The New Sudoku Players' Forum

[David P Bird]

This was previously an informative thread on Braid Analysis but it was very deliberately vandalised.

I have therefore pulled the opening post and renamed the thread Chat Room. It was not such a good post anyway as the example it used was flawed.

I hope that this will encourage Jason, whenever he returns, to take a firmer control of the loose cannons who post off topic throughout the forum.

David P Bird

[zealot17]

I don't think this elimination is valid. If you know for sure that 18 travels \ direction then it is okey to remove 8 from r5c8 since placing 8 will cause 3 to be placed r5c8 too. But 18 traveling \ way is not the only case as in the solution it appears it is 47 that travels / way.

[eleven]

Agreed. From stack 3 alone you cannot eliminate 8 in r5c8.
As the 4th drawback of braid analysis Ruud mentioned "Error Prone".

[David P Bird]

I don't think this elimination is valid. İf you know for sure that 18 travels \ direction than it is okey to remove 8 from r5c8 since placing 8 will cause 3 to be placed r5c8 too. But 18 traveling \ way is not the only case as in the solution it appears it is 47 that travels / way.

Hi zealot17 and welcome to the forum. I'm impressed that you chose such a difficult topic for your first post, and look forward to your future contributions.

On checking it I find you are right that the elimination is not justified by the chain alone. It leaves the option of the strands in r123c7 being 8\ 24/ when (8)r5c8 could be true.

I can't remember what other input I used to reach that conclusion, but it doesn't matter – it's not included in the justification.

Mischief maker Eleven is also right (for once), it's very easy to slip up using strand node inferences because they are a lot of work and are often forcing so they only work in one direction. The technique is best reserved for manually checking the options when solving simple puzzles without pencil marks.

David PB

[zealot17]

Hi David,
Thank you for your welcoming post. I am interested in sudoku for a while and I just meet the technique braid analysis. I have seen it for the first from sudopedia and this was the second article I saw that is written for braiding. Even sudopedia there is a missing 2 in second tower.

I am starting to think that this method isn't useful as I used to think. Simply because it requires lots of work and secondly you need to consider too many cases in order to make an elimination that (probably) can be made by easier techniques.

I once made an remarkable elimination with this technique but it required a lot of work and as I said there is probably chains that can make same job.

May be I can post a case that braiding can make a elimination

Zealot

[eleven]

Mischief maker Eleven is also right (for once),

Sorry David, there were 2 reasons for my comment.
The one is, that i never had read that before. Probably because i did not appreciate braid analysis.
The other was, that you have not posted for longer time, so i wanted to let the newcomer know, that i share his opinion - didn't know that zealot17 is so firm with that topic.
btw when was i not right ?

[David P Bird]

eleven,
Yes I'm not very active with Sudoku now as I've run out of new ideas and don't wish to contribute manually found AIC solutions against those that came from a computer solver or result from forcing or branched chains and possibly have weird and wonderful notations. You earned the 'mischief maker' title back in 2015 when you led me a merry dance on the puzzle for Aug 5th. The 'for once ' comment was my own bit of mischief.
David
.

[eleven]

You earned the 'mischief maker' title back in 2015 when you led me a merry dance on the puzzle for Aug 5th.

Old stuff. Still can't see, what was in your mind then.
But though i guess, you know it, let me assure, that i appreciate your contributions.

[rjamil]

Hi David P Bird,

Is braid analysis similar to as Tuleja’s (or Mr. T’s) Theorem?

Got an idea from this pdf link. In physical Page # 22 (printed page # 36), 2.2.2.15 Tuleja’s (or Mr. T’s) Theorem.

R. Jamil

[David P Bird]

Hi David P Bird,

Is braid analysis similar to as Tuleja’s (or Mr. T’s) Theorem?

You shouldn't have to ask. Yes it is - he came up with it in 2008, but I posted about it in 2006, and prior to me Hamilton had posted about finding 'travelling pairs' but had not analysed them very thoroughly.

It's not surprising that the same discovery or observation will be made independently by several researchers, and I believe it's foolish to try to claim one by putting one's name to it. Who knows what has been published in Japanese, the birthplace of Sudoku?

DPB

[rjamil]

Am still not fully grasp entire Sudoku solving strategies, not even co-relate several strategies either fully or partially overlap; or assumptive and non-assumptive difference.

By the way, this thread was created on Dec 2015. I need more time to analyze strategies one by one. But at the moment, interesting on pattern based and non-assumptive strategies only.

Thought that braid analysis is pattern based and non-assumptive, therefore, focusing on it now.

Regret for any misunderstanding.

R. Jamil

[David P Bird]

By the way, this thread was created on Dec 2015.

Responding to that insinuation, please look at <This Page>. At the top, you will find it was last revised in 2006 and at the bottom you will find a section on Braid Analysis.

[rjamil]

Hi DPB,

I have no objection regarding who invented sudoku puzzle? How, why, when and where? Whether sudoku born in Japan; or, born in America and reborn in Japan, the 'World's Most Influential People' of 2006 by Time magazine, and later if any, etc.

The main problem for me is that, there are lot of strategies that cover/overlap/supersede each other, partly or completely. Similarly, non-assumptive strategies that are very easy to spot by pencil-n-paper as well as to program, are superseded by very complex and assumptive strategies.

How do one know that either same strategy is define as separate name/title or two different strategies that are partly or fully overlap each other? And, assumptive strategy overlap with non-assumptive strategy or vice-versa?

Note: I thought that all pattern based strategies are non-assumptive (what I understand non-assumptive means non-recursive in programming) and easy to program too.

Sorry if I am too off-topic.

R. Jamil

[SpAce]

Off topic, but I've voiced my objection to the common misuse of the terms "strategy" and "assumptive/non-assumptive" here:

http://forum.enjoysudoku.com/help-to-solve-hardest-puzzles-t34876-45.html#p268399

People usually mean tactics or techniques when they talk about "strategies" so why not use those terms instead? A minor detail, perhaps, but what term should we use then if we *actually* talk about sudoku solving strategies??? Also, "assumptive/non-assumptive" seem to mean very different things to different people, so they don't really mean anything in this context and rarely seem to have much connection to the semantics of the words.

[rjamil]

Hi SpAce,

Many thanks for providing your post link, that thread caused me confusion in terminologies too.

What I understand is that all pattern based techniques are non-assumptive as they never produce contradiction later. Also, all those sudoku solver programs, that I analyzed, have programmed in such a way that some techniques are rollbacked when reached to some dead end, where as some techniques never rollbacked and when puzzle reached to dead end then program simply return with "unsolvable" or something like that message. Am unable to distinguish between them.

So I guess that all pattern based techniques contain non-assumptive property too; and rest, that can't be presented in pattern, contain assumptive property.

My question is that, is Braid analysis technique contain non-assumptive property?

Note: one more term that, not only confusing but hurt sometimes, is degenerative.

R. Jamil