paquita's SE 11.8 puzzle

Post puzzles for others to solve here.

Re: paquita's SE 11.8 puzzle

Postby Cenoman » Tue May 05, 2020 1:15 pm

Hi SpAce,
Thank you, first for reading carefully my heavy prose, second for your sound comments.
SpAce wrote:I recall from my few earlier experiments that it was sometimes difficult or even impossible [...] to make both orientations work. Thus, sometimes one or the other is the more natural fit (like in your second example); other times it might not matter much (like in the first example). Would you agree?
How could I not ? The few times when I was tempted to say it impossible, I left it aside and found later the trick for the second orientation (I was in search for a counter-example) I was not aware of David's statement about the complementary locked sets. Thank you for the quotation.
SpAce wrote:I have to complain about terminology
Surprise, surprise, I can't believe it ! :o
SpAce wrote:Hope you don't mind!
Definitely, I don't. :)
I adopted bad habits on sudokuwiki.org. But I catch your point. I am also in line with eleven's statement that the Home set and Away set are interchangeable and there is no need to give them names. From now on, I'll not use 'base', nor 'Home' nor 'Away'. Just the list of cells, the list of links and the list of eliminations (with tagged pm's, if possible, not the case on any forum). That's clear and comprehensive.

That you also for confirming the manual process with your coloured rectangles. My own presentation came from the need to use only text justifications.
Cenoman
Cenoman
 
Posts: 2711
Joined: 21 November 2016
Location: France

Re: paquita's SE 11.8 puzzle

Postby SpAce » Wed May 06, 2020 6:03 am

Hi Cenoman,

Glad to see that we're practically in full agreement!

Cenoman wrote:The few times when I was tempted to say it impossible, I left it aside and found later the trick for the second orientation (I was in search for a counter-example)

So you've always eventually found a working second orientation, instead of finding a true counter-example? That's encouraging! In fact, that sounds like a wonderful way to practice balancing tricks. If you've already found one working variant, it should work as a baseline to massage the more challenging variants into compliance, which in turn should help in cases when you really need such tricks. I should try more of those kinds of exercises!

I adopted bad habits on sudokuwiki.org.

A common source for many of them, I'm afraid :) I started my studies there, and it took a while to unlearn many non-optimal or even blatantly false ideas after I found better sources or figured things out on my own. (One of my general pet peeves is the abuse of the term "strategy". Not sure if it originated from sudokuwiki but at least many have adopted it because of it. I wonder what they call an actual solving strategy if they already use that term for individual tactics and techniques. A grand strategy? Or maybe they don't have one.) That said, I'm glad there is such a site as sudokuwiki despite its imperfections.

But I catch your point. I am also in line with eleven's statement that the Home set and Away set are interchangeable and there is no need to give them names. From now on, I'll not use 'base', nor 'Home' nor 'Away'.

I'm not sure what eleven might have meant. To me the Home set is always the one with four digits and the Away set has the other five, so they're not interchangeable per se because their contents are locked. Their roles are interchangeable, however, as both can be used to cover either rows or columns (presuming one can make both orientations work, as discussed above). I still think it's useful for communication purposes to be able to identify the two groups of digits with simple names, even though they serve no deeper purpose.

What makes "base digits" a really bad term is that it falsely implies that one of the two sets is a base set and the other a cover set, while both are in fact cover sets (when attached to corresponding rows and columns) for the intersection cells that form the base set. Home and Away imply no such falsehoods as they just mark the number of digits in each group without implying anything about their roles. That said, one can certainly live without such names, and I also think I'll mostly drop them. However, I don't think they're totally useless, as they're related to the process of finding and describing the MSLS pattern (not so much to using it, as that's normal set logic which doesn't need such concepts).

Just the list of cells, the list of links and the list of eliminations (with tagged pm's, if possible, not the case on any forum). That's clear and comprehensive.

I agree that that's all that's needed once the MSLS has been found. I use my compacted 'TxL (Rank R): {truths \ links}' notation for that, just like for any other set logic. It's totally generic and works for any combination of truth and link types. Unlike David, I think its counterproductive to use different notations for MSLS and other kinds of set logic, because it falsely implies that they're somehow different concepts. Any MSLS-specific information I optionally choose to include (separately) is only there to describe the pattern in a more readable format (MSNS 4x4: (1234)r1357, (56789)c1256), but all of that can be derived from the truths and links as well.

That you also for confirming the manual process with your coloured rectangles. My own presentation came from the need to use only text justifications.

No problem. I use both. The coloring phase is just for finding the digit groups and the truth cells visually, and then the eliminations once the Rank 0 balance is confirmed. I do the balancing calculations with the text-based grid like you. I'm a visually-oriented person so the coloring makes it easier to see the big picture. Apparently I share that with David because we both came up with the same coloring system for MSLS. Of course it would be dishonest to say that mine was created independently as I remembered reading that he used something like that, but I'd never studied the details (seemed too complicated) until I worked it out myself. Only after that I understood his system, and it turned out they were exactly the same.
User avatar
SpAce
 
Posts: 2671
Joined: 22 May 2017

Previous

Return to Puzzles