Log in

26 August 2010 @ 02:24 am
Friday Puzzle #64 - Shikaku and Masyu  
Two intermediate difficulty puzzles this week, both originally written for the Turkish Puzzle Championship (TBT) test as about 5 minute puzzles. Both are also indicative of what my next puzzle project will look like, with classic puzzles but different kinds of themes and (hopefully) fun solves.

Rules - Divide the grid into rectangles along the grid lines so that exactly one given circle is inside each rectangle and so that the number in that circle indicates the total area of the rectangle. See also here.

Rules - Draw a single closed loop in the grid passing vertically and/or horizontally through the centers of the white cells. The loop must pass through every cell containing a circle. When passing through a black circle, the loop must make a 90 degree turn and must pass straight through the first cells encountered on each side of this turn. When passing through a white circle, the loop must go straight through the white circle, but must make an immediate 90 degree turn in one of the adjoining cells. See also here.

stigant on August 27th, 2010 07:13 pm (UTC)
The center of the Masyu was an interesting construction, but I think it was a tad too trial-and-error for my taste (only a tad though). It's pretty obvious that you have to start there, though, so maybe that's not too bad (or maybe there's a more straight forward way to resolve the center that I'm not seeing). Having resolved the center pinwheel, though, the left and right ones had a nice trick and I think nailed the perfect balance of an interesting deduction that doesn't require too much branching. The type of thing you'd see in a larger Masyu from a certain other author who's name begins with a 'J'.
(Anonymous) on August 27th, 2010 08:44 pm (UTC)
I actually started (after the obvious edge effects) with the left pinwheel. I focused on the topmost black pearl in that cluster, and concluded that the path could not go either up/left or down/left at that corner, forcing the path to the right. Then the pinwheels sort of spammed across. It wasn't /too/ deep into trial and error, though I did have to think a bit ahead, certainly.
- Jack
stigant on August 28th, 2010 03:58 pm (UTC)
Ah, that is interesting, and (evidently) unintended.
motrismotris on August 27th, 2010 08:52 pm (UTC)
If you look at the *right* thing [any black segment going towards the inside of the box, and the contradiction that arises], the trial and error isn't so bad in the center, but it should teach something for the outside boxes. I did appreciate that the center would be a lot more fidgety than is typical for a Masyu though since its a 3-deep reasoning I suppose.

In playing with patterns, I found the 4x4 black gear pattern is somewhat constrained - with whites it is absolutely constrained - and there is some imbetween with variable positions and numbers of white circles. So a larger junoesque puzzle with lots of gears can likely be built, but I thought this size would have enough meat to be appropriate for a championship puzzle.

Edited at 2010-08-27 08:54 pm (UTC)
stigant on August 28th, 2010 04:00 pm (UTC)
Yeah, that was essentially my experience with the puzzle. I looked at what happens if the black pearls go in twice before I finally got it, but in both instances, my reasoning only went two deep. The third time, I finally saw the next step in that line.

(I did enjoy the puzzle... one's limit on how deep you can go on a backtracking expedition is always a personal preference, and I'm always trying to expand my own horizon.)
TH: pencilrpipuzzleguy on August 28th, 2010 02:21 am (UTC)
Solved the Shikaku by uniqueness; I assume that wasn't intended. Still stumped on the Masyu; I'm avoiding the other comments.
motrismotris on August 28th, 2010 02:30 am (UTC)
I never intend uniqueness, but a shikaku theme like that is really set to exploit if you trust me.
(Anonymous) on September 1st, 2010 03:20 am (UTC)
I too used uniqueness avoidance on the very bottom to push through at a certain moment in the solution. I felt dirty but was tired of looking boxes. I am a weak-hearted solver of these types.

(Anonymous) on September 1st, 2010 03:28 am (UTC)
Masyu melted before my eyes ...
I gotta say that the center of the Masyu unfolded in a hurry for me - possibly faster than for others because I accepted Tom's comment that it would be a 5-minute solve and figured that it would if pushed (confidence is a huge asset in solving - even when false confidence).

Here it goes:
1. Immediately saw that all whites had to be either all parallel to the center or perpendicular to the center.
2. If perp, then each black needed to have a leg going outward. If parallel, then each black needed to have a leg going outward.
3. I dropped down the outward black legs, and they started directing traffic in the pinwheels and it ran from there.

Go figure. I did something right (and in a hurry) for once.

Thanks for some nice post USPC diversions. Right after the competition, I am always starving for more.

MellowMelonMellowMelon [wordpress.com] on September 1st, 2010 04:01 am (UTC)
Re: Masyu melted before my eyes ...
Sometimes I think that confidence is one of the main reasons for my success in competitions. I've always gone into things like the Putnam and the USPC shooting for a perfect score and keeping that mindset for almost the entire time limit. Besides giving you that confidence you mentioned, it also makes sure you won't skip over a problem/puzzle that looks intimidating and turns out to be easy.

It's easy to have it when someone says a puzzle is quick. When someone says one is hard, I consider it a personally-directed challenge to my ability and get extremely determined to prove them wrong.

I think this is at the core of the reason why things like unlogicable puzzles at competitions (WSC4...) are incredibly frustrating. You fail to solve it and feel like the constructor beat you, then learn that they had to "cheat" to do it. No one likes losing to cheaters.
(Anonymous) on September 2nd, 2010 10:37 pm (UTC)
I finally got a chance to solve these. The inner ring in the Masyu was a little too much trial-and-error for my Masyu tastes (not to mention the fact that this one was almost *too* symmetric - large portions of repeated solving), but I enjoyed the pin-wheels of 4 blacks. Definitely worthy of being showcased in a big'un, if you ask me!

Shikaku was a bit fiddly for me. I think this type, along with nurikabe, is where nikoli.com really spoils the solver with its excellent interface; although perhaps the fiddliness was down to solving in MS paint :)