SudoCue Users

This one was a bit strange since there were many solving paths you could go, but I tried to keep it as short as possible. Overall, a rather easy assassin. I'm looking forward to V2.

1. C1234
a) Outies of C123 = 6(2) = {15/24}
b) 29(4) = {5789} locked for N5
c) Innies of C1234 = 11(2) = [56/74/83/92] -> R9C4 = (2346)
d) 11(2) @ N5: R7C4 = (5789)
e) 9(2) @ N5: R7C6 <> 1,2,4

2. C6789
a) 17(2) = {89} locked for C8 + N3
b) 8(3) = 1{25/34} -> 1 locked for C8
c) Innies C789 = 11(4) = {1235} -> locked for C6
d) 13(2) = {49/67}
e) Outies of C789 = 13(2) = {49/58/67}, R1C6 <> 4
f) Innies C6789 = 14(2) = {59/68}, R9C6 <> 8
g) Killer pair (58) of Innies C6789 blocks {58} of Outies of C789

3. C7
a) 21(3) must have 8 or 9 -> only possible @ R4C7
-> R4C7 = (89), 21(3) <> 4 ( {489} not possible )
b) R3C7 = (67) blocks {67} of 13(2)
c) 13(2) = {49} -> locked
d) R4C7 = 8
e) 21(3) = {678} -> 6,7 locked for N3 + R3
f) 15(2) @ C9: R7C9 <> 7
g) 15(4) <> 9 because 15(4) must have 6 xor 7 because of R5C7 = (67)
-> R45C8+R5C9 <> 6,7

4. C9
a) Innies+Outies: 5 = R9C8 - R5C9 -> R9C8 = (67), R5C9 = (12)
b) 4 locked in R123C9 @ 14(4)
c) 14(4) = 34{16/25} <> 7 -> other combinations blocked by R5C9 = (12)
d) 21(3): R89C9 <> 6 since R89C9 would be {68} -> blocked by killer pair (68) of 15(2)

5. N69
a) Innies+Outies: 1 = R8C6 - R4C9 -> R4C9 = (356)

6. C5
a) 6(3) = {123} locked

7. C6
a) Innies = 27(4) = 69{48/57} -> 6 locked
b) Killer pair (45) in 9(3) blocks [45] of 9(2)
c) 9(2) = [18/27]

8. C4
a) 6 locked in R46C4

9. C6 !
a) 16(3): R9C5 <> 6 since R9C6 has no 2,3,7,8 ( 6{28/37} )
b) 6 locked in R89C6
c) 6 locked in 19(3) @ C5 -> 19(3) <> 7
d) 13(3) = 1{39/57} -> no 2

10. N3
a) 1 locked in 13(3) for C7 + N3
b) 14(4) = {2345} -> locked for C9
c) R5C9 = 1

11. C9
a) 6 locked in 15(2) -> 15(2) = {69} locked
b) 21(3) = {678} locked for N9
c) R7C9 = 9 -> R6C9 = 6 -> R5C7 = 7 -> R3C7 = 6 -> R3C8 = 7 -> R9C8 = 6
d) R7C7 = 4 -> R6C7 = 9
e) Hidden Singles: R8C6 = 6 @ C6, R4C4 = 6 @ C4

12. N9
a) 11(3) = {236} -> 2,3 locked for C7 + N9
b) 8(3) = {125} locked in C8 -> R6C8 = 2 -> R6C6 = 1 -> R6C5 = 3 -> R6C4 = 4
c) R4C6 = 2, R7C4 = 7, R7C6 = 8

13. C456
a) Hidden Singles: R4C5 = 7 @ C5, R1C6 = 7 @ C6
b) 9(3) = {234} -> 3,4 locked for N2
c) 19(3) = {568} -> locked for C5 + N2

14. Rest is clean-up and singles

Again just my outline of solving path.


1. I/O N8 r8c4+r8c6=r6c456+3 > r8c46 min 9
2. r1c4+r8c4=6 <>3/3
3. r5c4+r9c4=11 no.1
4. r5c6+r9c5=14
5. I/O c9r9c8=r5c9+5 > r9c8=6/7 (8/9 blocked by r12c8=8/9) r5c9=1/2
6. r1c6+r4c79=20 ; r4c9 max 6 >r1c6 min =5
7. r67c6 9/2 cage<> 3/6 blocked by 9/3 cage r234c6
8. r459c5=20; r9c5<> 9

Consider position of 1 in c4

<> r23(blocked by 29/4 cage) <> r567 and <>r9 (see 3 above).
But 1 in r4c4 forces 1 into r78c5 and thus into r23c6 and 3 into r679c4/r678c5.This x-wing on 3 forces 3 into r23c 6 also.This places a 5 in r5c6 which is impossible.

Thus r18c4=1/5.
But,from 1. above a 1 in r8c4 > 8 in r8c6 > r6c456 =1/2/3.But a 3 cannot be placed in any of these cells (see 7. above)

Therefore r8c4=5 and r1c4=1

Hidden single in N5. A 6 must go in r4c4

In N8 r9c4= 2/3/4 (see 3. above)
But if a 2 cannot place a 9 in N8.
If r9c4=3 r9c56=49 (6/7 blocked by r9c8)
If r9c4=4 r9c56=39 but blocked by 6/3 cage
Therefore r9c456=349.

Just a mop up now.


A nice puzzle.

Regards

Gary

P.S. Did killersudokuonline number 90 in about 1 hr. Definitely easier than any assassins.

Hi all,

Not much discussion going on here! Where are all the regulars (Cathy?, Para?, Andrew?, ...)? So think I'll liven things up by adding my two cents worth:

Afmob wrote:This one was a bit strange since there were many solving paths you could go... Overall, a rather easy assassin.

Rating? 0.75? It was probably significantly easier than Ruud intended it to be for the following reason:

There's an early key move, which Afmob used, namely the C789 innies = 11(4) = {1235}, which effectively cracked the puzzle. Those who saw this move early on will probably have found the puzzle considerably easier than those who didn't. Interestingly, however, neither Sumocue nor JSudoku (in the default state, with all solvers enabled) pick up this move. Presumably, quadruple innies and outies are relatively low down in the solving technique pecking order, so they found the less productive but "simpler" moves first, causing them to make heavier going of the puzzle than necessary. Therefore, for this reason, it's possible that this move escaped Ruud's attention.

gary w wrote:Again just my outline of solving path.

Thanks, Gary. BTW, you chose the harder route in comparison to Afmob. However, you had some good ideas there! Here are a few comments you might find interesting:

Your move involving the " x-wing on 3" is something I would have expressed in a simpler form, as follows:

9. R23C6 cannot contain both of {13}, due to 5 being unavailable in R2C6
9a. The only other place for {13} in N2 is R123C4
9b. -> either R1C4 = 1, or...
9c. ...17(3) cage contains a 3 within R23C4 (or both)
9d. Either way, R4C4 cannot contain a 1

Also, I couldn't follow your contradiction: "But a 3 cannot be placed in any of these cells". I still had at least one 3 in R6C456 at this point. However, this contradiction could be avoided completely by eliminating the 8 from R8C6 in a simpler (separate) step. For example:

10. 1 in C6 locked in 9(3) cage at R234C6 or 9(2) cage at R67C6
10a. either 1 in 9(3) cage, then 9(3) = {135} ({126} blocked because 3 in C6 locked in this cage) -> R23C6 must contain a 5...
10b. ...or 1 is in 9(2) cage = [18] -> R7C6 = 8
10c. -> R237C6 must contain one of {58}
10d. -> {58} combo blocked for C789 outies at R18C6
10e. -> no 5,8 in R18C6

I also liked your logic in rejecting the [286] permutation for the 16(3) cage at R9C456. As mentioned several times before, I like to express such contradictions as closed loops, VIZ:

12. 17(3) cage at R9C456 = {2..} -> R9C4 = 2 -> R67C4 <> [29] -> R9C6 = 9 -> 17(3) cage at R9C456 = {9..}
12a. i.e., if 17(3) cage at R9C456 contains a 2, it must also contain a 9
12b. -> [286] combination/permutation blocked.

Afmob wrote:I'm looking forward to V2.

Me, too! Hopefully, Ruud will post it soon.

mhparker wrote:Hi all,

Not much discussion going on here! Where are all the regulars (Cathy?, Para?, Andrew?, ...)? So think I'll liven things up by adding my two cents worth:

Afmob wrote:This one was a bit strange since there were many solving paths you could go... Overall, a rather easy assassin.

mhparker wrote:Rating? 0.75? It was probably significantly easier than Ruud intended it to be for the following reason:

I wasn't going to post anything until I finished. At the moment I'm finding it at least as hard as A69. I'm currently on step 40, including preliminaries, and haven't yet fixed a cell although I feel that I'm fairly close to doing so. I must be missing something important if Mike rates it at 0.75.

[Edit.] That should have been "currently on step 49". However when I fixed a cell in step 50 I quickly found that I had reached an impossible position so I probably made a mistake earlier. Start again!

[Edit 2.] I discovered later, when I went through my steps again, that I hadn't reached an impossible position. I'd just been careless and eliminated too many candidates. After that it came out quickly.

Afmob wrote:I'm looking forward to V2.

mhparker wrote:Me, too! Hopefully, Ruud will post it soon.

Cant' keep you waiting any longer.



3x3::k:58884866:4866:435638454872:58884866:5132:435638454872:58882579:5132:43564888:4888:4872:5888:66845132:438348884872:6684:6684:6684:4383:4383:438336182349:5166:81536332355:51722349:5166:81536332355:51722367:5166:4929:49292884517223674929:5451:5451:54513911

The background image symbolizes the long series of moves that the puzzle demands. Unlike the fractal, it is finite.

Enjoy,
Ruud

I am around - just haven't had time to do the V1 yet, never mind the V2. Installing new computers at work so had no internet access on Friday and spent over 2 hours yesterday with my DH trying to resolve networking problems! Computers are wonderful things but they can sure be frustrating too.

Hopefully will get to the A70 some time today.

Edit: Have solved but didn't keep a WT. Key moves were having done the outies of c789 = 13 -> innies = 11 {1235} and conflicting combos in c6. I don't suppose the V2 will be so helpful in opening up the puzzle.

Since my first walkthrough was wrong I'll have to revise it. From my point of view A70 V2 is more difficult than A70 V3.

Edit: No walkthrough from me on this since Para has found a better solving path with less hypothetical moves.

Assassin 70 v2
Back to the drawing board for me too.
Have rechecked my solution for v3.Stands up so certainly,I believe,v2 is the tricky one.What is remarkable is that I made the same logical mistake as Afmob..can't really see why this should have been so compelling??

Gary

Well done Afmob and Gary for finishing A70V2 and so quickly! I'm still stuck - so must be missing something. You've spurred me on.

In the mean-time, here's an old fashioned V3 for you guys. It has the same solution as the original - you just have to work a bit harder to get it. I hope it closes up the way you solved the others .

I haven't been able to solve this one either - but have tried very hard. Sudoku Solver and JSudoku don't have any problems with it though I have no idea how they do it. But I'm guessing its a 1.75 rating - which my rule-of-thumb means it takes my leisurely meanderings at least 10 hours to find a nice solution. I used to not like to post these until I'd solved them - but more competition now .

Cheers
Ed

Assassin 70 V3 (est. rating 1.75) NOTE: it has 1 remote cage 14(3) at r4c258

3x3::k:4864281848683333:435948642818:4364:48683333:435948643859:4364:48685400:540048643859:43642318:5400359228525671:5671:567131142349:46542864235421002349:46542864235421003903:4654:4417:441728842100:54473903:4417:4171:4171:41715447:5447:

I think I'll take a break or else I get addicted . I'll wait until the regulars post their walkthrough for A70 and A70 V2 and see if they took a different approach to solve the assassin.

I think the remote cage in V3 should raise the difficulty a notch but maybe I'm mistaken and it reveals another weakspot .

sudokuEd wrote:I used to not like to post these until I'd solved them - but more competition now .

Gotta stick to your convictions, Ed I hope it's better than when you think you solved it, but made a flaw somewhere and end up with one of those unsolvable puzzles. By the way, Sudoku Solver and JSudoku having no tourble with it doesn't mean much. One of my V2's is still unsolved in the unsolvables section and neither solver had too much trouble with it.
Oh, btw. Love your doing an "old school" V2 again.

What about posting a walk-through for Assassin 70 instead of sticking to your convictions?

Sorry there's no walk-through coming for me this week or next week either(i'll be in Brazil then). I solved the original and did some work on Ruud's V2, but didn't find anything to crack it yet.

greetings

Para

ps. Afmob: there's nothing shameful in getting addicted to these(right, guys??). We have a rehab program called the unsolvables. Just start tackling those and at one point you'll really want a break from Killers.

deleted,,see below

Para wrote:Gotta stick to your convictions, Ed

You've hit my conscience hard Para - thanks. As punishment, I won't publish any more not-solved-first puzzles until (at least) a start has been made on all the unsolvables . Last week, I did start A39V3 - so will post that this week.

Para wrote:when you think you solved it, but made a flaw somewhere and end up with one of those unsolvable puzzles

Think this must be A41V2 here. What a blunder that turned out to be! Got there in the end though.

Para wrote:By the way, Sudoku Solver and JSudoku having no tourble with it doesn't mean much. One of my V2's is still unsolved in the unsolvables section and neither solver had too much trouble with it.

Assassin 44V1.5 I think. Don't know that anyone really gave it a serious go. Next week.

Para wrote:Oh, btw. Love your doing an "old school" V2 again

Thanks! I really enjoyed it after such a long break. I must confess I started it Friday night when PS 4 gave A70 an IQ rating. Don't think we've seen that since A1!

Para wrote: What about posting a walk-through for Assassin 70

More penance . Definitely only a 0.75 rating - maybe even 0.5 - except I used 1 little contradiction move. Maybe that speeded things up.

gary w wrote:Actually found a70v3 easier than version 1

Wow - we must solve in very different ways! I haven't looked at any of your outlines yet, or Afmod's walk-through. But am very curious since I'm still stuck on both V2 & V3. Later.

Cheers
Ed
(ps Please let me know of any corrections for the WT [edit: thanks Andrew])

Walk-through for Assassin 70

Prelims
i. 11(2)n1: no 1
ii. 11(3)n1 & n8: no 9
iii. 19(3)n2: no 1
iv. 17(2)n3 = {89}
v. 14(4)n3: no 9
vi. 9(3)n2: no 789
viii. 21(3)n3:no 123
ix. 29(4)n5 = {5789}
x. 9(2)n4 & n5: no 9
xi. 11(2)n4 & n5: no 1
xii. 6(3)n5 = {123}
xiii. 13(2)n6: no 123
xiv. 8(3)n6 = {125/134}
xv. 15(2)n6 = {69/78}
xvi. 21(3)n9 = {489/579/678}

1. 17(2)n3 = {89}: both locked for n3 & c8

2. "45"c9: r9c8 - 5 = r5c9
2a. r9c8 = {67}
2b. r5c9 = {12}

3. 21(3)n9 must have 6/7 = {579/678}(no 4)
3a. = 7{..}: 7 locked for n9
3b. no 8 r6c9
3c. no 6 r6c7

4. 14(4)n3 = {1346/2345}(no 7,8) (combo's with {12..} blocked by r5c9)

5. 8 in c9 only in n9: 8 locked for n9
5a. no 5 r6c7

6. "45" c789: 4 innies r1289c7 = h11(4) = {1235}: all locked for c7
6aa. forgot some clean-up: -> no 8 r6c7

6a. "45"c789:2 outies r18c6 = h13(2)
6b. = {49/58/67}(no 1,2,3)
6c. max r8c6 = 8 -> min r1c6 = 5

7. "45"c6: r59c6 = h14(2) = {59}/[86] = [5/8,6/9..]
7a. r5c6 = {589}
7b. r9c6 = {569}

8. h13(2)r18c6 = [94]/{67} = [6/9..]({58} clashes with r59c6)

9. Killer pair [69] in h14(2) & h13(2): both locked for c6

10. 9(3)n2 = {135/234} = 3[4/5..]
10a. 3 locked for c6
10b. -> 9(2) = {18/27}(no 45: blocked by 9(3))
[edit thanks Andrew]

11.29(4)n5 = {5789} all locked for n5
11a. no 12 in r7c6
11b. no 2346 in r7c4

12. 6(3)n5 = {123}: all locked for c5
12a. 19(3)n2 = {469/478/568}
12b. r45c5 must have 2 of {5789} -> 9 locked in 19(3) or r45c5
12c. no 9 r9c5

13. "45"c123: r18c4 = h6(2) = {15/24}

14. "45" c4: r59c4 = h11(2)
14a. = [56/74/83/92]
14b. r9c4 = {2346}

15. "45"c89: r3c8 = r5c7 = {467}

16. 21(3)n3 must have 2 of {467}
16a. = {678} only
16b. r5c7 = {67} (step 15)

17. r3c78 = {67}: both locked for r3 & n3

18. r4c7 = 8

19. r39c8 & r35c7 = {67}: both locked for c8 & c7

20. 8 in n5 only in r5: 8 locked for r5

21. "45" n69: r8c6 - 1 = r4c9
21a. r8c6 = {467} -> r4c9 = {356}

22. when 14(4)n3 = {1346} -> 6 must be in r4c9 -> 7 in r5c7 -> r67c9 = [96]; but this means 2 6's in c9
22a. 14(4) = {2345}(no 1,6): all locked for c9
22b. no 6 in r4c9 -> no 7 in r8c6 (step 21)
22c. -> no 6 in r1c6 (h13(2)r18c6

23. r5c9 = 1, r9c8 = 6 (i/oc9)

24. r3c78 = [67], r5c7 = 7

25. r89c9 = {78}(last combo): both locked for c9

26. r67c9 = [69], r67c7 = [94]

27. h14(2)r59c6 = {59}(last combo): both locked for c6

28. h13(2)r18c6 = [76]

29. r12c7 = {15}(last combo): both locked for c7 & n3

30. "45" n3: r4c9 = 5

31. "45" n6 1 innie r6c8 = 2


All simple from here.

Andrew wrote:I wasn't going to post anything until I finished. At the moment I'm finding it at least as hard as A69. I'm currently on step 40, including preliminaries, and haven't yet fixed a cell although I feel that I'm fairly close to doing so. I must be missing something important if Mike rates it at 0.75.

Finished it now and yes I did miss something important. I missed the key move 4 innies in C789! I don't know how I missed that because columns are clearly so important in the early stages. All I can think is that I may have looked at 4 innies in C123, which were no help, and forgotten to look at C789. Ed suggested to me, off-forum, that the 4 innies in C789 are hard to see because the 2 outies from C789 are so obvious. Maybe that was it.

mhparker wrote:There's an early key move, which Afmob used, namely the C789 innies = 11(4) = {1235}, which effectively cracked the puzzle. Those who saw this move early on will probably have found the puzzle considerably easier than those who didn't.

And it's harder still if you miss it completely. However it still comes out and would only be rated between 1.0 and 1.25 that way.

Afmob wrote:I'll wait until the regulars post their walkthrough for A70 and A70 V2 and see if they took a different approach to solve the assassin.

Sorry to disappoint you but I won't be posting my walkthrough for A70 for the reasons given above. I haven't looked at A70 V2 yet. Maybe the way that I solved A70 will help me with V2 if I try it?

An excellent solving path by Afmob!

I also looked at Gary's outline. Some interesting ideas there!