| View previous topic :: View next topic |
| Author |
Message |
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
 Posted: Tue Sep 21, 2010 4:56 am Post subject: Puzzle 10/09/21: B |
 |
|
| Code: | +-----------------------+
| 7 . . | 4 . . | 3 . . |
| . 9 3 | 8 7 . | 4 . . |
| . 4 6 | . 3 . | . . . |
|-------+-------+-------|
| 9 3 . | 5 1 . | 8 . 7 |
| . 8 5 | 6 2 7 | . . . |
| . . . | . 8 . | . . . |
|-------+-------+-------|
| 2 5 . | 1 . . | 6 3 . |
| . . . | . . . | 7 2 . |
| . . . | 7 . . | . . 5 |
+-----------------------+
|
Play this puzzle online at the Daily Sudoku site |
|
| Back to top |
|
 |
JC Van Hay
Joined: 13 Jun 2010
Posts: 494
Location: Charleroi, Belgium
|
| Posted: Tue Sep 21, 2010 8:04 am Post subject: |
|
|
| Quote: | 4-SIS AIC : (94)R5C8 (41)R5C1 1C3 8R1 : (9)r5c8=(8)r1c8 : => r1c8<>9
|
|
|
| Back to top |
|
|
peterj
Joined: 26 Mar 2010
Posts: 974
Location: London, UK
|
| Posted: Tue Sep 21, 2010 11:33 am Post subject: |
|
|
A wing...
| Quote: | m-wing(48) (4=8)r8c1 - r9c3=(8-4)r9c8=r5c8 ; r5c1<>4
.. or the corresponding ...
s-wing(48) (4)r5c8=r5c1 - (4=8)r8c1 - (8)r9c3=r9c8 ; r9c8<>4 |
|
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Tue Sep 21, 2010 7:26 pm Post subject: |
|
|
Updated information originally provided by storm_norm.
| Code: | Three Strong, Two Weak, Inference Set: 3-SIS/2-WIS
_______________________________________________________________________________________
= - = - =
fishy: A = A - A = A - A = A {1st and last A in the same house} loop
Cycle {includes 222 swordfish}
X-Cycle: A = A - A = A - A = A {1st and last A not in the same house}
W-Wing: (A=B) - B = B - (B=A)
gM-Wing: (A=B) - B = (B-A) = A
S-Wing: A = A - (A=B) - B = B {1st A and last B in the same house}
unnamed: A = A - A = (A - B) = B {1st A and last B in the same house}
(reverse) B = (B - A) = A - A = A {1st B and last A in the same house}
XY-Wing: (A=B) - (B=C) - (C=A)
inv/XY: A = (A-B) = (B-C) = C {1st A and last C are in the same house}
|
If you examine the table above, you'll notice there are four entries with only two candidate values present. Norm once referred to the S-Wing as a "Cyclops", but it was essentially unnamed until S-Wing came along. That leaves one unnamed entry, and I'm calling it E-Turbot. The first 2-SIS represent a turbot fish pattern that'd provide eliminations for the single digit A. However, the 2-SIS can be extended into the fifth cell of the turbot pattern through a strong link on a second candidate value -- thus the E part of the name on this 3-SIS.
Besides the M-Wing and S-Wing mentioned by Peter for this puzzle, there's also an E-Turbot present:
| Code: | E-Turbot: 4C1 4C8 8R9 : (4)r8c1 = r5c1 - r5c8 = (4-8)r9c8 = (8)r9c3 => r9c3<>4, r8c1<>8
+--------------------------------------------------------------+
| 7 2 18 | 4 59 159 | 3 89 6 |
| 5 9 3 | 8 7 6 | 4 1 2 |
| 18 4 6 | 2 3 19 | 5 7 89 |
|--------------------+--------------------+--------------------|
| 9 3 2 | 5 1 4 | 8 6 7 |
| *14 8 5 | 6 2 7 | 19 *49 3 |
| 6 7 14 | 39 8 39 | 2 5 14 |
|--------------------+--------------------+--------------------|
| 2 5 7 | 1 49 89 | 6 3 489 |
| *48 16 489 | 39 4569 3589 | 7 2 1489 |
| 3 16 @489 | 7 469 2 | 19 @*489 5 |
+--------------------------------------------------------------+
# 41 eliminations remain
__________________________________________________________________________________________
|
For those with a good memory, you'll recall that I mentioned this pattern awhile back in
Atypical Extension of Non-Productive Skyscraper.
Last edited by daj95376 on Tue Sep 21, 2010 8:09 pm; edited 2 times in total |
|
| Back to top |
|
|
ronk
Joined: 07 May 2006
Posts: 398
|
| Posted: Tue Sep 21, 2010 7:57 pm Post subject: |
|
|
| daj95376 wrote: | | That leaves one unnamed entry, and I'm calling it XY-Turbot. The first 2-SIS represent a turbot fish pattern that'd provide eliminations for the single digit A. However, the 2-SIS can be extended into the fifth cell of the turbot pattern through a strong link on a second candidate value -- thus the XY part of the name on this 3-SIS. |
Let's see now. In an xy-wing, 100% of the strong inferences are bivalued cells. In an xy-chain, 100% of the strong inferences are bivalued cells.
Using your proposed name, we would have an xy-turbot without even one bivalued cell. Consequently, that name seems ill-advised. |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Tue Sep 21, 2010 8:11 pm Post subject: |
|
|
| Ron, you slipped in right in the middle of my coming to the same conclusion. What do you think of the updated name? |
|
| Back to top |
|
|
peterj
Joined: 26 Mar 2010
Posts: 974
Location: London, UK
|
| Posted: Tue Sep 21, 2010 8:52 pm Post subject: |
|
|
I don't know but for me "turbot" says single digit only - whereas "wing" says two digits. "t-wing" for turbot wing?
btw I have seen mention of h-wings - but never a clear definition? |
|
| Back to top |
|
|
Marty R.
Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA
|
| Posted: Wed Sep 22, 2010 12:41 am Post subject: |
|
|
Coloring (4) + two extensions
XYZ-Wing (189)
Multi-coloring (9) |
|
| Back to top |
|
|
ronk
Joined: 07 May 2006
Posts: 398
|
| Posted: Mon Sep 27, 2010 12:47 pm Post subject: |
|
|
| daj95376 wrote: | | Ron, you slipped in right in the middle of my coming to the same conclusion. What do you think of the updated name? |
As you might guess, I'm reluctant to put names on too many patterns. If this one must be named, however, I suggest "L-Wing." The "wing" because there are 3 strong links and the "L" because the strong links are 100% biLocaLs. One can take this a step further and have an "L2-Wing" and an "L3-wing" for 2 and 3 digits, respectively.
Using your nomenclature:
L2-Wing: A = (A - B) = B - B = B
L3-Wing: A = (A - B) = (B - C) = C |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Mon Sep 27, 2010 5:22 pm Post subject: |
|
|
| ronk wrote: | I'm reluctant to put names on too many patterns.
If this one must be named, however, I suggest "L-Wing." The "wing" because there are 3 strong links and the "L" because the strong links are 100% biLocaLs. One can take this a step further and have an "L2-Wing" and an "L3-wing" for 2 and 3 digits, respectively.
Using your nomenclature:
L2-Wing: A = (A - B) = B - B = B
L3-Wing: A = (A - B) = (B - C) = C |
I understand your reluctance and agree that every pattern does not need a name. Since it's the only unnamed pattern (that I believe remains) in my "3:2 group", It seems awkward not to give it a name. I can accept "L-Wing". I hope others find it acceptable as well. Thanks!
Regards, Danny |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
