The New Sudoku Players' Forum

by **eleven** » Tue Jun 25, 2019 8:08 pm

Code: Select all: +-------+-------+-------+ | . . . | . 9 . | . . . | | . . . | 2 3 8 | . . . | | . . . | 1 . 4 | . . . | +-------+-------+-------+ | . . 9 | 3 . 5 | 1 . . | | . . 7 | . . . | 6 . . | | . 6 2 | . . . | 5 . . | +-------+-------+-------+ | . 9 . | . . . | . 1 . | | 4 3 . | 8 . 1 | . 7 9 | | 2 . 1 | . . . | . . 3 | +-------+-------+-------+

by **SpAce** » Tue Jun 25, 2019 9:40 pm

Code: Select all: .------------------.-------------.-------------------. | b17 28 b34 | 5 9 6 | b37 48 bc12 | | 69 17 cd4(5) | 2 3 8 | 79 456 16-5 | | 69 28 35 | 1 7 4 | 39 568 d(2)6-5 | :------------------+-------------+-------------------: | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 13 6 2 | 479 18 79 | 5 39 48 | :------------------+-------------+-------------------: | a57 9 8 | 67 2 3 | 4 1 a[5]6 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | '------------------'-------------'-------------------'

(57)r7c91 = (7134)r1c1973 - (2|4)r1c9,r2c3 = (25)r3c9,r2c3 => -5 r23c9; stte

3D Medusa: Show

by **Leren** » Tue Jun 25, 2019 10:36 pm

Code: Select all: *--------------------------------------------* | 17C a8-2aA 34 | 5 9 6 | 37 b48 12B | | 69 17 d45 | 2 3 8 | 79 c456 e156 | | 69 28b 35 | 1 7 4 | 39 568 256c | |----------------+-----------+---------------| | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 13 6 2 | 479 18 79 | 5 39 48 | |----------------+-----------+---------------| | 57D 9 8 | 67 2 3 | 4 1 56E | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | *--------------------------------------------* Kraken Column 9 Digit 5: 2 r1c2 - 8 r1c2 = (8-4) r1c8 = r2c8 - (4=5) r2c3 - 5 r2c9; 2 r1c2 - r3c2 = 2 r3c9 - 5 r3c9; 2 r1c2 - (2=1) r1c9 - (1=7) r1c1 - (7=5) r7c1 - 5 r7c9; => - 2 r1c2; stte

Leren

by **SteveG48** » Wed Jun 26, 2019 12:34 am

Code: Select all: *---------------------------------------------------* |b1-7 28 b34 | 5 9 6 |ab37 48 12 | | 69 c17 c45 | 2 3 8 | 79 456 c156 | | 69 28 35 | 1 7 4 | 39 568 256 | *----------------+----------------+-----------------| | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 13 6 2 | 479 18 79 | 5 39 48 | *----------------+----------------+-----------------| |d57 9 8 | 67 2 3 | 4 1 d56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | *---------------------------------------------------*

7r1c7 = (34)r1c37&7r1c1 - (4|7=156)r2c239 - (6=57)r7c19 => -7 r1c1 ; stte

by **Cenoman** » Wed Jun 26, 2019 10:22 am

Code: Select all: +------------------+------------------+-------------------+ | 1-7 28 B34 | 5 9 6 | B37 48 12 | | 69 b17 B45 | 2 3 8 | 79 456 Aa156y | | 69 28 35 | 1 7 4 | 39 568 256 | +------------------+------------------+-------------------+ | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 13 6 2 | 479 18 79 | 5 39 48 | +------------------+------------------+-------------------+ | 57z 9 8 | 67 2 3 | 4 1 56z | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | +------------------+------------------+-------------------+

Death blossom, stem (156)r2c9
(1)r2c9-(1=7)r2c2
(5)r2c9-(543=7)r12c3,r1c7
(6)r2c9-(65=7)r7c19
=>-7r1c1; ste

by **Cenoman** » Wed Jun 26, 2019 10:36 am

Code: Select all: +-------+-------+-------+ | . . . | . 9 . | . . . | | . . . | 2 3 8 | . . . | | . . . | 1 . 4 | . . . | +-------+-------+-------+ | . . 9 | 3 . 5 | 1 . . | | . . 7 | . . . | 6 . . | | . 6 2 | . . . | 5 . . | +-------+-------+-------+ | . 9 . | . . . | . 1 . | | 4 3 . | 8 . 1 | . 7 9 | | 2 . 1 | . . . | . . 3 | +-------+-------+-------+

eleven's pattern made me think of the Eiffel Tower (well, say almost...) Guess why ?

The genuine Eiffel Tower is symmetric. The pattern is not, because of 6r7c2 and 1r9c3.
So I propose an additional challenge to eleven's:

Code: Select all: +-------+-------+-------+ | . . . | . 9 . | . . . | | . . . | 2 3 8 | . . . | | . . . | 1 . 4 | . . . | +-------+-------+-------+ | . . 9 | 3 6 5 | 1 . . | | . . 7 | . . . | 6 . . | | . . 2 | . . . | 5 . . | +-------+-------+-------+ | . 9 . | . . . | . 1 . | | 4 3 . | 8 . 1 | . 7 9 | | 2 . . | . . . | . . 3 | +-------+-------+-------+

Note: a one-stepper is a bit challenging (not impossible though). Multi-steppers accepted.

FWIW, my own solutions:

Hidden Text: Show

Code: Select all: +-----------------------+------------------+-------------------+ | c17 28 34-1 | 5 9 6 | 37 48 12 | | 15679 1567 45-1 | 2 3 8 | 79 456 156 | | 569 28 35 | 1 7 4 | 39 568 256 | +-----------------------+------------------+-------------------+ | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | +-----------------------+------------------+-------------------+ | b57 9 8 | 67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 157 a15 | 679 4 79 | 8 56 3 | +-----------------------+------------------+-------------------+

Krakenless 3-steps:
1. XY-wing (1=5)r9c3-(5=7)r7c1-(7=1)r1c1 => -1 r12c3; 1 placement (+1r9c3) & basics

Code: Select all: +--------------------+------------------+-------------------+ |Bc17 28 B34 | 5 9 6 | B37 48 b2-1* | | 679-1* 167 A45 | 2 3 8 | 79 456 Aa15-6 | | 69 28 35 | 1 7 4 | 39 568 256 | +--------------------+------------------+-------------------+ | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | +--------------------+------------------+-------------------+ | c57 9 8 | 67 2 3 | 4 1 c56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | +--------------------+------------------+-------------------+

2. (1)r2c9 = r1c9 - (157=6)r17c1,r7c9 => -6 r2c9
3. (15=4)r2c39 - (437=1)rc137 => -1*r1c9,r2c1; ste

...or one-stepper:

Code: Select all: +-----------------------+------------------+-------------------+ | 17 28 134 | 5 9 6 | 37 48 12 | | 15679 1567 145 | 2 3 8 | 79 456 156 | | 569 28 35 | 1 7 4 | 39 568 256 | +-----------------------+------------------+-------------------+ | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | +-----------------------+------------------+-------------------+ | 57 9 8 | 67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 157 15 | 679 4 79 | 8 56 3 | +-----------------------+------------------+-------------------+ Double kraken (5)r3 & (1)r1 (5)r3c1 - (5=7)r7c1* || (5-3)r3c3 = 3r3c7 - (3=7)r1c7* || (5)r3c8 - r9c8 = r7c9 - (5=7)r7c1* || (5-2)r3c9 = (2-1)r1c9 || (1)r1c1* || (1)r1c3 - (15=7)b7p19* =>-7r1c1; ste

With a rather simple TM 8x8

Code: Select all: 7r7c1 5r7c1 7r1c7 3r1c7 3r3c7 3r3c3 5r7c9 5r9c8 5r9c3 1r9c3 1r1c1 1r1c3 1r1c9 2r1c9 2r3c9 5r3c1 5r3c3 5r3c8 5r3c9 =>-7r1c1

by **Leren** » Wed Jun 26, 2019 11:29 am

Code: Select all: *----------------------------------------------------* |Bb17bB 28 F134 | 5 9 6 |Aa7-3axA 48 G12 | | 15679 1567 145 | 2 3 8 | 79 456 156 | | 569d 28 35z | 1 7 4 | 39y 568F 256G | |-------------------+-----------+--------------------| | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | |-------------------+-----------+--------------------| | C57cC 9 8 | 67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 D157D E15D | 679 4 79 | 8 56E 3 | *----------------------------------------------------* Kraken Column 9 Digit 2: 3r1c7 - 7r1c7 = (7-1) r1c1 3r1c7 - 7r1c7 = 7r1c1 - 7r7c1 = (7-1) r9c2 = 1r9c3 - 1r1c3 *= 1r1c9 - 2r1c9; 3r1c7 - 7r1c7 = 7r1c1 - (7=5) r7c1 - 5r3c1; 3r1c7 - 3r3c7 = (3-5) r3c3; 3r1c7 - 7r1c7 = 7r1c1 - (7=5) r7c1 - 5r9c23 = 5r9c8 - 5r3c8 *= 5r3c9 - 2r3c9; => - 3 r1c7; stte

Leren

by **SpAce** » Wed Jun 26, 2019 11:51 am

Code: Select all: .----------------------.-------------.---------------. | a17 28 134 | 5 9 6 | b37 48 12 | | 1679-5 1567 145 | 2 3 8 | 79 456 156 | | 69-5 28 c3(5) | 1 7 4 | b39 568 256 | :----------------------+-------------+---------------: | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | :----------------------+-------------+---------------: | a[5]7 9 8 | 67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 157 1-5 | 679 4 79 | 8 56 3 | '----------------------'-------------'---------------'

Step 1: (57)r71c1 = (73)r13c7 - (3=5)r3c3 => -5 r23c1,r9c3

Added. A simpler alternative:

Step 1: (5=71)r71c1 - r12c3 = (1)r9c3 => -5 r9c3

Code: Select all: .-------------------.-------------.--------------------. | c17 28 B(3)4 | 5 9 6 | d(7)-3 B48 12 | | 1679 167 45 | 2 3 8 | 79 A456 a156 | | x69 28 35 | 1 7 4 | x(3)9 x568 a256 | :-------------------+-------------+--------------------: | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | :-------------------+-------------+--------------------: | b57 9 8 | 67 2 3 | 4 1 a56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | '-------------------'-------------'--------------------'

Step 2: Kraken 6B3

(6-4)r2c8 = (43)r1c837
||
(6,5)r237c9 - (5=7)r7c1 - r1c1 = (7)r1c7
||
(693)r3c817

=> -3 r1c7; stte

original step 2: Show

by **SpAce** » Wed Jun 26, 2019 1:23 pm

eleven, related to your frequent complaints about the clarity of my chains, most recently here, I'm interested in what you think of this "kraken":

Leren wrote:
Code: Select all
Kraken Column 9 Digit 2: 3r1c7 - 7r1c7 = (7-1) r1c1 3r1c7 - 7r1c7 = 7r1c1 - 7r7c1 = (7-1) r9c2 = 1r9c3 - 1r1c3 *= 1r1c9 - 2r1c9; 3r1c7 - 7r1c7 = 7r1c1 - (7=5) r7c1 - 5r3c1; 3r1c7 - 3r3c7 = (3-5) r3c3; 3r1c7 - 7r1c7 = 7r1c1 - (7=5) r7c1 - 5r9c23 = 5r9c8 - 5r3c8 *= 5r3c9 - 2r3c9; => - 3 r1c7; stte

No offense, Leren, but I think it's the ugliest mf I've seen in a long time

To be honest, I have much less interest in deciphering that than eleven's obfuscated C example which at least had the benefit of extreme brevity. Of course I can see how it works, but the way it's written and what it's called aren't in sync at all. It's a Nishio (contradiction net), not Kraken, and not intuitively followable even as such.

Since you're most likely generating your solutions with software in the first place, isn't it a piece of cake to change the way it outputs kraken solutions so they would have the normal left-to-right orientation? Of course, this last one would require quite a bit more than a simple reversal to make it readable, but at least it would fix the constant problem with your normal krakens. For special cases like this it shouldn't be too hard to put in some manual labor for the presentation.

Sorry about that, but I find it kind of funny that my chains are singled out as being "obfuscated"

by **SteveG48** » Wed Jun 26, 2019 2:22 pm

Code: Select all: *---------------------------------------------------------------------* | a17 28 c134 | 5 9 6 |a37 d48 12 | | 15679 1567 145 | 2 3 8 | 79 456 156 | | b569 28 35 | 1 7 4 |a39 d568 256 | *-----------------------+----------------------+----------------------| | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | *-----------------------+----------------------+----------------------| |af57 9 8 | 67 2 3 | 4 1 f56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 157 b15 | 679 4 79 | 8 e56 3 | *---------------------------------------------------------------------*

(7=93*)r13c7&(75)r17c1 - (5|9=16**)r3c1,r9c3 - (1|*3=4)r1c3 - (4|**6=58)r13c8 - (5=6)r9c8 - (6=57)r7c19 = -7 r1c1 ; stte

by **SpAce** » Wed Jun 26, 2019 8:58 pm

SteveG48 wrote:(7=93*)r13c7&(75)r17c1 - (5|9=16**)r3c1,r9c3 - (1|*3=4)r1c3 - (4|**6=58)r13c8 - (5=6)r9c8 - (6=57)r7c19 = -7 r1c1 ; stte

Hi Steve! That's a nice one! I think you should put the (7)r1c7 into its own term, though:

(7)r1c7 = (93*)r13c7&(75)r17c1 ...

It's also possible to simplify the last two terms a bit (no need to involve the 6):

... - (5=6)r9c8 - (6=57)r7c19 <-> ... - r9c8 = (57)r7c91

PS. The chain also proves +7 r1c7 directly via contradiction (two 5s on in r7). (Just an observation.)

by **eleven** » Wed Jun 26, 2019 10:09 pm

Sorry, have no time now for the new puzzle.

The other one can be solved without pencilmarks this way:

Code: Select all: +-------+-------+-------+ | . . . | 5 9 6 | . . . | | . . . | 2 3 8 | . . . | | . . . | 1 7 4 | . . . | +-------+-------+-------+ | 8 4 9 | 3 6 5 | 1 2 7 | | . X 7 | . . 2 | 6 . . | <--- | . 6 2 | . . X | 5 . . | <--- +-------+-------+-------+ | . 9 8 | . 2 3 | 4 1 . | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 * 1 | . 4 * | 8 . 3 | +-------+-------+-------+

5r5c2 & 7r6c6 is deadly, reverse BUG lite. (In a solution you could swap the other digits of the 2 rows to get a second solution)
Since 5r5c2 implies 7r6c6 (5r5c2 -> 7r9c2 -> 7r6c6), it must be false.

With pencilmarks:

Code: Select all: *---------------------------------------------------------* | 17 28 34 | 5 9 6 | 37 48 12 | | 69 17 45 | 2 3 8 | 79 456 156 | | 69 28 35 | 1 7 4 | 39 568 256 | |------------------+------------------+-------------------| | 8 4 9 | 3 6 5 | 1 2 7 | | #13+5 15 7 | #49 #18 2 | 6 #39 #48 | | #13 6 2 | #49+7 #18 79 | 5 #39 #48 | |------------------+------------------+-------------------| | b57 9 8 | a67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 57 1 | 679 4 79 | 8 56 3 | *---------------------------------------------------------*

BUG lite r56c14589: 5r5c1=7r6c4
5r5c1 == 7r6c4 - r7c4 = 7r7c1 => -5r7c1, stte

fyi, the BUG-lite pattern came accidentally, just looked for a triangular shape.

by **SteveG48** » Wed Jun 26, 2019 10:22 pm

SpAce wrote:
SteveG48 wrote:(7=93*)r13c7&(75)r17c1 - (5|9=16**)r3c1,r9c3 - (1|*3=4)r1c3 - (4|**6=58)r13c8 - (5=6)r9c8 - (6=57)r7c19 = -7 r1c1 ; stte

Hi Steve! That's a nice one! I think you should put the (7)r1c7 into its own term, though:

(7)r1c7 = (93*)r13c7&(75)r17c1 …

Agreed. Using the *s must have affected my brain. Starstruck?

It's also possible to simplify the last two terms a bit (no need to involve the 6):

... - (5=6)r9c8 - (6=57)r7c19 <-> ... - r9c8 = (57)r7c91

PS. The chain also proves +7 r1c7 directly via contradiction (two 5s on in r7). (Just an observation.)

True.

PS: I enjoy your chains. And Leren's too.

by **SpAce** » Thu Jun 27, 2019 5:33 pm

SteveG48 wrote:Using the *s must have affected my brain. Starstruck?

That must be it!

PS: I enjoy your chains. And Leren's too.

Thanks. It goes without saying that I enjoy yours too, and most of the others as well. Naturally some are more eye-pleasing or interesting than others, but that's mostly subjective, and all of them serve a purpose by displaying different approaches and POVs. Most importantly, everyone should be allowed to enjoy their own and learn from others' solutions. As long as they're more or less correct and understandable, I don't see a reason to force everyone into one narrow mold. Even truly bad examples provide contrast and learning opportunities.

That's why I don't really care how Leren prints his krakens. If I think they're too much of an eye-sore, I can choose to ignore them (which is exactly what I've advised to those who don't like my chains). I am, however, interested in why eleven thinks only my chains deserve constant criticism. Does he really see them as the worst ever, without any redeeming qualities? I've tried to ask this more or less directly a couple of times now, but no answer so far. I'm not allergic to criticism at all, but if I'm the only one getting it, it starts to feel a bit like a double-standard to me.

by **eleven** » Thu Jun 27, 2019 6:40 pm

Hi SpAce,
i just critisized, that by your abbreviation very useful information (for quick understanding), was hidden.
To be honest, i did not read Leren's 2nd kraken to the end.

I had a look at Cenoman's puzzle with the nice pattern now.

In both grids there's a complex invalid pattern:

Code: Select all: *-----------------------------------------------------* | . . 34 | . . . | . 48 . | | . . 45 | . . . | . 46 . | | 69 28 35 | . . . | 39 68 25 | *--------------------+--------------+-----------------*

(row 3 gives the links 3r3c3=6r3c8 and 5r3c3=8r3c8, so 5r3c3 -> 6r3c8 -> 4r2c8 -> 5r2c3 and 3r3c3 -> 3r1c3)

For Cenoman's:

Code: Select all: *-----------------------------------------------------* | 17 28 *34+1 | 5 9 6 | 37 *48 12 | | 15679 1567 *45+1 | 2 3 8 | 79 *46+5 156 | | *69/5 *28 *35 | 1 7 4 | 39 *68+5 *25/6 | |--------------------+--------------+-----------------| | 8 4 9 | 3 6 5 | 1 2 7 | | 135 15 7 | 49 18 2 | 6 39 48 | | 136 16 2 | 479 18 79 | 5 39 48 | |--------------------+--------------+-----------------| | 57 9 8 | 67 2 3 | 4 1 56 | | 4 3 6 | 8 5 1 | 2 7 9 | | 2 157 a15 | 679 4 79 | 8 6-5 3 | *-----------------------------------------------------*

5r23c8 == 1r12c3 - (1=5)r9c3 => -5r9c8, stte

The New Sudoku Players' Forum

leftover

leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover

Re: leftover