| View previous topic :: View next topic |
| Author |
Message |
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
 Posted: Tue Dec 09, 2008 9:16 am Post subject: accessible 12/9 vanhegan extreme |
 |
|
| Code: |
+-------+-------+-------+
| . 9 8 | . 5 . | 6 4 . |
| 6 . . | . 9 . | . . 5 |
| . . . | 6 . 8 | . . . |
+-------+-------+-------+
| 4 6 . | 9 8 5 | . 1 2 |
| . . 2 | . 6 . | 9 . . |
| 9 3 . | 2 4 7 | . 5 6 |
+-------+-------+-------+
| . . . | 8 . 6 | . . . |
| 7 . . | . 3 . | . . 8 |
| . 8 6 | . 7 . | 4 2 . |
+-------+-------+-------+
|
Play this puzzle online at the Daily Sudoku site |
|
| Back to top |
|
 |
Marty R.
Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA
|
| Posted: Tue Dec 09, 2008 5:44 pm Post subject: |
|
|
| This was one of those that bent but wouldn't break. I used a half-dozen moves, including X-Wing, UR, coloring, W-Wing, ERs and an XY-Wing before it succumbed to two more W-Wings. |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Tue Dec 09, 2008 7:27 pm Post subject: |
|
|
this was a weird day in the vanhegan puzzle group.
his "Very Hard" puzzle rated 7.1 SE and by far outweighs his other puzzles as far as vh tech's are concerned.
his "Fiendish" rated 7.1 SE and was considerably easier than his "very hard"
then the "Extreme" which I posted is rate 6.6 SE. but needs some vast coloring and/or vh + tech's. |
|
| Back to top |
|
|
tlanglet
Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills
|
| Posted: Tue Dec 09, 2008 7:31 pm Post subject: |
|
|
This style of puzzle is my favorite; they just keep going, and going, but also keep you interest by offering a wide variety of moves.
In this puzzle, I had seven steps including two URs, two skyscrapers, two xy-wings and an ER.
Ted  |
|
| Back to top |
|
|
Marty R.
Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA
|
| Posted: Tue Dec 09, 2008 10:05 pm Post subject: |
|
|
| Norm, what are these SE ratings? On the posted puzzle I see a Vanhegan rating of 2.2.0.1.1. |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Tue Dec 09, 2008 10:22 pm Post subject: |
|
|
| Marty R. wrote: | | Norm, what are these SE ratings? On the posted puzzle I see a Vanhegan rating of 2.2.0.1.1. |
SE is "Sudoku Explainer", its another solving program that uses a complex algorithm to not only solve puzzles but give ratings to the tech's used. the higher the rating the more difficult. its popularity would probably stem from the fact that it uses highly complex forcing chain systems to solve everything that is thrown at it.
I wouldn't go and put a 10.0 SE or higher in it just yet. my computer's processor took 20 mins to solve a 9.5. I was reading on the player's forum about a 11.0 rated puzzle in which a pentium 4 spent all night solving. |
|
| Back to top |
|
|
Asellus
Joined: 05 Jun 2007
Posts: 865
Location: Sonoma County, CA, USA
|
| Posted: Wed Dec 10, 2008 8:59 am Post subject: |
|
|
After reading the above, I was surprised that I only needed 3 steps:
2 Skyscraper r18
13 Type 1 UR r1c4=7
1 Skyscraper c15 |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Wed Dec 10, 2008 4:56 pm Post subject: |
|
|
| Asellus wrote: | After reading the above, I was surprised that I only needed 3 steps:
2 Skyscraper r18
13 Type 1 UR r1c4=7
1 Skyscraper c15 |
For all of its power, I believe Sudoku Explainer has some weak points -- including URs (I think). |
|
| Back to top |
|
|
cgordon
Joined: 04 May 2007
Posts: 769
Location: ontario, canada
|
| Posted: Wed Dec 10, 2008 9:29 pm Post subject: |
|
|
| Quote: |
2 Skyscraper r18
13 Type 1 UR r1c4=7
1 Skyscraper c15 |
Asellus: I have two UR's (13) and (39) - a skyscraper that gets rid of <1> in R3C7 - and two ER's on <2> that probably do the same damage as your r18 skys. But for the life of me I cannot see a skyscraper on c15 since I have four <1>s in c1 in box 1 and 7.
Craig |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Wed Dec 10, 2008 10:30 pm Post subject: |
|
|
| Code: | (39) UR1 [r37c89]
(1) Skyscraper [c59] -- (13) UR4 not needed
(2) Skyscraper [r18]
(2) XY-Wing [r1c9]/[r1c1]+[r3c7]
(1) Skyscraper [c15] -or-
(1) Skyscraper [r29]
|
Last edited by daj95376 on Wed Dec 10, 2008 10:38 pm; edited 1 time in total |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Wed Dec 10, 2008 10:37 pm Post subject: |
|
|
| Quote: | | For all of its power, I believe Sudoku Explainer has some weak points -- including URs (I think). |
we humans have hindsight. a program with a set algorithm will never look back and go, "wait a minute that wasn't needed  I'll scratch that step"
this is ironic because a program can be written to do this... a program can be written to solve a puzzle many times with different heuristics each time, then just print the one with the shortest list. BUT humans generally don't want to go back to look because of either impatience or the satisfaction of making it the way they did in the first place.
UR's are high on the heirarchy list for SE
this is the partial heirarchy of moves for Sudoku Explainer.
xy-wing
xyz-wing
UR/loop
naked quad
jellyfish
hidden quad
BUG
APE
forcing chains/cycles
so you can see UR's are looked for right away. its nice because a UR is sometimes very noticeable that early, but we have seen puzzles where 2 or 3 UR type 4's have no bearing on the solution. but from a human solver that might be where one start anyways.
---
this is a partial heirarchy list for Sudocue.
naked quad
hidden pair
hidden triple
hidden quad
x-wing
swordfish
jellyfish
remote pair
BUG+1
skyscraper
2-string kite
xy-wing
xyz-wing
color Wrap
color trap
ER
Unique Corner (UR type 1
Unique side (UR type 2)
Unique Subset (UR type 3)
Unique Pair (UR type 4)
Sue de coq
APE
...then it goes through all the finned fish
before getting to xy-chain.
then it looks for medusa.
so Sucocue looks for UR's much further down the list. I am not so sure about looking for UR's that far down because they are easy on the eye to catch. but Sudocue obviously has the single digit patterns early and the multidigit patterns later.
I remember there was a discussion on what tech's should be considered first in a priority list to determine difficulty. these two programs show a considerable difference in approaches to that question.
so now the question becomes, if I didn't catch that UR first, how much harder/easier did I make the puzzle?
hypothetical situation: give 100 people 20 puzzles, have them rate the puzzles 1-100-- 1 being the easiest. then have that repeated with 100 new people and so on. wanna bet the puzzles all score differently? or would you bet that the puzzles had around the same score? hmm?
Last edited by storm_norm on Wed Dec 10, 2008 11:04 pm; edited 3 times in total |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Wed Dec 10, 2008 10:55 pm Post subject: |
|
|
| Quote: | | After reading the above, I was surprised that I only needed 3 steps: |
yeah, after looking at the other puzzles I realized it was the easiest one for the day, I think. |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Wed Dec 10, 2008 10:56 pm Post subject: |
|
|
| storm_norm wrote: | UR's are high on the heirarchy list for SE
|
High, as in rating a difficulty of 6.6 for a UR Type 1  |
|
| Back to top |
|
|
cgordon
Joined: 04 May 2007
Posts: 769
Location: ontario, canada
|
| Posted: Wed Dec 10, 2008 10:57 pm Post subject: |
|
|
Dan. The xy wing!!! - very interesting. I was almost there - I'd covered all the other stuff with my ERs and URs. I am soo xy-wing challenged (along with other stuff).
PS I just came in after posts re difficulty. Interesting that a naked quad is higher than wings. Personally I think ERs and URs are no brainers - which is why I use them.
Last edited by cgordon on Wed Dec 10, 2008 11:05 pm; edited 1 time in total |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Wed Dec 10, 2008 11:00 pm Post subject: |
|
|
| Craig wrote: | Dan. The xy wing!!! - very interesting. I was almost there - I'd covered all the other stuff with my ERs and URs. I am soo xy-wing challenged (along with other stuff).
|
Me too. Welcome to the club!!!
| Asellus wrote: | After reading the above, I was surprised that I only needed 3 steps:
2 Skyscraper r18
13 Type 1 UR r1c4=7
1 Skyscraper c15 |
Would you please explain how you managed to get the (13) UR to be a Type 1 |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Wed Dec 10, 2008 11:04 pm Post subject: |
|
|
| daj95376 wrote: | | storm_norm wrote: | UR's are high on the heirarchy list for SE
|
High, as in rating a difficulty of 6.6 for a UR Type 1 |
no. since its higher heuristically, it would be rated lower. in other words it looks for the easier tech's first.
on the other hand, according to sudocue, it looks for them later and if you look at the scores that are given for UR's as compared to swordfish or x-wings and kites, UR's are scored as 200-240-280 somewhere in there. kites, x-wings swordfish are around 140-180-200.
so according to Sodocue's base default scores, UR's are considered a more difficult technique set. |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Wed Dec 10, 2008 11:14 pm Post subject: |
|
|
From the Sudoku Explainer home page:
| Code: | # Implements the following solving techniques:
* Naked Singles and Hidden Singles
* Direct Pointing, Claiming and Hidden Pairs
* Naked and Hidden Pairs, Triplets and Quads
* Locking, Claiming (intersections), X-Wing, Swordfish and Jellyfish
* XY-Wing and XYZ-Wing
* Unique Rectangles and Loops
* Bivalue Universal Grave
* Aligned Pair Exclusion
* Bidirectional Cycles
* Nishio
* Forcing Chains
|
The 6th asterisk down contains UR. My guess is that URs do qualify as Level 6. |
|
| Back to top |
|
|
storm_norm
Joined: 18 Oct 2007
Posts: 1741
|
| Posted: Wed Dec 10, 2008 11:50 pm Post subject: |
|
|
| Quote: | | The 6th asterisk down contains UR. My guess is that URs do qualify as Level 6. |
 I think you are right. I found a puzzle that can be solved with a single very easy to spot UR...
| Code: | . . 2|5 6 .|1 . .
. . .|. 8 3|. . .
6 . .|. . 9|. 2 .
-----+-----+-----
. 3 4|. . .|6 . .
5 . .|. . .|. . 2
. . 7|. . .|4 5 .
-----+-----+-----
. 5 .|7 . .|. . 4
. . .|1 5 .|. . .
. . 1|. 4 6|7 . . |
| Code: | .------------------.------------------.------------------.
| 4 9 2 | 5 6 7 | 1 38 38 |
| 1 7 5 | 2 8 3 | 9 4 6 |
| 6 8 3 | 4 1 9 | 5 2 7 |
:------------------+------------------+------------------:
| 289 3 4 | 89 279 5 | 6 789 1 |
| 5 1 89 | 6 379 4 | 38 3789 2 |
| 289 6 7 | 389 239 1 | 4 5 389 |
:------------------+------------------+------------------:
| 39 5 6 | 7 39 *28 |*28 1 4 |
| 7 4 89 | 1 5 *28 |*238 6 389 |
| 389 2 1 | 39 4 6 | 7 89 5 |
'------------------'------------------'------------------' |
obvious UR marked in the puzzle that solves it
but SE instead went with solving it with
1 swordfish
2 xy-wings |
|
| Back to top |
|
|
Asellus
Joined: 05 Jun 2007
Posts: 865
Location: Sonoma County, CA, USA
|
| Posted: Thu Dec 11, 2008 2:28 am Post subject: |
|
|
| Craig wrote: | | But for the life of me I cannot see a skyscraper on c15 since I have four <1>s in c1 in box 1 and 7. |
| daj95376 wrote: | | Would you please explain how you managed to get the (13) UR to be a Type 1 |
It seems I neglected to write down one of my steps!  I had 4:
2 Skyscraper r18
17 W-Wing r12
13 Type 1 UR r1c4=7
1 Skyscraper c15 |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
