Runners, welcome!

nh2offbanner

Welcome to another set of Nikoli Hurdles!
For the next 29 days, a track of 8 Nikoli puzzles will be opened for you to solve. Starting at Puzzle 1, the solution will be used to open Puzzle 2. The answer to Puzzle 2 will open Puzzle 3 and so on.

The different puzzle types will require different answer key extractions so pay attention to those. An entry form will be at the finish line for those who complete the entire track. More details can be found at the About page on the menu above.
Don’t forget to use capital letters, when required.

I will always be available to answer questions, so you can drop those in the comments or email me at roygbivpuzzles (at) gmail (dot) com. A prize will go to one lucky winner after this event ends.

ETA (27/9/14): 8 hours into the contest, early runners had sent me problems about some of the puzzles.
Puzzle 2 had a little tweak in one of its clues, I’m surprised people could still proceed. The last Puzzle 8 had one slight drawing error. Both puzzles have been fixed. Puzzle 7 was questioned as well, but upon reinspection, it is not broken and has a unique solution.
The finish line had the wrong password for about 8 hours, but it is now fixed. I apologize for all these mistakes.

Puzzles on Patreon

Ever since Grandmaster Puzzles offered opportunities to buy rewards on Patreon, we see more puzzle providers following the trend. Currently Grandmaster Puzzles enjoy a little over $500 a month of support from its audience. Considering that the patrons get a very good quality puzzle almost daily, a few dollars seem like a bargain. GMlogo

As a patron myself, I do enjoy the weekly puzzle PDFs sent promptly to my inbox before they are released the next week on GMP. You can also get access to solutions, walkthroughs, monthly large puzzles and even puzzle-packed PDF booklets. I can tell a lot of work has been put to maintain the standards at GMP so if you’re not already a patron, I highly recommend this innovative source.

In shocking contrast to Grandmaster Puzzles, Logic Masters India (LMI) has 9 patrons. Nine??
LMI has been like a home to me for the last 4 years and has over 6000 registered users with hundreds of active solvers logging in every month.

LMINewLogoLMI offer cute puzzle packs containing puzzle types from the latest online tests. The incentives on their Patreon page may not be as attractive as the offerings from GMP, but LMI has done such a hugely demanding job running regular online tests and constantly improving their web formats that active users on LMI should reconsider becoming a patron.

Most recently, the folks behind Akil Oyunlari launched Meraklisina Akil Oyunlari. I was late to discover the magazine when I first contributed to issue 76. 4 issues later, Akil Oyunlari underwent a drastic format change and all of sudden it became a children’s magazine. Growing up solving kids’ puzzle magazines, I realize how important these magazines play a role in attracting a child to still be a puzzler when he/she grows up.

Logo2Nonetheless I had to order a few back issues to satisfy my puzzle craving as ones in recent format are evidently not aimed at experienced readers. But now it is back, a 10-a-year online PDF puzzle magazine with logic puzzles for expert-level solvers. If anyone knows of any magazine containing just logic puzzles, do let me know. So far I know of Akil Oyunlari, Breinbrekers (which I saw the Dutch team solve at last year’s WPC – is there a way to buy this without going to Holland?) and a few titles by Gareth Moore at Puzzle Xtra.

Show your support and keep the puzzling world spinning.

In SOTR news…

previewnh2Nikoli Hurdles 2 is on its last corner of editing and I will post details about this event soon. Grab some Nikoli books and start warming up!

Light Read #3 – Let me flip that

So I was staring at the 2nd Pentopia of the WPC 2013 practise test written by Ko Okamoto thinking how I ended up with multiple solutions. I reread the rules at least four times and came up empty. How odd, I thought, since I have solved a few Pentopias before. I looked at the first Pentopia of the set, which was straightforward, and found nothing wrong.

Hmmmm…

25570910-225536-82536717.jpg

Pentopia by Ko Okamoto (Oct 2013)

A fifth look at the rules saw me emphasizing the last sentence: pentominoes may be rotated but NOT reflected. There we go. The puzzle solved itself.

There’s nothing upsetting about forbidding reflection, but it leaves an empty room for a flexibility-loving nut like me. When I construct puzzles using polyminoes I have Blokus pieces handy to quickly prove uniqueness. Blokus players will no doubt have the habit of twisting and reflecting polyminoes in their heads to optimize those nasty land-grabbing moves.

Any other puzzlers out there who plays Blokus?

Any other puzzlers out there who plays Blokus?

Which leads me to my point: why disallow reflection of pentominoes in puzzles?
Reflection affects 2 tetrominoes, namely the L (or J) and S (or Z), and thankfully most tetromino puzzles forgive reflection. Imagine a LITS where S and Z are two different pieces; there goes a boatload of beautiful deductions.

However when we’re dealing with pentominoes, 7 of 12 pieces form a “different” shape when they are flipped over. This reduces the number of possible orientations rather significantly. Puzzles with this rule will be easier since there will be less possible positions. Not only for solvers, but this also cut authors some slack when constructing the said puzzle.

Have I been solving pentopias all these years without ever knowing is rule?
A quick check at Puzzle Para Site, since one would associate Pentopias with Bram De Laat, revealed that reflections are in fact allowed.
Good.

Maybe there are certain puzzle types where not allowing reflection is better. But for now, I’ll have my seven extra orientations, thanks.

Solutions to July Giants – Part 2/2

The menu is taken down, the puzzles are available in PDF attached in the previous post.
Picking up from the last post after the giant Masyu:

Snake Pit
Following such a complex Mastermind Tapa was a tall order. The giant Masyu was met with so much criticism that I had to rethink what I should post for the upcoming Medusa. Most were quick to point out that the giant puzzle was too simple and a lot less exciting that the Tapa they solved a few days earlier.

Originally planned for Medusa was a giant word search. However after receiving mails about how dull the Masyu was, one had to be arrogantly audacious to post a simple word search. I returned to the drawing board and whipped up Medusa’s Snake Pit. It took exactly 60 minutes to draft the whole puzzle.

I looked at the Snake Variation Contest authored by Serkan Yurekli to see which types were suitable in a giant grid. Snake puzzles usually involve clues outside the grid so my choices were rather limited. Masyu Snake was a no-go since Argus was already on the list of giants. I settled on the 5 variants and the puzzle was born.

medusaansMy intended break-in was in the left-hand portion of the Schlange. I wanted solvers to start off linking BB and JE to get their first toehold of the puzzle. The puzzle gods (Medusa?) didn’t agree with me and the obvious “3” of Slitherlink Snake at the bottom left hand corner was definitely a buzzkill. When I was test-solving the puzzle, almost none of my planned deductions were required at all. I imagine this happen all too often in giant puzzles. Life goes on.

I feared it might be too easy because of the heavily-constrained nature of Snake puzzles. So it was a relief to still see so many compliments. Antonis Lalatsas observed that this very quickly degenerated into a length constrained Arukone. Which is good because I like Arukone but suck at Snake puzzles.

Similarly Stefan Tomlins noted: It’s very clear there’s only one answer once you’ve got it (except for the remaining sliver of paranoia).

However the medal of courage has to go to Prasanna Seshadri who created extra work for himself: I solved on paint, and its a bit messy now.

Yuhei Kusui disagrees since: This is the most beautiful Medusa I have ever seen.


Fillomino Match-Ups
The final puzzle was Cerberus. The monster concept came before the puzzle so I thought about something with 3 heads. Maybe three overlapping Fillominos?
I had a brief look at the two Fillomino Fillia tests and scouted the 3 variants that were used. The idea of having a matchmaker-type of puzzle quickly worked itself out. No Rectangles was the first to be constructed and due to its nature, 1-cell and 2-cell regions were not allowed. I then realize how difficult it will be for the remaining puzzles. In order to mask early matching deductions, I was prohibited from using any 1 and 2 as clues in the other Fillominos!
Ouch. Otherwise, oh this puzzle has a 1 (or 2); therefore it cannot be the No Rectangles.

No-Rectangles Fillomino
fillo1ans
Depriving a Fillomino author from using 1 and 2 will make his life very difficult. I learnt this the hard way when Star Battle keeps getting broken. I ended up with that ugly clunk of 7s in the middle.

Star Battle Fillomino

fillo2ansAfter the 3 puzzles were made, I originally had in mind 3 more classics mixed in. Then after unsuccessfully attempting to write the first classic without the use of 1 and 2, I just gave up.

Odd Even Fillomino
fillo3ans
There were many different approaches to Cerberus.
Antonis Lalatsas shared his: I excluded grid 5 due to the interplay between clue 22 and the 563 triplet. This left only grid 6 respecting the Yin-Yangish border constraint of Odd-Even, which then solved quickly. Grid 2 was out next due to top 3 being forced straight and a single 2×2 remaining space for two stars on the top two rows. And that’s all I did logically.
The bottom right quarter of grid 1 looked woefully underconstrained unless I could force the middle 5s into different polyominos by bringing down the 4, and this quickly gave me a working Star Fillomino.

Emboldened, I randomly tried grid 3 as No Rectangles, working ccw from the 10 and failing in the three 3s to the left. I used the same failing logic to working effect this time on the remaining grid, which solved pleasantly.
Was there an obvious logical step (or three) I missed?

Short answer: to be honest I can’t remember. During test solving I remembered disproving all 3 fakes before solving each variants. I recalled solving all 6 grids under every rule to make sure they had one unique solution. Odd-Even was first if I remembered correctly and there was a toss-up between Star Battle and No-Rectangles. I didn’t check to see if a puzzle also work under another rule but apparently [Matej Uher]  found that grid 1 has multiple solution in no rectangular, so i try other possibility (star battle) :-/

Edison He mentioned I was stk (sic) for a while on the Star Fillomino. Was there any better way to get it besides assuming that the top right square is not filled by a star?

I agree that Star Battle was the hardest of the lot. I remembered the puzzle started from the bottom left and worked its way up. The last bit I recalled using guess-and-check to prove uniqueness so it was probably a similar method everyone used.

And lastly Ivan Koswara pointed out that people are used with Star Fillomino with two stars
Oops I forgot.

I finish this post with a funny comment from James McGowan who promptly finished all four giant puzzles.
With the body of the final monster at his feet, Jameus wiped the gore from his face and looked around. The sun shone through a gap in the clouds, somewhere a bird burst into song. Local villagers were weeping with joy, busty maidens swooned and flirted. “Now we feast!” he bellowed, and the crowd roared back its approval. It was a good day to be a puzzler…

I feel everyday’s a good day to be a puzzler.
With all the solutions posted, July Giants has now come to a close. My next event is in the works and originally planned for October. However, I have a hunch that it might be held near the end of this month if things continue to go smoothly.
Thanks for dropping by.
See you next time!