e.g.:
BW
WB
If that occurs then one of the colours would have to be split into two pieces. Along with the normal 2x2 rule, this means you can find the 4th cell if you have three cells in any configuration like:
BW WB BB WW
W_ B_ B_ W_

2) Around the edge of a grid you can only have one length of White and one length of Black. So the edge is always the best place to start.
In these samurais you can't cut off the corner of the next grid, so the border lengths must always switch in the two overlap sections. This means there is one colour that goes around the outside of all 4 grids, and the other colour goes around the inside of the central 'gap'.

(I'm also keen to practise more 'Yin Yang' examples if anyone has a link?)

PuzzleScot
Site Admin
Posts: 2745
Joined: Fri 18 Jun, 2010 10:45 pm
Location: Edinburgh, Scotland

Re: Seven Samurais - LMI June Puzzle Test

Post by PuzzleScot » Wed 20 Jun, 2012 12:47 am

kiwijam wrote:(I'm also keen to practise more 'Yin Yang' examples if anyone has a link?)
To get you started:
Puzzle 3 here, on Cihan Altay's PQRST, was the first time this puzzle ever appeared: http://www.otuzoyun.com/pqrst/pqrst13.html

Rohan has some links too, including a mention of the Diogen Yin-Yang variation contest.

Nilz
Posts: 270
Joined: Fri 02 Sep, 2011 5:21 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by Nilz » Wed 20 Jun, 2012 7:02 pm

Can't help much on the Yin-Yang talk, I made a small guess to get started and lots of things flowed from there. Shame I only had about 3 minutes left when I started it. James's second point is quite important though, as it can help make a lot of quick deductions. This might not be new to anyway, but I'll try a verbal explanation anyway. Say the grid size is 5x5, 1-5 along the bottom a-e up the side. If 1a is white, 2a is black and 2d is white then: 2e cannot be white, since then 1d and 1e would also have to be white (by James' rule). So 2e is black, and hence lots more boundaries squares can be filled in in black (and clearly then, 4b and 4d must be white, etc.). Hope that helps a little.
PuzzleScot wrote:
kiwijam wrote:(I'm also keen to practise more 'Yin Yang' examples if anyone has a link?)
To get you started:
Puzzle 3 here, on Cihan Altay's PQRST, was the first time this puzzle ever appeared: http://www.otuzoyun.com/pqrst/pqrst13.html

Rohan has some links too, including a mention of the Diogen Yin-Yang variation contest.
Woah PQRST that's a blast from the past. First online comp I ever did: http://www.otuzoyun.com/pqrst/pqrst11.html. Alan beat me. :cry: And Ulrich Voigt was *only* 24th. Makes fascinating reading though, now that I recognise so many more of the names. What happened to some of those people- the top UK participant was Colin Backhurst, where is he now?

PuzzleScot
Site Admin
Posts: 2745
Joined: Fri 18 Jun, 2010 10:45 pm
Location: Edinburgh, Scotland

Re: Seven Samurais - LMI June Puzzle Test

Post by PuzzleScot » Wed 20 Jun, 2012 7:32 pm

Nilz wrote:Woah PQRST that's a blast from the past. First online comp I ever did: http://www.otuzoyun.com/pqrst/pqrst11.html. Alan beat me. :cry: And Ulrich Voigt was *only* 24th. Makes fascinating reading though, now that I recognise so many more of the names. What happened to some of those people- the top UK participant was Colin Backhurst, where is he now?
Lots of things come to mind here:
1) Wow - you knew of puzzles back then? You must have been 12 or 13? :roll:
2) Harumph! I'm not as dumb as current contests make me out to be - I'm just significantly slower than most! I was top Btrit in pqrst05 and pqrst10. Pretty sure I aced #15 too, but we may never know! Things went pear-shaped for me when I learned of the WPC and the speed element came into play.
3) Indeed! According to my records, 52 Brits (to 2010) had participated in the USPC, and Colin wasn't one of them. Brett Gilbert is another untested name. I'm acquainted with his sister (who contributed maze puzzles to this and last year's UKPC's), so should chase that up.
[edit] 4) There was a contestant back then named "Debadatta Mohanty"!

Nilz
Posts: 270
Joined: Fri 02 Sep, 2011 5:21 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by Nilz » Wed 20 Jun, 2012 8:08 pm

PuzzleScot wrote:
Nilz wrote:Woah PQRST that's a blast from the past. First online comp I ever did: http://www.otuzoyun.com/pqrst/pqrst11.html. Alan beat me. :cry: And Ulrich Voigt was *only* 24th. Makes fascinating reading though, now that I recognise so many more of the names. What happened to some of those people- the top UK participant was Colin Backhurst, where is he now?
Lots of things come to mind here:
1) Wow - you knew of puzzles back then? You must have been 12 or 13? :roll:
2) Harumph! I'm not as dumb as current contests make me out to be - I'm just significantly slower than most! I was top Btrit in pqrst05 and pqrst10. Pretty sure I aced #15 too, but we may never know! Things went pear-shaped for me when I learned of the WPC and the speed element came into play.
3) Indeed! According to my records, 52 Brits (to 2010) had participated in the USPC, and Colin wasn't one of them. Brett Gilbert is another untested name. I'm acquainted with his sister (who contributed maze puzzles to this and last year's UKPC's), so should chase that up.
[edit] 4) There was a contestant back then named "Debadatta Mohanty"!
I know you're a good solver Alan, I was kidding. ;) From the date on the website, I would've been 16 when that took place. Solving puzzles was more fun than studying for A-levels or whatever I was supposed to be doing at that time. :) IIRC, I did try to take part again, I just never really had enough free time during the right weeks. And then I guess after those competitions stopped I fell out of the online puzzling loop for about 6 years! The perils of trying to do a degree! ;) Having said that, it seems that lots of other UK solvers have fallen out of the loop for even longer, which is a shame, since there are quite a lot of UK solvers on those results.

detuned
Posts: 1907
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: Seven Samurais - LMI June Puzzle Test

Post by detuned » Wed 20 Jun, 2012 9:28 pm

That link is fascinating... thanks for sharing!

I took the test during a brief window of free time last week, and bombed quite spectacularly. I really couldn't get my head around the fences and the samurai interactions. I'm fairly sure I lucked out with the hitori and not really understanding the rules what with the lack of simply connectedness and all and many misinterpretations.

I did the black and white later, after getting the initial deductions James mentioned. It turned out one of the outer conditions got you a contradiction in the west-positioned grid fairly quickly, and after that things weren't so bad. I found myself using what in sudoku terminology would be called "painting" - if square X has value A then square Y has value C, but if square X has value B then square Y also has value C - but other than that the puzzle seemed like steady progress...

puzzlemad
Posts: 878
Joined: Mon 21 Jun, 2010 10:20 pm
Location: Oldham

Re: Seven Samurais - LMI June Puzzle Test

Post by puzzlemad » Thu 21 Jun, 2012 7:09 am

Just had a very quick glance through this lot - will find time to read properly later on.

Didn't get much chance for preparation (but that's no excuse).

I started off with the fences - didn't get too far with it before eventually abandoning, managed the ABCD and the Hitori - found those fairly straightforward. Quick glance at the star battle, but I'm normally not much good on those - today's croco demonstrates that!, so left it pretty quickly (solved after and found it straightforward so should have persevered). Eventually went back to the fences and managed the top one (with a guess in the bottom right corner as I'd not got anywhere with the east one) - answer submission was wrong and I just presumed that it was the corner - had I checked I could have claimed some points there as I'd miscounted a simple line. Got going with the minesweeper as the time was ticking away and got three of the grids out (apart from one corner), so put in an answer for that. After time was up I looked at it again and it took me about another 30-60 seconds to solve the last one (why didn't I try whilst the last bit of extra time was ticking away).

Sitting around in hospitals waiting for appointments whilst the medical profession make differing conclusions about my damaged finger (yes I fell of a horse!) has allowed me to spend more time on the remaining ones so have now solved all - apart from the YingYang - but a report for work is far more important. Back to hospital on Friday morning - so may get to look at the YingYang then!

(Quick puzzle for you - how can different health professionals look at the same X-ray and come up with different answers - 2 say no break (1 of which says boutonnierre injury) and another says broke in 2 places!) - treatments of course totally opposite to each other.

david mcneill
Posts: 424
Joined: Mon 21 Jun, 2010 2:02 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by david mcneill » Thu 21 Jun, 2012 12:05 pm

I took this test the other night and solved 20 of the puzzles in 100 minutes, including writing down the answer keys. Found it a very interesting set of puzzles and it has stimulated a very interesting thread here.

I skipped the Battleships puzzle and started with the Hitori, in which pretty quick logical progress could be made on all 4 sub-grids. The Starbattle I found quite tricky and after making a few deductions, I decided to move on. I thought the Easy as ABCD was a beautiful puzzle and was pleased to get it completed in decent time and entirely by logic. Then I went on to the Minesweeper. This was extremely fiddly. The overlapping areas helped a bit, but the total number of mines in each sub-grid had to be checked repeatedly. Although I solved the puzzle, I spent a long time on it. Next up was Yin Yang. Not sure where to start, so back to Starbattle. One or two further deductions and suddenly the whole thing fell. Was able to complete all 4 sub-grids very quickly at this point. Another beautiful puzzle. Hate Fences, so back to Yin Yang. I was pleased to make the second of James' deductions i.e. that in each sub-grid there would be a boundary colour switch in the overlapping areas which meant that the outer boundary of the whole grid would be one colour. But which colour? I guessed (correctly) but even then I could only fill sub-grids C and D. Resorted to guesswork for sub-grids A and B and eventually found solutions. It was very hard to visualise what was going on, especially at the guessing stage. Lucky to complete this before the time ran out.

Afterwards I solved the Battleships puzzle. Thought this was a lovely puzzle too, but as someone else pointed out, it was really one big puzzle, starting with grid A, making deductions about the top 3 rows of grids B and C, and then flitting between grids B-D, before finally choosing one of two possible solutions for grid A. I still have got next to nowhere with the Fences puzzle. James shared an insight into the constraints in the overlapping areas. I haven't grasped this yet. Any chance of a further explanation.

Thank you,

David.

emma
Posts: 314
Joined: Wed 22 Jun, 2011 4:58 pm

Re: Seven Samurais - LMI June Puzzle Test

Post by emma » Thu 21 Jun, 2012 1:21 pm

david mcneill wrote: Afterwards I solved the Battleships puzzle. Thought this was a lovely puzzle too, but as someone else pointed out, it was really one big puzzle, starting with grid A, making deductions about the top 3 rows of grids B and C, and then flitting between grids B-D, before finally choosing one of two possible solutions for grid A.
Or in my case, giving up, but not without noticing that the ambiguity in grid A was at the right-hand side!

PuzzleScot
Site Admin
Posts: 2745
Joined: Fri 18 Jun, 2010 10:45 pm
Location: Edinburgh, Scotland

Re: Seven Samurais - LMI June Puzzle Test

Post by PuzzleScot » Thu 21 Jun, 2012 5:59 pm

david mcneill wrote:James shared an insight into the constraints in the overlapping areas. I haven't grasped this yet. Any chance of a further explanation
I think I know what he's referring to - I deduced that the loops must visit all 8 grid intersections.

To visit an overlap region, you must cross into the grid area for the adjacent puzzle. From THAT puzzle's perspective, there cannot be a line appearing from nowhere. Therefore at the cross-over point, there is a node that has 3 lines emanating from it (can't be 2 or 4). That can only occur at the common corner of the overlap area. The same applies when the line exits the overlap area, hence both corners of each overlap area need to be used.

It's equally valid, in principle, that the entire overlap between 2 sub-puzzles is missed out entirely, but the presence of clues therein discards this as a possibility.

edit: I've just noticed another constraint at this junction. Clearly the 3 lines emanating from the junction point cannot all be within any one grid. Therefore, 2 of these lines must be at 90deg to each other against the grid border. (eg, in the centre hole, 8 of the 12 lines can be filled in immediately.)

edit 2: An implication of the first observation is that no line can travel to the overlap border, except at the aforementioned corners. This means, for example, both lines of the '2' (R14C16) can be drawn in immediately. (vertically parallel)

kiwijam
Posts: 700
Joined: Wed 01 Jun, 2011 10:03 am

Re: Seven Samurais - LMI June Puzzle Test

Post by kiwijam » Fri 22 Jun, 2012 5:31 am

Yes, I think Alan has explained most if it.
To summarize, the loop can only enter/exit a 3x3 corner region along the very outside edge of the grid.
So for each 3x3 overlap zone, you can immediately (around the outside of it) draw in 4 segments, and mark 8 segments as not possible.

And a secondary point is that a 1 or a 3 diagonally outside an overlap corner (there are 4 of these clues, 1 per grid) behaves like it would if it were in a normal grid corner, so the 3 gets two definite lines and the 1 gets two not-lines.

Post Reply