View previous topic :: View next topic |
Author |
Message |
daj95376
Joined: 23 Aug 2008 Posts: 3854
|
Posted: Thu Jul 21, 2011 3:05 pm Post subject: Puzzle 11/07/21: ~ Advanced |
|
|
After so many singles, you'd think the puzzle wouldn't need much to crack it. Maybe you're right.
Code: | +-----------------------+
| 8 . . | 9 . 5 | . . 6 |
| . . 9 | . 8 6 | . . . |
| . 6 . | . . 2 | 9 . . |
|-------+-------+-------|
| 7 . . | 3 5 4 | . 9 2 |
| . 4 . | 2 . 9 | . . . |
| 2 9 5 | 8 7 . | . . 4 |
|-------+-------+-------|
| . . 2 | . . . | 4 . 7 |
| . . . | 4 . . | . 5 . |
| 4 . . | 5 . 8 | 6 . 9 |
+-----------------------+
|
Play this puzzle online at the Daily Sudoku site
For those who don't need experience finding Singles:
Code: | +--------------------------------------------------------------+
| 8 137 1347 | 9 34 5 | 17 2 6 |
| 13 2 9 | 17 8 6 | 157 4 135 |
| 5 6 1347 | 17 34 2 | 9 137 8 |
|--------------------+--------------------+--------------------|
| 7 18 6 | 3 5 4 | 18 9 2 |
| 13 4 138 | 2 6 9 | 1578 17 15 |
| 2 9 5 | 8 7 1 | 3 6 4 |
|--------------------+--------------------+--------------------|
| 9 5 2 | 6 1 3 | 4 8 7 |
| 6 138 138 | 4 9 7 | 2 5 13 |
| 4 137 137 | 5 2 8 | 6 13 9 |
+--------------------------------------------------------------+
# 40 eliminations remain
|
|
|
Back to top |
|
|
tlanglet
Joined: 17 Oct 2007 Posts: 2468 Location: Northern California Foothills
|
Posted: Fri Jul 22, 2011 2:17 pm Post subject: |
|
|
I initially found a couple of single digit (1) steps and decided to attempt to complete the puzzle by deleting only (1s). Here is the minimum set I found.
Code: | *-----------------------------------------------------------*
| 8 137 1347 | 9 34 5 | 17 2 6 |
| 13 2 9 | 17 8 6 | 157 4 135 |
| 5 6 1347 | 17 34 2 | 9 137 8 |
|-------------------+-------------------+-------------------|
| 7 18 6 | 3 5 4 | 18 9 2 |
| 13 4 138 | 2 6 9 | 1578 17 15 |
| 2 9 5 | 8 7 1 | 3 6 4 |
|-------------------+-------------------+-------------------|
| 9 5 2 | 6 1 3 | 4 8 7 |
| 6 138 138 | 4 9 7 | 2 5 13 |
| 4 137 137 | 5 2 8 | 6 13 9 |
*-----------------------------------------------------------* |
#1: Kraken x-wing(1)r25c19=(1-3)r8c9=r2c9=(3=1)r2c1; r2c47<>1
#2: skyscraper(1)r2c19 plus transport: (1=3)r5c1-(3=1)r2c1-r2c9=r8c9-r9c8=r5c8; r5c37<>1
Code: | *-----------------------------------------------------------*
| 8 137 1347 | 9 34 5 | 17 2 6 |
| 13 2 9 | 7 8 6 | 5 4 13 |
| 5 6 347 | 1 34 2 | 9 37 8 |
|-------------------+-------------------+-------------------|
| 7 18 6 | 3 5 4 | 18 9 2 |
| 13 4 38 | 2 6 9 | 78 17 5 |
| 2 9 5 | 8 7 1 | 3 6 4 |
|-------------------+-------------------+-------------------|
| 9 5 2 | 6 1 3 | 4 8 7 |
| 6 138 138 | 4 9 7 | 2 5 13 |
| 4 137 137 | 5 2 8 | 6 13 9 |
*-----------------------------------------------------------* |
#3: Code: | axy-wing(37-1)b1q24+r1c7=(1)r1c2-r1c3*=(1)r8c3-(8)r8c3=(8-3)r5c3=(3-1)r5c1=r2c1; r1c3*,r2c9<>1
| =(1)r9c3-r9c8=r8c9;r1c3*,r2c9<>1 |
Ted |
|
Back to top |
|
|
susume
Joined: 13 May 2011 Posts: 36 Location: Southeastern US
|
Posted: Sat Jul 23, 2011 2:53 am Post subject: |
|
|
Found several deletions of 1s, but in the end only needed two of them:
XYZ wing 137 pivot r1c2 => r1c3<>1
anp (18=3)r48c2-r8c9=r2c9-(3=1)r2c1 => r1c2<>1
Ted, I learned the ANP from your posts, and now they jump out at me when I'm looking for XY wings. |
|
Back to top |
|
|
peterj
Joined: 26 Mar 2010 Posts: 974 Location: London, UK
|
Posted: Sat Jul 23, 2011 8:37 am Post subject: |
|
|
Having played an RP in my first pass on this (confession), I wondered if... an oddagon one-stepper!
Code: | *-----------------------------------------------------------*
| 8 13(7) 1347 | 9 34 5 |(17) 2 6 |
|*13 2 9 | 17 8 6 | 157 4 *13+5 |
| 5 6 1347 | 17 34 2 | 9 137 8 |
|-------------------+-------------------+-------------------|
| 7 (18) 6 | 3 5 4 |-18 9 2 |
|*13 4 *13+8 | 2 6 9 | 1578 17 (15) |
| 2 9 5 | 8 7 1 | 3 6 4 |
|-------------------+-------------------+-------------------|
| 9 5 2 | 6 1 3 | 4 8 7 |
| 6 138 138 | 4 9 7 | 2 5 *13 |
| 4 13(7) *13+7 | 5 2 8 | 6 *13 9 |
*-----------------------------------------------------------*
oddagon(13) r2c1,r5c1,r5c3,r9c3,r9c8,r8c9,r2c9 ==> r4c7<>1
(8)r5c3 - (8=1)r4c2
||
(5)r2c9 - (5=1)r5c9
||
(7)r9c3 - r9c2=r1c2 - (7=1)r1c7 |
And an elimination on 1 to boot! |
|
Back to top |
|
|
Luke451
Joined: 20 Apr 2008 Posts: 310 Location: Southern Northern California
|
Posted: Sat Jul 23, 2011 1:36 pm Post subject: |
|
|
peterj wrote: | Having played an RP in my first pass on this (confession), I wondered if... an oddagon one-stepper! |
Oddagonagain! Very nice and concise. |
|
Back to top |
|
|
tlanglet
Joined: 17 Oct 2007 Posts: 2468 Location: Northern California Foothills
|
Posted: Sat Jul 23, 2011 2:26 pm Post subject: |
|
|
Hey, you folks are as crazy as I am solving puzzles in weird ways; hope you enjoyed yourselves.
So, I had a "hidden" reason for making this post: Oddogans. This technique has come up a couple of times recently, primarily posted by Luke and Peter, and I tried unsuccessfully to use it on this puzzle.
I found very little info on the topic but have a grasp of how it works; what I don't have is an understanding of how to "spot" the critical pattern. I would greatly appreciate any info on this technique.
Ted |
|
Back to top |
|
|
Luke451
Joined: 20 Apr 2008 Posts: 310 Location: Southern Northern California
|
Posted: Sat Jul 23, 2011 4:17 pm Post subject: |
|
|
tlanglet wrote: | I found very little info on the topic but have a grasp of how it works; what I don't have is an understanding of how to "spot" the critical pattern. I would greatly appreciate any info on this technique.
Ted |
One place to start is the Sudopedia Broken Wing article. As far as spotting them, candidate filtering will make them pop out.
To my knowledge, the "oddagon" term has come into use only in the last couple of years. The first I saw of it was in the work of Steve Kurzhals, who goes deep with the idea. I'm not sure if he originated the term, though.
Here is a "Type 1" oddagon from last year, not deep at all.
Just another "DP/impossible pattern" technique for the quiver. |
|
Back to top |
|
|
peterj
Joined: 26 Mar 2010 Posts: 974 Location: London, UK
|
Posted: Sat Jul 23, 2011 4:50 pm Post subject: |
|
|
Ted, Luke,
These are fun patterns - but I have to say I have not yet found a BrokenWing i.e. odd number of strong bi-location links in a loop, that did anything more than perhaps compress two turbot moves (kite or sky) into one. I have read they exist but not so far for me. Now Oddagons i.e. odd number of identical bivalues in a loop (my working definition anyway), seem to do more damage.
Couple of things on spotting them...
- you always seem to need a strong link or pair of bivalues in a diagonal orientation in a block (b9 in this case). Otherwise it's hard to form a loop with an odd number of cells.
- if you get that "remote pair feeling" start looking! An "almost remote pair" is going to be an oddagon. A remote pair which makes both eliminations is a type-1 oddagon I guess. |
|
Back to top |
|
|
Luke451
Joined: 20 Apr 2008 Posts: 310 Location: Southern Northern California
|
Posted: Sat Jul 23, 2011 5:34 pm Post subject: |
|
|
This makes me wonder: is the "oddagon" term reserved only for bivalue patterns?
It's the same concept as guardians/broken wings, but I can't recall a single digit pattern labeled as an oddagon. |
|
Back to top |
|
|
ronk
Joined: 07 May 2006 Posts: 398
|
Posted: Sat Jul 23, 2011 6:18 pm Post subject: |
|
|
Luke451 wrote: | This makes me wonder: is the "oddagon" term reserved only for bivalue patterns?
It's the same concept as guardians/broken wings, but I can't recall a single digit pattern labeled as an oddagon. |
Steve Kurzhals coined the oddagon term AFAIK, so this question should be to him. Steve generally likes broadly-defined terms, so my guess is the answer would be "no". |
|
Back to top |
|
|
daj95376
Joined: 23 Aug 2008 Posts: 3854
|
Posted: Sat Jul 23, 2011 10:09 pm Post subject: |
|
|
* I'm not surprised that an oddagon pattern existed.
* I'm not even surprised that three of the cells contained only one extra candidate.
* But Peter knocked my socks off when he managed to get those extra candidates to lead to a common elimination that cracked the puzzle!!! |
|
Back to top |
|
|
|