The New Sudoku Players' Forum

[easysurf]

There is no need for an X-Wing.
The problem can be solved through logic.

There is a hidden pair in Box 3.
See: cells (row 1 column 9) and (row 3 column 9)

For more help, see Example # 12 @

http://www.easysurf.cc/su-doku.htm

[Karyobin]

I didn't say there was any need for an X-wing. Someone expressed an interest in what they were, I just pointed them in the right direction for an explanation, that's all.

[Doyle]

Congratulations on solving that puzzle. Your {18}{18}{138} is actually what is called a naked triple. A "complete" naked triple would be {138}{138}{138}, but there can be missing values in the set, which makes them easy to overlook. A distribuition that's especially hard to see is {13}{38}{18}; you can't place any of those values immediately, but you can eliminate all three from all other cells in the unit. Rather common and very useful.

[Anette]

Doyle, you said earlier that there might be an easier solution not involving naked triples.

I did it with only several naked pairs and a lot of eliminations.

Since I still have problems spotting naked triples (especially in the form {13} {38} {18}), I would probably not have cracked this one if had required naked triples.

[The Central Scrutinizer]

Thanks Doyle. I assume you were talking to me.

Congratulations on solving that puzzle. Your {18}{18}{138} is actually what is called a naked triple. A "complete" naked triple would be {138}{138}{138}, but there can be missing values in the set, which makes them easy to overlook. A distribuition that's especially hard to see is {13}{38}{18}; you can't place any of those values immediately, but you can eliminate all three from all other cells in the unit. Rather common and very useful.

I’m seeing those devious little {1 2} {2 3} {1 3} combinations everywhere! Do they have a name? Because it seems like just about everything has a cool name.

At first (and second) glance it seems like they’re ripe to be solved, but that combo really tells me nothing more about those three cells.

What I’m finding is that moving along a row or column to a new side of the puzzle (or widening my concentration zone to include a 3x3) usually yields more information than I thought. This happened in today’s Sun-Times puzzle.

I kept staring at this maddening pattern of {1 7 8} {1 7 8} {7 8} {7 8} in various rows of the center 3x3.

In the middle right 3x3, there was a pair of {7 8} {7 8} in column 7.

Then, in the upper right 3x3, there was this lone {1 7}, also in column 7. If I understand you right, the above two statements comprise a naked triple. (just like yesterday)

Realizing that the pair of {7 8} dominated those two cells, I deduced that the 1 had nowhere else to go. This set off a firestorm of confirmations that took me to the top middle 3x3, back down to the middle-middle and over to the middle right. When the dust had cleared, I had the entire ‘upper 6’ 3x3s completely solved.

What was so liberating about this deduction was that the bottom three 3x3s were, for all intensive purposes, empty. I’m always taking baby steps--when I work on a Patternless, I rarely write a word down until I can double or triple check it. It seemed so wrong to be filling in those numbers in pen without having more information. But it worked.

So far, none of the 4 puzzles I’ve worked in the Sun-Times have had to apply more complex logic than these naked triples. The numbers have pretty much isolated themselves.

Just to get my terminology straight:

Now, what I’m wondering about is the ‘Hidden’ pair or triple. More terms that I’ve come across recently. I think in working the puzzles, reading a few posts and writing this one, I’ve come to figure out the difference. The example above {1 3 8} {1 3 8} {1 3 8} is considered ‘naked’ because there are no other numbers cluttering the mix?

Would {1 2 3 8} {1 3 6 8 9} {1 3 5 7 8} be an example of a ‘hidden triple?’ The same three numbers are present in all 3 cells but there are other numbers camouflaging them. From the looks of it, you could apply the same logic and eliminate 1 3 8 from the rest of their respective units. (?)

[Karyobin]

I must admit before starting, I've just called (bellowed) a pub quiz. Whereas this in itself may not appear incriminating, they supply me with free Stella. 'nuff said, so here goes...

If you have a row with the candidate structure {1,7,8} {1,7,8} {7,8} {7,8}, then you've gone wrong.

Sorry.

The reasoning's quite simple...

A pairing of {7,8} {7,8} would mean there are no other possible candidate 7's or 8's in that row. So, removing these from the two cells {1,7,8} {1,7,8} would leave a 1 (and only a 1) in each cell. That's wrong isn't it - coz you can't have two 1's (or anything else, for that matter) in a row.

I haven't read the rest, coz my eyes aren't focussing too well...

Sorry again.

(By the way, the best way to debug your own working or get a clue as to the next logical step is always to enter the starting grid. There are some very clever posters here, but none yet who have turned out to be physic [sic].)

Oh, just spotted this one...

Would {1 2 3 8} {1 3 6 8 9} {1 3 5 7 8} be an example of a ‘hidden triple?’ The same three numbers are present in all 3 cells but there are other numbers camouflaging them. From the looks of it, you could apply the same logic and eliminate 1 3 8 from the rest of their respective units. (?)

This would be an example of a hidden triple, but only if none of the other cells in the group contained those numbers - then you could erase all other numbers from these cells (e.g. {2}, {6 & 9} and {5 and 7}. You see, they have to be in these cells alone.

Gotta go to bed now, cricket in ten and a half hours...

[The Central Scrutinizer]

Free Stella is a beautiful thing.

I realize I'm prattling on and really doing you guys a disservice by not posting a grid for a visual example. It's plain as day when I describe it.

Karyobin, I actually said in various rows in reference to that first string of 4 candidates I named. Not that it matters, because in Soduku, a 3x3 is no different than a row. Looking back at the paper, I realize I did omit an innocent 6 due to a typographical error: {1,7,8} {1,7,8} {6,7,8} {7,8} is what should have been shared.

That 6 obviously owns the cell, so what I really should have posted was: {1,7,8} {1,7,8} {7,8}, I had no business talking about that cell.

You're psychic enough to catch my booboo.

I will now devote the rest of this post to editing and re-editing a portion of the grid.

Let's see how this format looks. I stole it from another thread.

Code: Select all

.5.|9..|.4.
.1.|...|5..
8..|3.2|.9.
-----------
6..|4..|.1.
53.|.9.|.64
.9.|..5|..2
-----------
.7.|2.6|..9
..4|...|.7.
.6.|..4|.8.

There wasn't really any crazy logic involved. I just get jazzed when I solve a puzzle. Forgive me.

Tomorrow will be a different story, I'm sure.

[MCC]

Where can I join the Free Stella campaign.

We should lobby our MP's.

'FREE STELLA-SHE's INNOCENT'





Sorry, can't blame the booze, free or otherwise.

[Doyle]

Central S: Actually the congrats were for the original poster, JCell, but feel free to accept some anyway. This one puzzle has sure generated a lot of comment; it was rated "Hard" by the Wash Post (or by Pappocom), it wasn't actually that hard, but it was most instructive to those at a certain position on the learning curve.

[The Central Scrutinizer]

I've perverted the conversation with my own stuff. I tend to go off on tangents and I'm sorry.

I was looking for some pointers on terminology and just started rambling, rambling, rambling.

I'm going to start the puzzle this thread references on my way home tonight. (thanks for posting the starting grid Doyle)