Friday 25 June 2010

Friday Puzzles #58

This week dear readers, I am appealing to you for a little feedback. It seems every puzzle designer worth their salt has their own pet sudoku variant they’ve created, nurtured and loved. Oh the long days I looked on in envy!

Anyhow, today that all changes as I’ve come up with something that is (a) original to the best of my knowledge and (b) isn’t totally lame. I call it Sudoku Islands. The idea is basically a hybrid of sudoku and nurikabe. You have an NxN grid, and the idea is to fill in each column, row and box with the digits 1-M, where M < N. Moreover these digits form a connected region of the grid, with no 2×2 blocks, which surround islands. The size of these islands is clued by the grey squares, and there is exactly one given clue per island in the solution. Unfortunately if you are used to nurikabe much, the roles of white and grey (black) squares have become reversed, but you’re playing by my rules now so I guess that’s hard cheese!

What you get is in my mind something that works out to be an improved version of extra space sudoku, where you not only have to fill in the grid, but also have to work out where the extra space is. That’s all beginning to sound a bit wordy, so here’s a really easy toy example for you to get the idea. Fill this grid with digits 1-4 in each row/column/box:
    #068 Sudoku Islands – rated very easy
And this one with digits 1-5. Enjoy!
    #069 Sudoku Islands – rated medium
All puzzles © Tom Collyer 2009-10

Friday 18 June 2010

Friday Puzzles #57

Another week, another Masyu puzzle. As promised from last week, this features a novel new trick which you might have missed last week as it wasn’t required in the solve. So whilst I haven’t really had the time to do a juno with it this time round, you will at least have to spot it. Actually I claim it’s a novel new trick, but that’s only to the best of my knowledge.

Additionally this week, I’ve tried to give this total twisted symmetry, but it manages to break down in a couple of places. It turns out that a nikoli.com author by the name of cubic function has simultaneously been incorporating twisted symmetry into his puzzles – including in much larger sizes than mine for a while now too. His designs have perhaps been less striking than mine, and I’ve only picked up on it after solving a second time.
    #067 Masyu – rated hard
All puzzles © Tom Collyer 2009-10

Whilst I’m on my soap-box, actually, I read this week that the “world record” for sudoku speed solving has been broken. Now the world record criteria – here denoted as STD-27 – are somewhat arbitrary, presumably to ensure a hard puzzle, but I was the first person to question this, and made a puzzle satisfying world record criteria over a year ago, solvable in the order of 2 minutes. I believe a certain Mr Snyder held that unofficial record. Anyhow, the Czechs have made an even easier puzzle than mine, and the record was broken a total of three times at this event, ultimately by Jakub Ondrousek.

Friday 11 June 2010

Friday Puzzles #56

And so it’s back to the weekly schedule. I think I’ll leave it at that, suffice to say after the uniqueness hiccough with last week’s puzzle I will be using the following phrase a little more carefully!
Actually, I’ve got a bit of a puzzle (sudoku) project in the pipeline so I probably won’t be releasing too many puzzles of any great size in the next few weeks. But don’t let that put you off: as everybody knows, small is beautiful. And this week’s masyu is indeed a thing of great beauty. Even if I say so myself. Enjoy!
    #066 Masyu – rated easy
All puzzles © Tom Collyer 2009-10

Friday 4 June 2010

Friday Puzzles #55

Mission accomplished! Weekly numbering is now completely in order. Modulo the missing #7 and #17 that is. Still, there were a couple of weeks where I made bonus puzzles so I think that just about balances out the equation. I must add that I revisited the double-toroidal sudoku, but I have to say that with its current grid I lost a bit of interest.

Anyhow, a sudoku variant this week which I think I’m right in claiming has only ever been pulled off by yours truly – although please correct me if I’m wrong. In case you missed the original post the idea with 10’s sudoku is normal sudoku rules, except that every time a pair of adjacent cells had digits summing to 10, they are marked. Where there are no markings between adjacent cells, the conclusion is then that the relevant digits cannot sum to 10.

Prior to that post, I can’t find an example where 10’s sudoku had no such marked occurrences between pairs of digits, but it turns out with a bit of work that it’s perfectly possible. I’m sure a computer-y type would easily be able to write a program to churn out more of these without working up any sort of a sweat, but that’s not the way I roll. There’s plenty of potential with this variant to churn out some really fiendish challenges; this example is fairly plain sailing, but perhaps with a couple of surprises along the way. Enjoy!

UPDATE: My apologies to those who had a go at an earlier version of this puzzle, which had two solutions, and my thanks to Jack who pointed this apparently obvious blooper I managed to completely miss! My reputation for errors seems to be growing, but stick with me anyway – at the very least you won’t be bored!
    #065 No Tens Sudoku – rated medium
All puzzles © Tom Collyer 2009-10

Thursday 3 June 2010

Friday Puzzles #54

Again, I find myself in two minds as to how this puzzle ought to be classified. You see, the problem is most Hashi that I do are fairly generic, and probably require a solving set of about five moves applied with various degrees of imagination. With this Hashi, there are two polarised solving experiences that I can envisage you, dearest reader, having. Either something clicks, and everything gets pretty standard pretty quickly, or else you’ll struggle a little bit and the solve will turn into something of a slog.

It seems to me that the best case scenario for me as the author is that everyone manages to click. In which case there seems to be only a minor pay-off before everything gets standard. Still, having steadfastly argued earlier this week that an otherwise unspectacular puzzle still had some merit to it based on a novel and interesting move to finish it off, I suppose I should be happy with this. Enjoy!

Update: My thanks to Bryce Herd who pointed out that an earlier version of this puzzle failed to satisfy a basic theorem of graph theory!
    #064 Hashiwokakero – rated hard
All puzzles © Tom Collyer 2009-10

Friday puzzles #53

Well, there are two puzzle due before the end of the day if I’m going to get my numbering on track, and both of them are going to be Hashi. If you’ve never seen this puzzle before – after all, it is the first time they’ve been featured on this blog – then as ever there’s a handy link in the “How To Play” section there on the right.

This should be a nice warm-up for you – there’s a harder one coming this evening. After all, if you can’t win a game of noughts-and-crosses (or, I guess, tic-tac-toe for my American readers) then at least you should be able to solve a puzzle! :)
    #063 Hashiwokakero – rated easy
All puzzles © Tom Collyer 2009-10

Contact Form

Name

Email *

Message *