The New Sudoku Players' Forum

[richardm]

For me this is one of the most difficut puzzles I have done. Far more difficult than any Times Super Fiendish to date. I don't use PM. Why is this one so hard. Can others do it realtively quickly without PM? If so How? It not that I don't know how to solve it, I just don't know how to do it quickly.

Code: Select all

... ... .3.
.93 .74 ...
1.. .3. 26.

... ..2 ..1
.5. 1.9 .7.
4.. 6.. ...

.62 .1. ..3
... 42. 95.
.7. ... ...

Your thought please.
Richard

[RW]

It's hard without pm's because it requires several consecutive moves that don't directly solve any cells. I tried it without pm's, got here:

Code: Select all

 *-----------*
 |...|.61|.3.|
 |693|274|...|
 |1..|.3.|26.|
 |---+---+---|
 |...|..2|..1|
 |.5.|1.9|.7.|
 |4..|6..|...|
 |---+---+---|
 |962|.1.|..3|
 |...|42.|95.|
 |574|.9.|...|
 *-----------*

and then I was stuck for a while. Eventually I noticed the naked triplet in row 4 (actually, I noticed the hidden quad) and started looking for implications by that. I first looked at the quad cells and almost immediately saw this short chain:
if r4c5=4 => r4c7=5 => r4c3=6 => r5c3=8 => r5c5=Empty => r4c5<>4

and the puzzle was solved. I then put the grid into SS to see what solution I had missed as I had to resolve to a chain, and it turned out that if I had started by looking at the cells in the triplet I would have found the easier solution... But this is the beauty of manual solving, you don't need to look for the most simple solution.

RW

[ravel]

Hi richardm,

i only can add that i dont like these puzzles for manual solving. Without having RW's skills, i probably would end up filling in the PM's, until the naked triple pops up (its always disappointing then, when the puzzles is solved with that).
There are a lot of sudokus, that are rated higher, but can be solved without PM's in an elegant way.

And you might be interested in RW's great thread Solving without pencilmarks.

[richardm]

Thanks for the reference. I note RW's opening remark about being the only one to enjoy solving without PM. Well, I'm there too. If I have to use PM the it;s a moral defeat for me.

I have to disagree about the enjoyment factor. For me these Superiors are beautiffly elegant because they don't use fancy tricks yet manage to keep their secrets from you. I believe there's a common misconception ttat the use of the so called advanced techniques results in very hard puzzles. Once you are aware of a few of these techniques then I find that these puzzles generally solve very quickly - and quite often with a variety of methods. The Times proimised that Super Fiendish would be at least as hard as the former Superios, but this has not turned out to be the case.

I'm going to print off and study RW in detail.

Thanks

Richard

[tarek]

This is the same issue that made me start the "Superior 2" thread, where you would have no consecutive non single moves. It turned out that these puzzles are very rare.........

What consecutive moves would be still comfortable for MANUAL solvers ???? (I could smell a superior 3 coming very soon )

tarek

[richardm]

Superior 3 was ok for me, though I can't remember much about it. I missed the early superiors up to 26, which is why I'm doing them now.

So, I read though RW's article and that approach agrees with mine (good article btw), in particular the enumeration of what can't go into a cell as oposed to what can go there. All the triples and quads are much easier to spot (relative term of course).

The other thing I do is test for some form of forcing chain when I get a pair of candidate cells in a box for a single number that are not co-linear.

Richard

[ravel]

What consecutive moves would be still comfortable for MANUAL solvers ???? (I could smell a superior 3 coming very soon )

I would very appreciate a thread with good puzzles for non PM solvers, but i suppose that they are hard to create with usual puzzle generators.

We saw that naked triples are very hard to spot without PM's (and in the case above also the hidden quad combined with a box/line).

But sometimes hidden quads jump into your eyes before any naked singles (imagine you have 1234 in a row and a column), also some unique rectangles are very easy to find, or 2 strong links, that eliminate a number from a hidden pair and so on.

[richardm]

tarek wrote:
What consecutive moves would be still comfortable for MANUAL solvers ???? (I could smell a superior 3 coming very soon Very Happy)

I've found my Superior 2 and 3. Both very nice puzzles.
Superior 2: two pairs with 9. one pair alternates (9/5) the other doesn't (9/2). the pairs shar two rows hence 9 must go.
Superior 3: an x-wing on 2 with a nested quasi x-wing also on 2. This pinned down 2 in r3c1. I realise now that I inadvertently used a 4-fish on this one. Slapped writsts all round I didn't intend to do that - but it was quite clear to see without PM. Perhaps I should have another go.

richardm

[Pat]

hey richardm,

tarek was referring to the 2 existing topics on "superior" puzzles --

1. The SUPERIOR thread
2. The SUPERIOR THREAD (Second collection)

-- and to the possibility of starting a 3rd topic

~ Pat

[richardm]

Two very interesting threads. Thanks for pointing these out.

I was going to ask whether there were any published collection of Superiors?

Richard