The New Sudoku Players' Forum

[richardm]

Does any one have, or know how I can get to the set of superior puzzles published in the Sunday Times? I wasn't aware of the early ones and would like to opportunity to solve them. Tray as I might the search facility in the online edition of the times only surfaces a few of the early puzzles.

Richard

[udosuk]

I only know that if you search by "Su Doku" using the timesonline search engine you'll get the puzzles posted on weekdays, but if you search by "Sudoku" you'll get those Sunday puzzles... Not sure how many you can backtrack though...

[richardm]

The seach engine used y Time Online seems to be broken, It matters not whether I use "Su Doku" or "sudoku" neither brings up a list of past puzzles. Yet, googling for sudoku brings up some of the eary puzzles on the Times' website - e.g. Superior 8 can be found this way. It seems that the Times is not indexing the sudoku puzzles, hence I was wondering whether anyone had kept a private archive of the Superior puzzles?

Richard

[udosuk]

The search engine works for me as I'm typing now... Perhaps you've disabled the javascript function in your browser security setting or something...

The "Superior Sudoku" title was not used anymore after Oct 08 (the last one being www.timesonline.co.uk/article/0,,18209-2386866.html { broken link } Superior Sudoku 62)... Since then it's merely "Sudoku" starting on Oct 15 (www.timesonline.co.uk/article/0,,18209-2392354.html { broken link } Sudoku 63)...

From their search engine the earliest available one would be "Superior Sudoku 61" on Oct 08 (the solution of "Superior Sudoku 60" is available, but not the puzzle)... Perhaps you could contact the timesonline staff to see if there is any way they could send you the archive of earlier puzzles...

[richardm]

OK, thanks.. I think I know what the problem is. Tines online is locked into using IE and a particular version of Java - stupid programming practice.

I'm an exclusive Linux user so this is why I see the problem.

Richard

[Pat]

Does any one have, or know how I can get to the set of superior puzzles published in the Sunday Times? I wasn't aware of the early ones and would like to opportunity to solve them. Tray as I might the search facility in the online edition of the times only surfaces a few of the early puzzles.

http://forum.enjoysudoku.com/viewtopic.php?p=22065#p22065

[udosuk]

That's a very nice deed, Pat... Good on you mate!

[richardm]

Hear! Hear!

[tarek]

I gather that you also had a look at the superior thread.........

I had a go on trying to rate these puzzles & by that I discovered several points to improve, however, in their present form, they still -to me at least- are very very good superiors.

tarek

[richardm]

I agree, and frankly I don't think the current set of Times' Super Fiendish puzzles come close to these pappocom Superiors from the Sunday Times.

[Pat]

I don't think the current set of Times' Super Fiendish puzzles come close to these pappocom Superiors from the Sunday Times.

they say their Super Fiendish may be even tougher than the Pappocom Very Hard puzzles -- possibly requiring Swordfish

How to Solve Superfiendish

• The x-wing technique
  
• The Swordfish technique
  The first puzzle for which it is required appeared in Times2 on Friday, March 16, 2007.

[richardm]

Well, you say that I used neither technique - at least not knowingly. I stared at the puzzle for a an hour then saw that in the top right and top centre 3x3 boxes, 7 was confined to one of two alternating positions in each box. This pinned down 8 in the top row then 2 on bottom 3x3 and the rest fell out using simple elimination.

But the point I was making is that hard does not necessarily = good. Each of the superios seems ot have a rather sime idea at the core which is usuallyextremely well hidden. If you get the core idea you wonn't need any rought working at all, nor will you need to apply any stock so-called advanced techniques.

What I observe with the super fiendish puzzles is that to date they are very much quicker to solve, but seem to be rather clumsily constructed. Maybe they will improve in style over time - let's hope so.

[Pat]

back on Friday,
when i saw the article in The Times,
i was rather annoyed --
they had stated explicitly:
The Swordfish technique
The first puzzle for which it is required appeared in Times2 on Friday, March 16, 2007.
the way i see it,
when i'm told in advance that a Swordfish (or whatever) is required,
this information is a (partial) spoiler,
somewhat spoiling the fun for me --
now i'll be looking for a Swordfish much sooner than otherwise.

therefore,
i'm glad to report that they've re-phrased the above article --

How to Solve Super Fiendish

• The x-wing technique
  
• The Swordfish technique
  From today onwards, you will find the swordfish technique become a regular feature in Times2.

and now,
the Friday puzzles have appeared on Times Online --
see for yourself, what exactly does the Super Fiendish need?

#1236 -- Super Fiendish

[ 28 cells given ]

Code: Select all

 9 . 5 | . 6 . | . 4 .
 . 7 . | . . . | . . .
 . 8 . | 1 . . | . 5 9
-------+-------+------
 5 . . | . . 3 | . 8 1
 . 6 . | . . . | . 3 .
 3 1 . | 8 . . | . . 6
-------+-------+------
 1 3 . | . . 9 | . 6 .
 . . . | . . . | . 7 .
 . 9 . | . 4 . | 3 . 5



[richardm]

Interesting. I don't see one needs the contrivance of swordfish or x-wing. Simple elimination results in.

Code: Select all


 9 2 5 | . 6 . | 1 4 .
 . 7 1 | . . . | . 2 .
 . 8 3 | 1 . . | . 5 9
-------+-------+------
 5 4 . | 6 . 3 | . 8 1
 . 6 . | . . . | . 3 .
 3 1 . | 8 . . | . 9 6
-------+-------+------
 1 3 . | 5 . 9 | . 6 .
 . 5 . | . . . | 9 7 .
 . 9 . | . 4 . | 3 1 5
 

At which point we note that either (7,3) or (9,1) is 7.
If (9,1) is 7 then (6,1) is 8
If (7,3) is 7 then (6,6) is 7 and (6,1) is still 8

Hence (6,1) is 8. From here (1,8) is 2 and (4,8) is 3.
The rest falls out with simple elimination.

My technique is not to learn all these stock patterns but to look for a simple switching pair in a 3x3 box that don't share the same row or column. They generally imply some constraint elsewhere. If I see the patterns I'll use them, but I generally don't use anything more complicated than x-wing.
However I would say that the patterns are useful if you're trying to set a puzzle of a desired difficulty.

Richard

[udosuk]

To recap Richard's move in the lingo we use here:

Code: Select all

 *--------------------------------------------------------------------*
 | 9      2      5      | 37     6     *78     | 1      4     #378    |
 | 46     7      1      | 349    3589   458    | 68     2      38     |
 | 46     8      3      | 1      27     247    |@67     5      9      |
 |----------------------+----------------------+----------------------|
 | 5      4      279    | 6      279    3      | 27     8      1      |
 | 278    6      2789   | 2479   12579  12457  | 2457   3     @247    |
 | 3      1      27     | 8      257   @2457   | 2457   9      6      |
 |----------------------+----------------------+----------------------|
 | 1      3      2478   | 5     @278    9      | 248    6      248    |
 | 2468   5      2468   | 23     1238   1268   | 9      7      248    |
 |@2678   9      2678   | 27     4      2678   | 3      1      5      |
 *--------------------------------------------------------------------*

In b3, either r1c9=7 or r3c7=7:
r1c9=7 => r1c6=8
r3c7=7 => r5c9=7 => r9c1=7 => r7c5=7 => r6c6=7 => r1c6=8

Therefore r1c6=8 => 8 in b7 locked in r9c13 => r8c1=2 (naked pair {46} in r23c1)
And all naked singles from here.

This technique is called "forcing net" and many consider it more powerful than swordfish etc.

Some people who are allergic to trial & error don't like to resort to these moves.
I guess for them a swordfish is a easier path to take...

But certainly, if you're solving without PM or participating in a timed race then it is more favourable to do it this way...