Identical Pair Forcing Pattern

Advanced methods and approaches for solving Sudoku puzzles

Re: re: degenarate

Postby denis_berthier » Sun Sep 07, 2008 4:40 am

Pat wrote:the simplest definition of a quartet
will be so simple that it does allow degenerate case
-- e.g. two duos.
personally, i like those simple definitions.
(not that i would ever take 2 duos and use them as a quartet.)

That's the whole question: should our definitions correspond to what the real players do or should they be unnecessarily general?
I'm clearly on the first side. In my book and in this forum, I've always tried to stick as much as possible to the basic concepts and practices of real players. It allowed me to define a consistent global conceptual framework based on pure first order logic and it led me to discover really new types of rules. I've never even thought of allowing ghost candidates in them but, now that the question has been raised, it'd be totally stupid for me to allow this in nrc(z)(t) chains.

I understand that others, with minds molded in a different way, may need simpler and overly general definitions, with no real gain and at the cost of proposing patterns such as the first one in this thread as something new, but that's not really my problem - except that it has led me (and may lead others) to waste too much time on it.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Postby ronk » Sun Sep 07, 2008 5:54 am

Denis, are you suggesting that if you removed the naked pair technique from your SudoRules, that the remaining naked quad technique wouldn't pick up the degenerate case of two naked pairs:?:
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: re: degenarate

Postby tarek » Sun Sep 07, 2008 6:09 am

denis_berthier wrote:
Pat wrote:the simplest definition of a quartet
will be so simple that it does allow degenerate case
-- e.g. two duos.
personally, i like those simple definitions.
(not that i would ever take 2 duos and use them as a quartet.)

That's the whole question: should our definitions correspond to what the real players do or should they be unnecessarily general?
I'm clearly on the first side.
I'm on the 2nd side, "unnecessarily" is your own opinion which I disagree with... I always try to build concepts on firm grounds, this would make things more future proof ...

You forget that REAL players are REAL humans. A human on several occasions did spot the Swordfish elimination where a smaller x-wing could have done the job. By assuming that no REAL player would spot two naked doubles as a naked quad is assuming that humans are incapable of comitting mistakes.

By the way Denis, Your book that you keep referring to ... Is it Freely available or is it commercial? If it is commercial (correct me if I'm wrong) ... Then I would be reluctant in participating in discussions where the DIRECT result is on the pages of a commercial book profiting one party of that discussion.

tarek
User avatar
tarek
 
Posts: 2624
Joined: 05 January 2006

Re: re: degenarate

Postby denis_berthier » Sun Sep 07, 2008 6:17 am

tarek,
The views and concepts I'm discussing in this forum I have also defined in this forum.
If you have problems with the idea that there exists books and that not all is free, that's not my problem.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Postby denis_berthier » Sun Sep 07, 2008 6:19 am

ronk wrote:Denis, are you suggesting that if you removed the naked pair technique from your SudoRules, that the remaining naked quad technique wouldn't pick up the degenerate case of two naked pairs:?:

Why should I remove NP and not NQ? In my approach, that'd be stupid.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Postby ronk » Sun Sep 07, 2008 6:22 am

denis_berthier wrote:
ronk wrote:Denis, are you suggesting that if you removed the naked pair technique from your SudoRules, that the remaining naked quad technique wouldn't pick up the degenerate case of two naked pairs:?:

Why should I remove NP and not NQ? In my approach, that'd be stupid.

Nonetheless, what's your non-evasive answer to my question:?:
Last edited by ronk on Sun Sep 07, 2008 2:32 am, edited 1 time in total.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: re: degenarate

Postby tarek » Sun Sep 07, 2008 6:32 am

denis_berthier wrote:tarek,
The views and concepts I'm discussing in this forum I have also defined in this forum.
If you have problems with the idea that there exists books and that not all is free, that's not my problem.
True. Unfortunately, this IMO would mean that every discussion you start, is probably commercially motivated & people engaging in these discussions are pawns serving your own commercial purposes.

Due to the probability that I'm wrong & to not distract others from the topic, I will abstain from discussing matters anymore.

tarek
User avatar
tarek
 
Posts: 2624
Joined: 05 January 2006

Re: re: degenarate

Postby denis_berthier » Sun Sep 07, 2008 6:57 am

tarek wrote:
denis_berthier wrote:tarek,
The views and concepts I'm discussing in this forum I have also defined in this forum.
If you have problems with the idea that there exists books and that not all is free, that's not my problem.
True. Unfortunately, this IMO would mean that every discussion you start, is probably commercially motivated & people engaging in these discussions are pawns serving your own commercial purposes.
Due to the probability that I'm wrong & to not distract others from the topic, I will abstain from discussing matters anymore.

Why didn't you ask if Myth's book is free?
How many times did you see any reference to my book in my 319 posts in this forum?
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Postby denis_berthier » Sun Sep 07, 2008 6:59 am

ronk wrote:
denis_berthier wrote:
ronk wrote:Denis, are you suggesting that if you removed the naked pair technique from your SudoRules, that the remaining naked quad technique wouldn't pick up the degenerate case of two naked pairs:?:

Why should I remove NP and not NQ? In my approach, that'd be stupid.

Nonetheless, what's your non-evasive answer to my question:?:

This was a non-evasive answer: it'd be stupid.
As it'd be stupid to remove nrczt2 and not nrczt3 and the next ones.
In SudoRules, rules are organised hierarchically; the hierarchy can't be destroyed in any arbitrary way.
Notice that this is a question about implementation, not about rules.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Re: re: degenarate

Postby tarek » Sun Sep 07, 2008 7:06 am

denis_berthier wrote:Why didn't you ask if Myth's book is free?
How many times did you see any reference to my book in my 319 posts in this forum?
In your defence, I think I know who you are & most of your cards are open for other people to see which is not the case for many other people (in general). I am guilty for judging people for what they say & not for what they don't.

I don't know how many times you mentioned it, but it was enough for me to ask a question about it. If everything is present in this forum, referring to your book is redundant & serves only to generate negativity towards your motives.

tarek
User avatar
tarek
 
Posts: 2624
Joined: 05 January 2006

Re: re: degenarate

Postby denis_berthier » Sun Sep 07, 2008 7:15 am

tarek wrote:
denis_berthier wrote:Why didn't you ask if Myth's book is free?
How many times did you see any reference to my book in my 319 posts in this forum?
In your defence, I think I know who you are & most of your cards are open for other people to see which is not the case for many other people (in general). I am guilty for judging people for what they say & not for what they don't.

I don't know how many times you mentioned it, but it was enough for me to ask a question about it. If everything is present in this forum, referring to your book is redundant & serves only to generate negativity towards your motives.

tarek

I never counted. But likely not more than 10 times.
Things such that nrczt chains were even defined and discussed here before I made a second edition including them.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Postby ronk » Sun Sep 07, 2008 7:29 am

denis_berthier wrote:This was a non-evasive answer: it'd be stupid.

"This was a non-evasive answer" ... in your opinion. My guess is your naked quad technique would pick up the degenerate case of two naked pairs. Since this would appear inconsistent with your argument, you are avoiding a direct answer to my question.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Postby denis_berthier » Sun Sep 07, 2008 7:38 am

ronk wrote:
denis_berthier wrote:This was a non-evasive answer: it'd be stupid.

"This was a non-evasive answer" in your opinion. My guess is your naked quad technique would pick up the degenerate case of two naked pairs. Since this would be inconsistent with your argument, you are avoiding a direct answer to my question.

NO: IT WOULD NOT. The rule for quads is only for non-degenerate ones.
And if you want to know, if I delete the rule for some nrczt(p), there will be NO nrczt(q) for any q>p. Doing differently would be stupid (in particular because it'd destroy confluence).
Again, this is a question of implementation and I've always been careful not to enter into implementation discussions. All I do can be defined autonomously at the pure logic level and I don't want to introduce any confusion between the two levels.
denis_berthier
2010 Supporter
 
Posts: 1253
Joined: 19 June 2007
Location: Paris

Re: re: degenarate

Postby DonM » Sun Sep 07, 2008 12:18 pm

denis_berthier wrote:In my book and in this forum, I've always tried to stick as much as possible to the basic concepts and practices of real players.


I'm sure you really believe that. However, I bought your book and I've read your threads and, speaking as a purely manual solver (and also one who has programmed and sold a computer software package, non-Sudoku, in the past and so has no gripe against computers), your contributions are IMO far removed from the 'basic concepts and practices of real players'.

That doesn't make the contribution bad nor unappreciated. In fact, you have come up with some very interesting concepts, but they are IMO, in the theoretical, academic ivory tower, heavily computer output related category (a few solvers using your methods on a French website nonwithstanding). Over the last year or so, I have waited, in vain, to see if there would ever be anything remotely close to a practical tutorial (or even some practial, handholding examples) that would describe your methods to 'real players', only to see reams of overwhelming computer output substituting as examples.

However, to end on a positive note, your threads have a very useful function & purpose on a Sudoku forum, but just to keep things real, they do not contribute much to the everyday 'concepts and practices' of the typical manual solver.
DonM
2013 Supporter
 
Posts: 475
Joined: 13 January 2008

Postby Myth Jellies » Sun Sep 07, 2008 1:11 pm

Denis wrote:NO: IT WOULD NOT. The rule for quads is only for non-degenerate ones.
And if you want to know, if I delete the rule for some nrczt(p), there will be NO nrczt(q) for any q>p. Doing differently would be stupid (in particular because it'd destroy confluence).


Ah, so keeping it simple for the moment, to really find an xy(12) chain connecting two points, one must first guarantee that there is no xy(n<12) connecting those two points under your paradigm. Otherwise that really wouldn't be an xy(12) chain and you would not have the right to make the resultant deduction (even though of course we know that the smaller chain would also give you the right to make the deduction). I just don't see that as being an especially intelligent way of looking at things myself. Essentially your paradigm forces an implementation where every possible chain of a shorter length has to be considered without error first before moving on to considering longer chains. Intelligent if you are interested in keeping stats on what is needed to solve a puzzle perhaps, but really stupid if you want to be able to build on work already accomplished. It is like working a maze by always going back to the start point when you hit a node until you've marked all the first nodes. Repeat the process for all the second nodes from start making sure they aren't first nodes, then all the third nodes, et cetera until you finally reach the exit.

By the way, I have no real book. "In my book" is just another way of saying "in my opinion".
Myth Jellies
 
Posts: 593
Joined: 19 September 2005

PreviousNext

Return to Advanced solving techniques