SudokuSolver Forum http://www.rcbroughton.co.uk/sudoku/forum/ |
|
Archive Part D. Assassins 101-125, variants, other puzzles http://www.rcbroughton.co.uk/sudoku/forum/viewtopic.php?f=3&t=881 |
Page 1 of 7 |
Author: | Andrew [ Thu Apr 21, 2011 11:56 pm ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
This is Part D of the Assassin Forum Archive, moving on to include puzzles posted on this site and provide easy links to the puzzles and threads. Please read the first part of the Archive Index to get the background to this archive including Mike (mhparker)'s original post about ratings. Old SSv3.2.1 scores: Puzzle rating table, with links to archive entries; each of these has a link to the puzzle thread. Abbreviations used in Rating Table: +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ +------------------------+-----------+-------+-------+-------+-----------------------+-------+ Some of the selected quotes in the puzzle entries have been edited to remove "spoilers"; the full rating comments are included with the walkthroughs. In some cases the puzzle makers gave hints; these are included in tiny text in the selected quotes. Many thanks to Ed for providing the original rating tables, including links to the puzzle threads. I think it was a great idea that in the rating tables he provided separate columns for Afmob and myself, the most regular posters of walkthroughs. Thanks also to Børge for generating so many diagrams, which are also in the Images with "udosuk Style Killer Cages" thread. If you solve any of these puzzles and decide that there aren't enough walkthroughs for that puzzle or, better still, that you've found an interesting way to solve it which hasn't been posted, please feel free to post your walkthrough in the puzzle thread. Your comments and walkthrough will then be added to this archive. |
Author: | Andrew [ Fri Apr 22, 2011 12:09 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Messy One #1 by Ed (April 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Ed: Start of an occasional series - Messy One. Must get a SudokuSolver Score from 0.95-1.05 and is messy (NOT symmetrical). Afmob: We can't have a Killer on our new home without a walkthrough, can we? Nice idea Ed though it will be quite difficult to get the SS rating right since the range of accepted ratings is quite small. This is probably one of the rare cases where it's harder to create a Killer than to solve it. Rating: 1.0. Mike: Nice, enjoyable, and (above all!) messy killer, Ed! Thanks! Just to give Afmob a bit of company, here's my WT. I used a move (step 5), which almost certainly is not necessary, but which added to the fun, so I thought "why not?". Like Afmob, I'll rate this one at 1.0. Andrew: This was a fun puzzle and a fairly easy one. Even so the solving path is fairly well defined so that my route was very similar to those already posted by Afmob and Mike. I'll also rate Messy One #1 at 1.0. I thought about a lower rating but it's still a reasonably long path and harder than newspaper puzzles ... so I feel that 1.0 is justified. It's my first walkthrough for many months, probably over a year, that doesn't include the word killer ... Walkthrough by Afmob: Walkthrough by Mike (mhparker): Walkthrough by Andrew: |
Author: | Andrew [ Fri Apr 22, 2011 12:15 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Messy One #2 by Afmob (April 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Afmob: Here is my entry to the beauty of the asymmetrical. If you don't find the cracking move which is available right from the start, then you'll have a harder time than SS suggests. SS Score: 1.04. Estimated rating: (Hard) 1.0, depends on how long it takes you to find that move. Hint: Look at Innies/Outies of R1234. Mike: Nice puzzle, Afmob! Don't think I saw the cracking move you intended though, because I didn't see anything like that that was available "right from the start". Haven't read your TT hint. Nevertheless very enjoyable and still only about a 1.25 rating the way I did it... gary W: I thought this one was excellent. Many thanks Afmob Andrew: Thanks Afmob for a challenging puzzle. Yes, it's available from the diagram at the start. That move isn't a difficult one but it's hard to spot (my step 30). Maybe that's because the eye is drawn away from the move. I'll rate the way I solved it as Hard 1.25. I used 16 different 45 tests, still slightly short of my record! Ed: Agree. It was my key move and was very hard to find. Here's an alternative way to get past Mike's step 8. It is the same as Andrew used, just gets to it a lot quicker. I'd still give this way a 1.25 rating. Walkthrough by Mike (mhparker): Solving outline by gary W: Walkthrough by Andrew: Alternative ending by Ed: |
Author: | Andrew [ Fri Apr 22, 2011 12:23 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Cyclone V1 by Mike (mhparker) (April 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Mike: Welcome to my first puzzle(s) on the SS forum! Tried to create a nice killer with a whirling cage pattern, and always ended up with one row and column left over. So decided to simply cut them out and go for an 8 x 8 killer instead! Est. rating: 1.25. Have fun with those "36" tests! Afmob: Thanks Mike for providing a nice Killer which had an interesting move (step 2f). But I don't know if I tackle V2 because I am used to solve Killers with SumoCue and it's quite tedious to make eliminations with other programms, so I had quite a fight to solve it using JSudoku. Rating: Hard 1.0. Andrew: It shouldn't be any harder using my Excel worksheet than a standard 9 x 9 killer; the eliminations get done by me either way. I guess the hardest thing may be getting used to 2 x 4 blocks instead of 3 x 3 nonets. Thanks Mike for a fun puzzle. ... Thanks for that table, Mike! In a position similar to where Afmob found his interesting move I found a different variation. It wasn't quite as direct as Afmob's one but mine also led to a fairly quick solution. I agree with Afmob and rate Cyclone Killer V1 at Hard 1.0. Mike wrote: BTW, here is a killer combination table for 8x8, which you may find useful: Walkthrough by Afmob: Walkthrough by Andrew: |
Author: | Andrew [ Fri Apr 22, 2011 12:30 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Cyclone V2 by Mike (mhparker) (April 2008) here (below Cyclone V1) Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Mike: Welcome to my first puzzle(s) on the SS forum! Tried to create a nice killer with a whirling cage pattern, and always ended up with one row and column left over. So decided to simply cut them out and go for an 8 x 8 killer instead! Est rating: 1.75. Have fun with those "36" tests! Afmob (after going through my WT) also commented about the 1.75 rating being too high. I agree, but would still rate it as 1.5, because there's at least one important move in there that is difficult to see. Andrew: Cyclone Killer #2 was a really challenging puzzle. It took me a few sessions but I eventually got it out. Mike's step 15b was a nice one; it really broke the puzzle open. I also missed Mike's step 12a but found a couple of similar steps. I realised later that my step 27 is the complement of Mike's step 12a. Mike rated this puzzle at 1.5, after feedback from Afmob. I agree with that. It was Hard 1.5 the way I solved it. Mike wrote: BTW, here is a killer combination table for 8x8, which you may find useful: Walkthrough by Mike: Walkthrough by Andrew: |
Author: | Andrew [ Fri Apr 22, 2011 12:41 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Assassin 101 by Nasenbaer (May 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Nasenbaer: A little bit earlier than I intended but I'm not at home so it's now or never to give you the first Assassin for this forum. Back to the roots - that's what we should call this Assassin. After I created and solved it I realized that I had seen this design before (except for the diagonals) - in Assassin 2! Ed convinced me to go along with it (the old ones are often the best - the design, not Ed ), so here it is: SSolver 3.0 rating: 1.23 Afmob: Thanks Nasenbaer for providing the first Assassin on this forum. I think SudokuSolver's rating is a bit too high. After looking at SS solving path it is my step 6b which SS misses and which explains the difference between my and SudokuSolver's suggested rating. Rating: (Hard?)1.0. Mike: Actually, the Assassin 2X is (like Ed ) not all that old... Fortunately (talking about being old...) my memory is obviously even shorter, so had to "re-invent the wheel" again! As to the rating, I'd personally score it at 1.25, because it took quite a lot of chipping away at the candidates before the puzzle finally gave in, with lots of small moves being required, making the puzzle IMO too "finicky" for a 1.0 rating. I didn't keep a WT this time, but would like to take this opportunity to thank Nasenbaer for a challenging but rewarding "re-match"! Andrew: Thanks for posting the first Assassin on this forum! I think it's also the first time that the "Assassin of the Week" has been a Killer-X. It will be good to have them as the main puzzle occasionally, maybe once a month. Yes, it's very like Assassin 2, A2X and A2X-Lite except that the split cage is on the other diagonal. I'll agree with Mike and rate A101 at 1.25. My solving path is fairly similar to Afmob's, so maybe there's one clear path, but a number of my steps are very different. Walkthrough by Afmob: Walkthrough by Andrew: |
Author: | Andrew [ Wed Apr 27, 2011 11:17 pm ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Assassin 102 by Afmob (May 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Afmob: This Assassin offers lots of Innies and Outies to analyze. Use the right ones and together with basic killer techniques you can solve fast. Note that this Killer was derived from a famous Sudoku. Do you know which one? Hint : This Sudoku is as hard as metal. Hint 2: Ed said a part of its name in his wt post. SS Score (v3.1): 1.09 Estimated rating: 1.0 - Hard 1.0. It felt nearly (but not quite) as difficult as A101. gary W: Now to Afmob's puzzle which based on my solving time (2 hours plus) I'ld rate about 1.25 Ed: No idea. Don't know any famous vanilla! Bit one-eyed! A102 is a really fun puzzle - lots of hidden nuggets. ... agree with gary on the 1.25 rating. I really enjoyed this puzzle. Thanks very much Afmob! Mike (mhparker): In the land of the blind, the one-eyed man is king . As Børge would say, I'm completely clueless! Nasenbaer: A fun puzzle, all right. But it took me a lot longer to solve it on my first go than A101 because I didn't see the killer pair {89} in c1 which sets r2c1 to {89} . My notes from the second run look like Ed's walkthrough so there will be none by me. Rating: for me it was a bit harder than A101... I would give it a 1.25 Afmob: The famous Sudoku is: Solving outline by gary W: Walkthrough by Ed: Afmob's steps which explain why he estimated A102 at Hard 1.0: Walkthrough by Andrew: |
Author: | Andrew [ Wed Apr 27, 2011 11:41 pm ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Assassin 102X by Mike (mhparker) (May 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Mike (mhparker): Nasenbaer's A101 broke the mould by being the first original Killer-X Assassin. Afmob's A102 did likewise by being the first Assassin to contain mulitple diagonal cages. The puzzle below continues both themes by simply combining them. Unless I'm very much mistaken, it should make a refreshing change from the usual "fiddly" V2... SudokuSolver v3.1 scores this at 1.06, which is probably not far off the mark. Happy manual solving! Afmob: There is nothing like coming back from a sunny (and nearly killer-free) week and writing your shortest walkthrough (good bye, uA 99) so far. Though my walkthrough is very short this assassin is not a push-over. Rating: 1.0. Andrew: Thanks Mike for a fun variant! Quite a bit different from A102 but clearly still a variant. In the introduction to Messy One #2, Afmob wrote: If you don't find the cracking move which is available right from the start, then you'll have a harder time than SS suggests. That also applies to A102X.I'll rate A102X as Hard 1.0. Walkthrough by Afmob: Walkthrough by Andrew: |
Author: | Andrew [ Thu Apr 28, 2011 1:04 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Assassin 103 by Mike (mhparker) (May 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: Code: Select, Copy & Paste into solver: Solution: Quote: Mike (mhparker): The cage pattern for this Assassin reminded me of a mounted photo, so I decided to stick with this analogy whilst coloring it. A picture may be proverbally worth a thousand words, but this one sure ain't givin' anything away! Est. rating: 1.25. SudokuSolver v3.1 scores this at 1.24, which seems pretty accurate again. Ed: Good luck! I get an SS(v3.1) score of 1.74. Might need that good luck and 2 weeks! Mike replied: Hmm. You're right. On my Windows Vista installation, I get 1.75. However, I've just repeated the same test on my Windows XP installation, and got the published score of 1.24 again. Looked at the solver list and found that the default scoring options were not set. For some reason, SS did not alert me to this fact like it normally does. Despite this, I do quite a bit of analysis before posting a puzzle, and would not have published it as a V1 if I had felt that it was a genuine 1.75. The moral of this story is that we shouldn't always take what the automated solvers say for granted, even though they have an invaluable role to play in the puzzle evaluation process these days. HATMAN: Interesting puzzle - I managed to do it with a bit of T&E and I think I can now see how to do it without. Afmob: This was an interesting Killer where the hardest move (step 4b) was available right from the start. Thanks Mike! I'm not sure about the rating. My walkthrough could also justify a 1.25 depending on how you rate step 4b. My first gut feeling went for 1.5, so I'll stay with that. Andrew: Thanks Mike for a challenging puzzle. Yes, it (the key breakthrough step) was there right after the Prelims. It took me a very long time to find this key move. Then I just came across it after I found that the other moves I was working with didn't seem to be making any more progress. With hindsight I feel that I ought at least to have got it after step 19 but I didn't find it until step 54! Solving comment by HATMAN: Walkthrough by Afmob: Walkthrough by Andrew: |
Author: | Andrew [ Thu Apr 28, 2011 2:50 am ] |
Post subject: | Archive Part D. Assassins 101-125, variants, other puzzles |
Human Solvable 1 by HATMAN (May 2008) here Puzzle Diagram: Images with "udosuk style Killer Cages" by Børge: A much easier version (not within the Human Solvable classification) without r5c5 being part of a cage, i.e. r5c5 is a Zero Killer cell Warning: Viewing these diagrams may be a spoiler for the full puzzle: Code: Select, Copy & Paste into solver: Solution: Quote: HATMAN: One of my interests (and I think Matt's) is creating human solvable puzzles that the programs have difficulty with. HATMAN's full introduction and various discussions are included in the first hidden section below, so that any "spoilers" are hidden. Afmob: Finally a long walkthrough because my present wts kept getting shorter and shorter. But this difficult Killer deserves a long walkthrough. udosuk: after studying Afmob's step 11 I realise there is a easy way to solve this puzzle. Kudos to Afmob for spotting the breakthrough point. Andrew (in 2012): I tried Human Solvable 1 for the first time a few days ago and managed to finish it a couple of days later. Thanks HATMAN for a challenging puzzle! I enjoyed some of the steps in the other walkthroughs, breakthroughs and outlines… Rating 1.75. HATMAN's full introduction and various discussions: udosuk's original breakthrough steps: Walkthrough by Afmob: udosuk's easier breakthrough, after studying Afmob's step 11: Walkthrough by Andrew (in 2012): |
Page 1 of 7 | All times are UTC |
Powered by phpBB® Forum Software © phpBB Group https://www.phpbb.com/ |