Assassin 60

Para · Post by **Para** » Wed Jul 25, 2007 3:51 pm

Hi all

Guess Cathy just beat me to it. But here's mine. It is a bit adjusted because the original was a big mess. loads of steps through each other. I missed far too many things early on to make it easy to read.

Walk-through Assassin 60RP-Lite

1. 10(3) at R1C1, R2C6 and R6C4 = {127/136/145/235}: no 8,9

2. 19(3) at R2C2 and R4C4 = {289/379/469/478/568}: no 1

3. 9(3) at R3C1 = {126/135/234}: no 7,8,9

4. 23(3) at R3C3 = {689}

5. 13(4) at R3C8 = {1237/1246/1345}: no 8,9; 1 locked in 13(4) cage -->> R56C8: no 1

6. 20(3) at R4C6 = {389/479/569/578}: no 1,2

7. 11(3) at R6C6 = {128/137/146/236/245}: no 9

8. 45 on N5689: 2 outies: R3C8 + R8C3 = 7 = {16/25/34}: no 7,8,9

9. 45 on R123: 1 outie and 2 innies: R4C3 + 5 = R3C18: Max R3C18 = 11 -->> Max R4C3 = 6: R4C3 = 6; R3C18 = {56} -->> locked for R3
9a. Naked Pair {89}at R3C34 -->> locked for R3
9b. Clean up: R8C3 = {12}(step 8)
9c. 13(4) at R3C8 = {1246/1345}; R3C8 = {56}; R4C89 + R5C9 = {124/134}: no 5,6,7 -->> 1,4 locked for N6

10. 9(3) at R3C1 needs one of {56} in R3C1: 9(3) = 5{13}/6{12} = {2|3..},{2|5..},{3|6..} -->> R45C1 = {12/13}: no 4,5; 1 locked for C1 and N4

11. 10(3) at R1C1 can’t have {23},{25} or {36} in R12C1 because of 9(3) at R3C1
11a. 10(3) at R1C1 = {27}[1]/{35}[2]/{45}[1] -->> R12C1 = {27/35/45} = {2|5..}: no 6, R1C2 = {12}
11b. Killer Pair {25} in R12C1 + 9(3) at R3C1 -->> locked for C1
11c. 5 in C1 locked for N1
11d. 1 in N1 locked for R1
11e. 1 in R6 locked for N5

12. 45 on C123: 3 innies: R138C3= 12 = [381]/{1[9]2} -->>R1C3 = {123};1 locked for C3; R13C3 = {3|9..}
12a. 19(3) at R2C2 = {469/478}({289} blocked by R3C3; {379} blocked by R13C3): no 2,3; 4 locked for N1; 6 only in R2C2 -->> R2C2: no 9
12b. 9 in N1 locked for C3

13. 10(3) at R1C1 = {27}[1]/{35}[2]: 2 locked for N1; R12C1 = {27/35}= {2|3}
13a. Killer Pair {23} in R12C1 + 9(3) cage at R3C1 -->> locked for C1

14. 45 on R89: 2 innies: R8C27 = 6 = {15/24} = {1|2..}: no 3,6,7,8,9
14a. Killer Pair {12} in R8C27 + R8C3 -->> locked for R8

15. 45 on N7: 1 outie and 2 innies: R6C1 = R78C3: R6C1 = {4789} -->> R78C3 = [52/71/72/81]: [31] blocked by R1C3 -->> R6C1: no 4; R7C3: no 2,3,4
15a. 4 in C1 locked for N7
15b. Clean up: R8C7: no 2(step 14)

16. 18(3) at R6C2 = [927]/{38}[7]/{37}[8]/[945] -->> R6C23 = {37/38}/[92/94]: no 5

17. 5 in N4 locked within 16(3) cage at R4C2 -->> 16(3) = {259/358/457} = {3|7|9..}

18. Combining 15 and 16
18a. R6C123 = [794]/[892]/[9]{38}: [9]{37} blocked by 16(3) at R4C2: 9 locked for R6 and N4: R6C2 = {389}; R6C3 = {2348}
18b. Clean up: R7C3 = {57}(step 16); 16(3) at R4C2 = {358/457}: no 2
18c. CPE: R7C2: no 9 -->> sees all 9’s in R6

19. Combination analysis 22(4) at R6C1 from step 15: When R6C1 = 7: no {25}(must be in R78C3); When R6C1 = 8, no {17}(must be in R78C3); When R6C1 = 9, no {27}(must be in R78C3)
19a. 22(4) = [7]{68}[1]/[8932]/[8635]/[9481]: R7C1 = {4689}; R7C2 = {368}; 8 locked within 22(4) cage -->> R89C1: no 8
19b. 8 in C1 locked within 22(4) cage -->> R7C2: no 8
19c. 22(4) = [7861]/[8932]/[8635] -->> R7C1: no 4; R6C1: no 9
19d. R6C2 = 9(hidden)
19e. Clean up: R6C3 = {24}
19f. R6C13 = [74/82] = {2|4..},{2|7..},{4|8..},{7|8..}

20. 4 in N7 locked within 23(4) cage at R8C2 -->> 23(4) = {2489/3479/4568}: no 1
20a. 1 in N7 locked for R8
20b. Clean up: R8C2: no 5
20c Naked Pair {12} in R8C23 -->> locked for N7
20d. 23(4) = {3479/4568} = {3479}/{46}{58} -->> R9C2: no 6

21. 22(4) at R6C1 = [7861]/[8932] -->> R7C1: no 6
21a. R7C123 = [865/937] = {6|7..}
21b. 17(3) at R7C7 = {39}[5]/{48}[5]/{58}[4]: {67}[4] blocked by R7C123: 5 locked in 17(3) cage for N9; R7C78 = {34589} = {39/48/58} = {8|9..}
21c. Killer Pair {89} in R7C1 + R7C78 -->> locked for R7

22. 45 on N9: 2 innies: R7C9 + R9C7 = {17/26}: no 3,4,8,9

23. 45 on R6789: 1 innie and 1 outie: R5C8 = R6C5 + 3 -->> R5C8 = {56789}; R6C5 = {23456}

24. 45 on N5: 3 innies: R4C6 + R6C46 = 11 = {128/137/146}(1 in N5 locked within these cells): no 5,9

25. 45 on R6789: 4 innies: R6C589 + R7C9 = 18
25a. R6C89 can’t have both {23} because of 13(4) at R3C8(step 9c)
25b. R6C589 can’t have both {24},{27},{48},{78} because of R6C13(step 19f)
25c. R6C4 can’t be 3 less than any of the other innies(step 23; R5C8 sees all other innies)
25d. R7C9 = 1: R6C589 = [6]{38} (others blocked by 25b and 25c)
25e. R7C9 = 2: R6C589 = [2]{68}/[3]{58}/[6]{37}(others blocked by 25b and 25c)
25f. R7C9 = 6: R6C589 = [4]{35}(others blocked by 25b and 25c)
25g. R7C9 = 7: R6C589 = [2]{36} (others blocked by 25a, 25b and 25 c)
25h. Conclusion: R6C5 = {2346}: no 5; R6C89 = {35/36/37/38/58/68}: no 2
25i. Clean up: R5C8: no 8(step 23)

26. 5 in R6 locked for N6
26a. Clean up: R6C5: no 2
26b. R6C589 = [6]{38}/[3]{58}/[6]{37}/[4]{35} -->> R6C89: no 6; R7C9: no 7
26c. Clean up: R9C7: no 1

27. 20(3) at R4C6 = {389}/{479}: no 6 -->> 9 locked within 20(3) cage in R45C7 -->> locked for C7 and N6
27a. Clean up: R6C5: no 6
27b. R6C589 = [3]{58}/[4]{35} -->> R6C89 = {35/58} = {3|8..}: no 7; R7C9: no 1; 5 locked within R6C89 for R6
27c. Clean up: R9C7: no 7
27d. Naked Pair {26} in R7C9 + R9C7 within N9

28. Killer Triple {348} within R6C13 + R6C5 + R6C89 -->> locked for R6

29. 11(3) at R6C6 = [173/164/263/623] -->> R6C6: no 7; R7C6 = {34}

30. 1 in R7 locked for N8 and 10(3) cage at R6C4
30a. R6C6 = 1(hidden)
30b. Clean up: R6C7: no 2(step 29)

31. Hidden Triple {124} in N6 within R4C89 + R5C9 -->> R4C89 + R5C9 = {124}
31a. 13(4) at R3C8 = {1246} -->> R3C8 = 6
31b. R3C1 = 5; R5C8 = 7; R6C7 = 6; R7C6 = 4; R9C7 = 2; R7C9 = 6
31c. R7C2 = 3; R7C13 = [97] (step 21a); R6C3 = 2; R8C23 = [21]
31d. R6C1 = 8; R6C4 = 7; R6C5 = 4(hidden); R1C23 = [13]; R2C2 = 6(hidden)
31e. R4C2 = 7(hidden); R3C2 = 4; R23C3 = [98]; R3C4 = 9; R5C23 = [54]
31f. R9C23 = [85]

32. Naked Pair {58} in R7C78 -->> locked for R7 and N9
32a. R8C7 = 4; R89C1 = [64]

33. 45 on N5: 1 innie: R4C6 = 3

34. 33a. R45C1 = [13]; R5C9 = 1(hidden); R9C8 = 1(hidden)

35. 15(3) at R5C5 = {29}4 -->> R5C56 = {29} -->> locked for N5 and R5
35a. R45C7 = [98]; R5C4 = 6; R9C4 = 3; R7C78 = [58]; R1C7 = 7; R12C1 = [27]

36. 45 on N3: 1 innie + 1 outie: R1C6 = R3C7 + 4 -->> R1C6 = 5; R3C7 = 1
36a. R23C6 = [27]; R2C78 = [34]; R1C89 = [98]; R23C9 = [52]; R4C89 = [24]
36b. R6C89 = [53]; R5C56 = [29]; R1C45 = [46]; R3C5 = 3; R7C45 = [21]
36c. R89C6 = [86]; R2C45 = [18]; R4C45 = [85]; R8C4 = 5; R89C5 = [97]
36d. R8C8 = 3; R89C9 = [79]

greetings

Para

Andrew · Post by **Andrew** » Wed Jul 25, 2007 4:14 pm

CathyW wrote:The 7th book. And no I'm not going to give anything away!!

The 5th film we will be going to see but probably not until after our holiday (5th-12th August in Chamonix, France) so I'll have to hope the A62 can be solved before I go and catch up with the A63 when I get back.

Enjoy Chamonix! We once stopped there for a few hours on a touring holiday and took the train up to the Mer de Glace. Well worth doing. The glacier and the surrounding mountains are magnificent.

mhparker wrote:Talking about books, have you read Andrew Stuart's "The Logic of Sudoku" yet, or is it still sitting on the shelf in pristine condition?

My copy is falling apart at the seams and is rapidly becoming a collection of individual pages!

You clearly need to buy a second copy!

I must admit that mine is near pristine condition. I have glanced at a few things but not yet studied it properly. Since our move it has been sitting on the bookshelf.

CathyW · Post by **CathyW** » Thu Jul 26, 2007 4:31 pm

Andrew wrote:Enjoy Chamonix! We once stopped there for a few hours on a touring holiday and took the train up to the Mer de Glace. Well worth doing. The glacier and the surrounding mountains are magnificent.

I'm sure we will. My hubby has been to the area ski-ing a number of times in the winter. I love the mountains but am not a skier so looking forward to some spectacular scenery, cable car rides, walks and almost certainly a trip to a 'natural swimming pool' not far away. Plus of course reading and puzzle solving while I'm waiting for the family to get up in the mornings!!

Probably going to take Andrew Stuart's book with me too - already well thumbed but there are still a number of advanced techniques that I haven't yet got to grips with (AICs and ALSs in particular).

Just noticed I'm officially an addict now (not merely hooked!!)

mhparker · Post by **mhparker** » Fri Jul 27, 2007 6:00 am

CathyW wrote:Probably going to take Andrew Stuart's book with me too - already well thumbed but there are still a number of advanced techniques that I haven't yet got to grips with (AICs and ALSs in particular).

Sounds good. I'll test you on it when you get back.

Para · Post by **Para** » Thu Aug 02, 2007 7:22 pm

As Cathy was asking this question as well. Did anyone else solve the Assassin 60 RP-Lite yet?

greetings

Para

mhparker · Post by **mhparker** » Fri Aug 03, 2007 9:19 pm

Para wrote:Did anyone else solve the Assassin 60 RP-Lite yet?

I heard from Ed a couple of days ago that he's finally finished it. (Ed, if you're reading this, how's the WT coming on?

).

I've also got a WT for it, which I can post if Ed doesn't beat me to it (another

there, Ed!).

Para · Post by **Para** » Fri Aug 03, 2007 11:59 pm

Ok. I guess it's safe to enlarge then. Awaiting WT's then. Wanted to compare notes. Really liked my way through that puzzle, just wondering for the different paths through this puzzle.

Just a general question, Mike. Do you solve your V2's(V2 just meaning a second version, no difficulty intended) before posting them?

greetings

Para

ps. I also have the unedited version of my WT saved for if anyone wants to see how i got to the moves in my walk-through.

mhparker · Post by **mhparker** » Sun Aug 05, 2007 8:43 am

Hi Para,

Para wrote:Just a general question, Mike. Do you solve your V2's(V2 just meaning a second version, no difficulty intended) before posting them?

No, I don't. Don't forget, I'm not keeping the solution and tweaking the cage pattern. Instead, I'm keeping the cage pattern constant, but modifying the solution, thus creating totally different puzzles each time. To assess the relative suitability of these puzzles by actually manually doing them would be impractical. Instead, one is dependent on (quicker) heuristical methods of evaluation based on stepping through the puzzle using automated solvers.

The RP Lite was a bit of an exception in that I only looked for trying to find a way forward myself after working through Cathy's WT. Therefore, to give her due credit, I may just add my steps to her WT (thus making an implicit tag solution out of it) instead of posting one done from scratch.

mhparker · Post by **mhparker** » Sun Aug 05, 2007 9:18 am

deleted (duplicated post)

sudokuEd · Post by **sudokuEd** » Mon Aug 06, 2007 10:35 am

Wow Mike. What a puzzle is RP Lite! I've been trying and trying to get ready for this post - but still don't feel I've found a solution I'm happy with so have some more work to do yet.

I want to respond to how you used software steps before posting this puzzle. Hmmm.

Para wrote:Do you solve your V2's(V2 just meaning a second version, no difficulty intended) before posting them?

mhparker wrote:No, I don't. .... by actually manually doing them would be impractical. Instead, one is dependent on (quicker) heuristical methods

I'm disappointed about this. I thought you had solved RP Lite yourself manually before posting it. I would have liked to have known that this was not the case. I assume that we are all manual solvers/puzzle makers except where stated otherwise.

My strong preference is for puzzles/versions that have been manually solved before being posted. I hope I can try and persuade you that this is not impractical for Assassin versions that have a different solution. I try to make these as well and don't find them impractical. Time consuming yes - but no more impractical or time consuming than the other types of versions in my experience.

Here are some of the shortcuts I do to try and get the difficulty level right.

1. Paste the potential puzzle string into Perfect Sudoku - if it rates at anything less than Insane then its not going to be a V2.

2. Paste the potential puzzle string into SumoCue then get it to solve by pressing F10. If SumoCue has to grunt to find a solution, then its probably going to be hard. But an instant solution means it could also be V2 standard so this is no general rule of thumb. Of 34 puzzles in my very hard file (rating 1.5 and up), 24 of them make SumoCue grunt.

3. If I can make lots of progress solving it manually, but can't finish it, it could be a really worthwhile V2 as is. It could be worth posting with that as an intro.

4. If I can't make much progress manually, I try looking for 2 digits that do not share any cage. With a complete swap of those two digits, the puzzle should still be unique (still worth checking of course), but could make it solvable manually.

5. Break a big cage in 2. This is very easy to do with Perfect Sudoku and get a new puzzle code. Then try and solve again. Sure, the cage pattern is a bit different, but there is no doubt what cage pattern inspired it.

6. This is a new one I'm planning on trying. From the ratings we have been developing, a general rule of thumb so far is that 100+ steps w/0 singles by Sudoku Solver is probably going to be a very tough puzzle.

It is very easy to get the steps from Sudoku Solver to run without accidentally getting any hints. I have the puzzle pic right off the screen and have a piece of card stuck over the steps so that all I can see is "naked/hidden single" (to get a step count) without knowing where that single is.

Of course, there are many great puzzles under 100 steps, so this is not an absolute. Again, for me, manual solving is the only way to really tell.

Time consuming? Yes! Quick? Not usually. Sometimes I will do numerous versions of a puzzle to get the one I want. Sometimes I have never posted any of those versions because they were not at the right level of difficulty. That's the risk.

I understand that this manual way may be impractical for you. If you want to use a solvers log to design/rate a puzzle, would you mind saying this in the intro to the puzzle?

Cheers
Ed

mhparker · Post by **mhparker** » Tue Aug 07, 2007 8:08 am

sudokuEd wrote:Wow Mike. What a puzzle is RP Lite! I've been trying and trying to get ready for this post - but still don't feel I've found a solution I'm happy with so have some more work to do yet.

Thanks, Ed. I'm glad you liked it. It's always nice to get some encouraging feedback and know that one's work is appreciated!

P.S.: BTW, if anyone is interested in how much manual work was involved in the production of the RP Lite, we're talking about roughly 8 hours. In other words, the equivalent of a full day at the office!

Andrew · Post by **Andrew** » Wed Aug 08, 2007 11:16 pm

Ruud wrote:The solving skills of the regular crowd has improved tremendously since I first posted these Assassins.

Definitely!

CathyW wrote:Did anyone else try Mike's A61X?

Andrew wrote:I started it last night but it was nearly bedtime so I didn't get much further than the preliminary steps and some easy 45s.

Must find time to work through the posted A60 walkthroughs and decide whether to post my one. Then I'll have a proper try at A61X.

That was written in the Assassin 61 thread on 1st August. I must admit I got sidetracked, first by A61X which I thought I was about to finish but then reached an impossible position (I've since managed to solve it. I'd eliminated a combination that contained the solution.

) and then by A62.

I eventually found time to work through Richard's and Mike's walkthroughs yesterday followed by Para's and Cathy's ones today. All contained excellent stuff and took different routes to reach the key eliminations from R3C7 and R7C3.

I'll have to make an effort to be first to post a walkthrough for some future Assassin. Then others will have to go through my walkthrough first before they post their ones!

I particularly liked Para's which was so direct. He said that he'd done a major rewrite after Word had crashed before he'd saved his walkthrough. Maybe it was a good thing that Word crashed on him!

I also lost my partial walkthrough but that was just carelessness on my part. However my reconstruction was pretty well how I had done it first time.

Mike (steps 21 and 22) and Cathy (step 37) had chains based on R3C7 (and R7C3) which were used to achieve results away from R3C7 although Cathy also got an elimination from R3C7 out of it. I hope that my step 28, which produced an elimination from R3C7 and R7C3 is acceptable as a contradiction move. Maybe in hindsight I should have continued looking for other moves before I used that one.

Anyway, having gone through the other walkthroughs, here is how I solved Assassin 60.

1. 6(3) cage in N1 = {123}, locked for N1

2. 22(3) cage in N1 = {589/679} = 9{58/67}, 9 locked for N1
2a. 45 rule on N1 3 innies R1C3 + R3C13 = 17 = 4{58/67}

3. 9(3) cage at R3C3 = {126/135/234}, no 7,8,9
3a. R3C3 = {456} -> no 4,5,6 in R3C4 + R4C3

4. 21(3) cage in N3 = {489/579/678}, no 1,2,3

5. 11(3) cage in N4 = {128/137/146/236/246}, no 9

6. 10(3) cage at R6C6 = {127/136/145/235}, no 8,9

7. 20(3) cage in N7 = {389/479/569/578}, no 1,2

8. 20(3) cage in N8 = {389/479/569/578}, no 1,2

9. 10(3) cage at R8C6 = {127/136/145/235}, no 8,9

10. 20(3) cage in N9 = {389/479/569/578}, no 1,2

11. 45 rule on N7 3 innies R789C3 = 9 = {126/135/234}, no 7,8,9

12. 45 rule on N3 3 innies R3C789 = 10 = {127/136/145/235}, no 8,9

13. 45 rule on N9 2 innies R7C9 + R9C7 = 9 = {27/36/45}/[81], no 1,9 in R7C9

14. 45 rule on R89 2 innies R8C27 = 11 = [29]/{38/47/56}, no 1,9 in R8C2

15. 45 rule on C89 2 innies R27C8 = 13 = {49/58/67}, no 3

16. 45 rule on R1 1 innie R1C7 – 2 = 2 outies R2C19, min R2C19 = 3 -> min R1C7 = 5, max R1C7 = 9 -> max R2C19 = 7 -> max R2C9 = 6

17. 45 rule on C1 2 outies R19C2 – 4 = 1 innie R7C1, max R19C2 = 12 -> max R7C1 = 8

18. 45 rule on N5689 2 innies R4C9 + R9C4 = 14 = {59/68/77}

19. 45 rule on N124 2 outies R3C7 + R7C3 = 4 = {13/22}
19a. Max R7C3 = 3 -> min R6C23 = 14, no 1,2,3,4

20. 45 rule on R6789 2 innies R6C15 = 7 = {16/25/34}, no 7,8,9
20a. Max R6C5 = 6 -> min R5C56 = 11, no 1

21. 45 rule on C6789 2 innies R15C6 = 13 = {49/58/67}, no 1,2,3

22. 45 rule on N7 1 outie R9C4 – 4 = 1 innie R7C3 -> no 8,9 in R9C4, clean-up: no 5,6 in R4C9 (step 18)

23. Killer triple 1/2/3 in R4C3 and R789C3 (step 11), locked for C3

24. Min R5C3 = 4 -> max R45C2 = 7, no 7,8

25. 45 rule on C89 4 outies R1278C7 = 28 = {4789/5689} = 89{47/56}, 8,9 locked for C7, no 3, clean-up: no 8 in R8C2 (step 14)

26. 45 rule on R12 4 outies R3C2567 = 24, max R3C7 = 3 -> min R3C256 = 21, no 1,2,3
26a. Min R3C5 = 4 -> max R2C45 = 9, no 9

27. R3C4 = {123} -> R3C789 (step 12) must contain two of 1,2,3 = {127/136/235} (cannot be {145}), no 4
[I suppose this is called a Hidden Killer Triple. I’m just used to writing the logic as I spot it.]

28. R3C7 + R7C3 (step 19) = {13/22}
28a. If R3C7 + R7C3 = {22} => R3C4 + R4C3 = {13} => R3C3 = 5 => R3C789 = {127} ({136} clashes with R3C4 = {13}) => R3C89 = {17} = 8 => R3C4 = 3, R3C2567 = 24 (step 26), R3C34 = [53] = 8 -> R3C1 = 5 clashes with R3C3 -> R3C7 + R7C3 cannot be {22}
28b. R3C7 + R7C3 = {13}, clean-up: no 6 in R9C4 (step 22), no 8 in R4C9 (step 18)
[Step 28a was originally a contradiction for the 4 outies from N4 but then I saw the contradiction along R3 which is more direct and easier to present.]

29. 16(3) cage at R3C8 = {169/259/367}
29a. {367} must have 7 in R4C9 -> no 7 in R3C89
29b. R3C789 = {136/235} = 3{16/25}, 3 locked for R3 and N3

30. 13(3) cage at R8C3 = {157/247/256} (cannot be {346} because no 3,4,6 in R9C4), no 3

31. R3C789 contains 1/2 (step 29b) -> 14(3) cage in N1 must contain 1/2 = {149/158/167/239/248/257}

32. R3C2567 (step 26) = {1689/3489/3579/3678} (cannot be {4569/4578} which don’t contain 1,3)

33. Killer triple 4/5/6 in R3C2567 (step 32), R3C3 and R3C89 (step 29b), locked for R3

34. 45 rule on R123 2 outies R4C39 – 4 = 1 innie R3C1, R3C1 = {78} -> R4C39 = 11,12 = {29/39} -> R4C3 = {23}, R4C9 = 9, R9C4 = 5 (step 18), R7C3 = 1 (step 22), R3C7 = 3 (step 28b), clean-up: no 4,6 in R7C9 (step 13)
34a. R7C3 = 1 -> R6C23 = 16 = {79}, locked for R6 and N4
34b. R9C4 = 5 -> R89C3 = 8 = {26}, locked for C3 and N7 -> R4C3 = 3

35. R3C789 = 3{16/25} (step 29b)
35a. R4C3 = 3 -> R3C34 = 6 = [42/51]
35b. R3C89 cannot be {25} which clashes with R3C34 -> R3C89 = {16}, locked for R3 and N3 -> R3C34 = [42] -> R1C3 + R3C1 = [58] (step 2a) -> R2C2 = 6 (hidden single in N1), R5C3 = 8
35c. R5C3 = 8 -> R45C2 = 3 = {12}, locked for C2 and N4 -> R1C2 = 3

36. 5 in R3 locked in R3C56, locked for N2

37. R3C7 = 3 -> R23C6 = 12 = [75] (only remaining permuation) -> R2C3 = 9, R3C25 = [79], R6C23 = [97]

38. R456C1 = {456}, locked for C1

39. 21(3) cage in N3 = {489} (only remaining combination because the remaining 7 is in the same cell as the 9) -> R1C7 = 9, R2C78 = {48}, locked for R2 and N3
39a. 14(3) cage in N3 = {257} -> R2C9 = 5, R1C89 = {27}, locked for R1 -> R12C1 = [12]

40. R5C6 = 9 (hidden single in C6) -> R56C5 = 8 = {26/35]/[71], no 4

41. 1 in N8 locked in R89C6, locked for C6, no 1 in R9C7
41a. 10(3) cage at R8C6 = 1{27/36}, no 4
41b. {127} has 7 in R9C7 -> no 2 in R9C7
41c. {136} has 6 in R9C7 -> no 6 in R89C6

42. Grouped X-Wing in 1, R3C89 = {16}, 1 in N9 locked in R89C89, 1 locked for C89
[Alternatively 1 in C7 locked in R456C7, locked for N6]

43. 20(3) cage in N9 = {479/569/578}
43a. {479} has 9 in R7C8 -> no 4 in R7C8
43b. {569} has 9 in R7C8 -> no 6 in R7C8

44. Killer pair 6/7 in 20(3) cage (step 43) and R9C7, locked for N9

45. 16(4) cage in N9 = {1249/1258/1348}
45a. Killer pair 8/9 in 16(4) and 20(3) cages in N9, locked for N9

46. 16(3) cage at R6C8 = {268/358} = 8{26/35}, no 4
46a. 8 locked in R6C89, locked for R6 and N6
46b. R7C9 = {23} -> no 2,3 in R6C89

47. 14(3) cage at R4C6 = 2{57}/6{17}8{15/24} -> no 4 in R4C6, no 6 in R45C7

48. 3 in N6 locked in R5C89, locked for R5
48a. 15(3) cage = {357} (only remaining combination), locked for N6
48b. 5 in N6 locked in R45C8, locked for C8

49. R6C89 = {68} -> R7C9 = 2 -> R1C89 = [27], R5C9 = 3
49a. Naked pair {68} in R6C89, locked for R6

50. 14(3) cage at R4C6 = {248} (only remaining combination) -> R4C6 = 8, R45C7 = {24}, locked for C7 -> R2C78 = [84], R6C7 = 1
50a. R6C7 = 1 -> R67C6 = 9 = [36] (only remaining permutation) -> R1C6 = 4, R6C4 = 4, R6C1 = 5, R6C5 = 2, R5C5 = 6 (cage sum), R1C45 = [68], R45C1 = [64], R4C5 = 5 (hidden single in N5) and the rest is naked singles and cage sums

sudokuEd · Post by **sudokuEd** » Fri Aug 10, 2007 1:48 am

mhparker wrote: manual work... involved in the production of the RP Lite.....roughly 8 hours.

We must be bonkers to make these versions.

Finally found a nice solution to RP Lite - after nearly 3 weeks. Congratulations to Para in getting it (and a walk-through!) soooo quickly.

Haven't had a chance to look at Cathy's way.

Para wrote:11. 10(3) at R1C1 can’t have .....{25} ..... in R12C1 because of 9(3) at R3C1

Para wrote:12. 6 only in R2C2 -->> R2C2: no 9

Missed these on early attempts which make a big difference. But my final solution, on the whole, mirrored Para's to step 24.

Finished it quite differently though. End of Para's step 24 here (first use of a bookmark from Sudoku Solver:

Paste these marks into Sudoku Solver

)

Code: Select all

-------------------------------.-------------------------------.-------------------------------.
| 2357      12        13        | 23456789  23456789  23456789  | 23456789  23456789  23456789  |
| 2357      4678      4789      | 123456789 123456789 1234567   | 123456789 123456789 123456789 |
| 56        47        89        | 89        12347     12347     | 12347     56        12347     |
&#58;-------------------------------+-------------------------------+-------------------------------&#58;
| 123       34578     6         | 2345789   2345789   3478      | 35789     1234      1234      |
| 123       34578     34578     | 23456789  23456789  23456789  | 356789    56789     1234      |
| 78        9         24        | 123467    23456     1234678   | 235678    235678    235678    |
&#58;-------------------------------+-------------------------------+-------------------------------&#58;
| 89        36        57        | 1234567   1234567   1234567   | 34589     34589     1267      |
| 4679      12        12        | 3456789   3456789   3456789   | 45        346789    346789    |
| 4679      3578      3578      | 123456789 123456789 123456789 | 1267      12346789  12346789  |
'-------------------------------.-------------------------------.-------------------------------'

Alternate step 25.
[1] in r7c9 -> 7 in r9c7 (h8(2)n9)-> 7 for n6 in 21(4) = [1]{578} only
[2] in r7c9 -> 6 in r9c7 (h8(2)n9) -> 6 in n6 in 21(4) = [2]{568} only
[6] in r7c9 -> 21 (4) = [6]{357} ([6]{258} blocked by clash with r6c5: i/o r6789)
[7] in r7c9 -> 21(4) = [7]{356} ([7]{239} clashes with 13(4) in n6)
-> no 2 r6c89, no 9 r5c8
-> 9 in n6 only in 20(3)
which leads to the solution very quickly through the h11(3) in n5, then the 11(3) cage at r6c6.

Now, onto Assassin 61.

Cheers
Ed