The New Sudoku Players' Forum

by **ArkieTech** » Mon Jul 15, 2019 11:12 am

Code: Select all: *-----------* |...|...|..6| |4.8|5.7|..2| |.7.|9..|...| |---+---+---| |12.|..6|.59| |...|..9|...| |53.|...|.18| |---+---+---| |...|..4|.2.| |6..|2..|7.1| |3..|...|...| *-----------*

Play/Print this puzzle online

by **SteveG48** » Mon Jul 15, 2019 1:22 pm

Code: Select all: *-----------------------------------------------------------* | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | *-------------------+-------------------+-------------------| | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | *-------------------+-------------------+-------------------| | 7 189-5 1-5 | 6 19 4 |d3589 2 d35 | | 6 ab4589 a45 | 2 39 35 | 7 ac48 1 | | 3 b1459 2 | 8 7 15 |c59 6 45 | *-----------------------------------------------------------*

(5=489)r8c238 - (8r8c2)|(9r9c2) = (89)b9p57 - (8|9=35)r7c79 => -5 r7c23 ; stte

by **Ngisa** » Mon Jul 15, 2019 2:30 pm

Code: Select all: +-------------------+-----------------+-------------------+ | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | +-------------------+-----------------+-------------------+ | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | +-------------------+-----------------+-------------------+ | 7 1589 c15 | 6 d19 4 | 3589 2 35 | | 6 b4589 b45 | 2 39 a35 | 7 48 1 | | 3 1459 2 | 8 7 e1-5 | 59 6 45 | +-------------------+-----------------+-------------------+

(5)r8c6 = r8c23 - (5=1)r7c3 - r7c5 = (1)r9c6 => - 5r9c6; stte

Clement

by **SpAce** » Mon Jul 15, 2019 4:32 pm

Code: Select all: .-------------------.----------------.---------------------. | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | :-------------------+----------------+---------------------: | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | :-------------------+----------------+---------------------: | 7 c15(8)-9 15 | 6 a1[9] 4 | c(35)8-9 2 c(35) | | 6 4589 45 | 2 39 35 | 7 48 1 | | 3 1459 2 | 8 7 a15 | b59 6 b45 | '-------------------'----------------'---------------------'

Almost Hidden Triple

(9=15)b8p29 - r9c79 = (538)r7c972 => -9 r7c27; stte

as 3D Medusa Multi-Coloring: Show

by **eleven** » Mon Jul 15, 2019 4:45 pm

Code: Select all: *-------------------------------------------------* | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | |-----------------+--------------+----------------| | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | |-----------------+--------------+----------------| | 7 158-9 a15 | 6 a19 4 | 3589 2 35 | | 6 b4589 b45 | 2 3-9 35 | 7 b48 1 | | 3 1459 2 | 8 7 15 | 59 6 45 | *-------------------------------------------------*

(9=15)r7c53 - (5=489)r8c238 => -9r8c5, r7c2; stte

or 2 steps in 1 (for SpAce):
(45=1)r78c3 - (1=9)r7c5 - (9=354)r8c356 => -4r8c28, -5r8c23; stte

by **SpAce** » Mon Jul 15, 2019 5:16 pm

eleven wrote:or 2 steps in 1 (for SpAce):
(45=1)r78c3 - (1=9)r7c5 - (9=354)r8c356 => -4r8c28, -5r8c23; stte

Thanks for that

Very nice. The chain proves a slightly stronger conclusion too: +4r8c3, -5r8c2. (Then again, a placement conclusion usually indicates that something could be simplified, as is the case here too: -5r8c23 gives the same result with a simple XY-Chain.)

by **rjamil** » Tue Jul 16, 2019 1:56 pm

Code: Select all: ........64.85.7..2.7.9.....12...6.59.....9...53.....18.....4.2.6..2..7.13........ +----------------+--------------+---------------+ | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | +----------------+--------------+---------------+ | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | +----------------+--------------+---------------+ | 7 1589 (15) | 6 9-1 4 | 3589 2 35 | | 6 4589 45 | 2 39 35 | 7 48 1 | | 3 459-1 2 | 8 7 (15) | 59 6 45 | +----------------+--------------+---------------+

W-Wing: 15 @ r7c3 r9c6 Strong Link (5 @ r8c23 r8c6 or 5 @ r7c79 r9c79) => -1 @ r7c5 r9c2; stte

OR

Code: Select all: +----------------+---------------+---------------+ | 9 15 135 | 4 2 138 | 1358 7 6 | | 4 16 8 | 5 136 7 | 13 9 2 | | 2 7 1356 | 9 136 138 | 1358 48 345 | +----------------+---------------+---------------+ | 1 2 7 | 3 8 6 | 4 5 9 | | 8 46 46 | 1 5 9 | 2 3 7 | | 5 3 9 | 7 4 2 | 6 1 8 | +----------------+---------------+---------------+ | 7 1589 (15) | 6 (19) 4 | 3589 2 35 | | 6 489-5 4-5 | 2 (39) (35) | 7 48 1 | | 3 1459 2 | 8 7 15 | 59 6 45 | +----------------+---------------+---------------+

WXYZ-Wing: 1359 @ r7c35 r8c56 => -5 @ r8c23; stte

R. Jamil

by **SpAce** » Wed Jul 17, 2019 6:52 am

Hi rjamil,

Nice solutions! You asked about the Eureka notation in the other thread, so I'll try and help you by translating your moves into that.

rjamil wrote:W-Wing: 15 @ r7c3 r9c6 Strong Link (5 @ r8c23 r8c6 or 5 @ r7c79 r9c79) => -1 @ r7c5 r9c2; stte

First, note that your move is not a basic W-Wing so calling it just that is a bit misleading. I would call it "Grouped W-Wing" because it's using at least one group node. Thus:

Grouped W-Wing:

Code: Select all: (1=5)r7c3 - r8c23 = r8c6 - (5=1)r9c6 => -1 r7c5,r9c2 or: (1=5)r7c3 - r7c79 = r9c79 - (5=1)r9c6 => -1 r7c5,r9c2

WXYZ-Wing: 1359 @ r7c35 r8c56 => -5 @ r8c23; stte

There are actually several ways to name and notate this, because it can be seen as an XY-Chain (all bivalue cells) and as a few different ALS-XY-Wings and ALS-XZs (depending on which bivalues are combined into ALS nodes). Lastly, it's one of those rare patterns that can be written as a one-linker (like all WXYZ-Wings, if my hypothesis holds), but I don't know what it should be called then (just ALS-Z maybe, because there's no explicit X!). So, here goes:

Code: Select all: Name Details Eureka Conclusion -------------------------------------------------------------------------------------- XY-Chain (5=1)r7c3 - (1=9)r7c5 - (9=3)r8c5 - (3=5)r8c6 => -5 r8c23 ALS-XY-Wing (X=1,Y=9,Z=5) (5=1)r7c3 - (1=9)r7c5 - (9=35)r8c56 => -5 r8c23 ALS-XY-Wing (X=1,Y=3,Z=5) (5=1)r7c3 - (1=93)r78c5 - (3=5)r8c6 => -5 r8c23 ALS-XY-Wing (X=9,Y=3,Z=5) (5=19)r7c35 - (9=3)r8c5 - (3=5)r8c6 => -5 r8c23 ALS-XZ (X=1,Z=5) (5=1)r7c3 - (1=935)b8p256 => -5 r8c23 ALS-XZ (X=9,Z=5) (5=19)r7c35 - (9=35)r8c56 => -5 r8c23 ALS-Z (Z=5) (51)r7c35 = (935)b8p256 => -5 r8c23 --------------------------------------------------------------------------------------

Only I would write the last one, so don't take it too seriously

Anyway, all of those chains are bidirectional (like all properly written AICs), so they can be read (or rewritten) from right-to-left as well. Btw, when presenting any ALS move like those with Eureka, it's mostly unnecessary clutter to provide the details or even the name of the pattern, as they can be easily read from the chain itself. One or the other is enough (and I much prefer Eureka).

One of the hardest things (but part of the fun) about Eureka is the abundance of ways to write the same thing correctly. For example, all of these chains containing ALS nodes are equivalent even though they look different:

Code: Select all: (5=19)r7c35 - (9=35)r8c56 (my usual style, also SteveG48's, Clement's, and usually eleven's) (5=19)r7c35 - (93=5)r8c56 (51=9)r7c35 - (93=5)r8c56 (Cenoman's current style) (51=9)r7c35 - (9=35)r8c56 (5=9)r7c35 - (9=5)r8c56 (Leren's style, sometimes eleven's style)

Personally I find it easiest to read (and write) ALS nodes if all digits are listed and the larger group of candidates is on the right hand side of the '=', because I normally read from left to right, and it also matches the way I see ALS progressions on the grid. It's also the style recommended by the (partly obsolete) Eureka documentation. As can be seen, tastes vary widely on both accounts, so it's best to get used to all of the variants. A couple more equivalent chains:

Code: Select all: (5=1)r7c3 - (1=935)b8p256 (my style) (5=1)r7c3 - (19=35)b8p256 (5=1)r7c3 - (13=95)b8p256 (5=1)r7c3 - (193=5)b8p256 (Cenoman's style) (5=1)r7c3 - (1=5)r7c5,r8c56 (Leren's style)

As you can see, there are two ways to specify a cell group in a box. I much prefer the bBpP style because it makes it immediately clear that the group is within the same box (plus it's shorter). I use the comma-style grouping when the node is split into multiple houses. But, again, both are valid.

Another thing to note is that the digit ordering within the parentheses of an ALS node doesn't matter (as long as the linking digits are on the proper side of the '='). So, once again, these are all logically equivalent:

Code: Select all: (5=1)r7c3 - (1=935)b8p256 (my style) (5=1)r7c3 - (1=395)b8p256 (could be my style as well) (5=1)r7c3 - (1=359)b8p256 (SteveG48's style)

I like to keep the right-linking digit(s) furthest right (again, as recommended by the Eureka documentation, and because I find it easiest to follow). I also like to match the digits and the cell positions as much as possible, though that's secondary. Steve likes to keep both in the numerical order, which has its benefits too, so no one right answer here either. In Steve's style the internal link marker can't be moved as freely (as in the previous example) because it might leave the linking digits on the wrong side of it.

by **rjamil** » Wed Jul 17, 2019 1:31 pm

Hi SpAce,

First of all, many thanks for such a detail response.

However, regret to inform that I need to study the AIC before understanding single version of simple logical English text move presentation in to very complex (non-standard, immature, illogical, trial and error) and compressed eureka notation manually.

I am happy as long as my solutions are, to solve few (around 50%) puzzles, without knowing complex moves. If each move is readable from left to right and right to left in eureka form then it should also be acceptable in one way obsolete English literal text form as long as pencilmark grid helps to visualize/explain the criteria of elimination(s).

As, not all computer programming languages and spoken/written languages are perfectly compatable with each other respectively.

Wish to learn eureka notation later/slowly and keep participating with old format.

R. Jamil
_______________________________
A professor wrote the following sentence on the board and asked his class to punctuate it:

Woman without her man is nothing.

Half of the class punctuated the sentence in the following way:

Woman: without her, man is nothing.

The other half of the class responded with the following:

Woman, without her man, is nothing.

It’s clear, from this example, that punctuation is very important indeed!
______________________________
Say ‘NO’ to Drugs from the NMB Police D.A.R.E Officers.

Oops! That one missing comma turned police officers into drug dealers.

by **SpAce** » Wed Jul 17, 2019 4:19 pm

rjamil wrote:However, regret to inform that I need to study the AIC

Yes, of course. Eureka is a language for presenting AICs, so the most important thing is to learn how AICs work. I just think that process can go hand in hand, because you can learn about AICs most easily by understanding Eureka (since almost all of the AIC examples on this forum are in Eureka format, except very old ones).

before understanding single version of simple logical English text move presentation in to very complex (non-standard, immature, illogical, trial and error) and compressed eureka notation manually.

Once again you lost me. Are you saying Eureka is those (colored) things?

Wish to learn eureka notation later/slowly and keep participating with old format.

You're very welcome to do that. There's nothing wrong with your solutions or their understandability (for me at least). However, if you ever wish to present more complex solutions, you'll find that your notations won't scale to those levels. They're heavily dependent on the reader knowing the name and the logic of the pattern beforehand, so they only work for some small and well-known patterns. Eureka doesn't depend on the reader knowing any patterns at all (unless those patterns are used as shortcut nodes in the chain, like the XY-Wing in my solution today), because the full logic is there for everyone to see.

We might have a little different idea about simplicity too. You seem to think that learning a bunch of ready-made patterns is somehow simpler than knowing one almost-all-powerful method that encompasses them all. I think it's much simpler to learn a single generic concept that works for almost everything, including but not limited to those ready-made patterns. I learned AICs first and only later learned the names of common patterns. That's why I've never needed those patterns for solving because I could already find them as anonymous chains. Learning the names just made communication easier because many people use them. Things are of course different with some complex patterns that aren't AICs (like JExocet).

by **rjamil** » Thu Jul 18, 2019 3:16 am

Hi SpAce,

SpAce wrote:We might have a little different idea about simplicity too. You seem to think that learning a bunch of ready-made patterns is somehow simpler than knowing one almost-all-powerful method that encompasses them all. I think it's much simpler to learn a single generic concept that works for almost everything, including but not limited to those ready-made patterns. I learned AICs first and only later learned the names of common patterns. That's why I've never needed those patterns for solving because I could already find them as anonymous chains. Learning the names just made communication easier because many people use them. Things are of course different with some complex patterns that aren't AICs (like JExocet).

Well, I started learning from small pattern based moves with the help of exemplar aid, whereas, it seems to me that, your learning skills are very high as compared to mine.

Is there any visual tool to learn chains/eureka?

Similarly, is there any tool to convert English text move directly in to eureka without learning/converting in to complex moves first?

I think, it's easy to imagine ninety different ways to represent one ALS move for me then to think one gigantic move having unlimited different ways and superceded many small moves without knowing each one. Yes, it will take homework to build all different ways of one ALS move, but later, developing routine(s) to cover all ways will be easy with efficient in speed in mind.

Being not an expert in Sudoku, need basic learning material from start to understand. It's same as school/college going student trying to learn University level course.

Actually, my level in Sudoku techniques is same as learning to write alphabets over doted A B C.

R. Jamil

by **SpAce** » Fri Jul 19, 2019 7:28 am

rjamil wrote:Is there any visual tool to learn chains/eureka?

Hodoku is clearly the best for learning chains (and pretty much everything else), and it shows chains visually. Even SudokuWiki works for that but in a much more limited fashion. Unfortunately both use the obsolete Nice Loop notation instead of Eureka, which is why I actually learned that first when I started with chaining. Eureka is much better, so I immediately switched to that when I found out about it. I don't actually know if any public solver outputs Eureka (those two are the only ones I've used (edit: not true, see * below), but it looks like SudoCue might. Even if it does, it's probably a bit outdated style.

(*) Added. I forgot Phil's solver, which is a shame. It seems to have Eureka support, which is only logical since Phil's been a regular here (as pjb). It also has other nice features (such as high-end patterns like JExocet and SK-Loop) missing from Hodoku. The UI isn't quite as friendly, though, so I've only used it when studying those missing patterns.

Similarly, is there any tool to convert English text move directly in to eureka without learning/converting in to complex moves first?

Probably not, but any of us here can help with that. It's no problem for me to translate your moves into Eureka if you want.

Actually, my level in Sudoku techniques is same as learning to write alphabets over doted A B C.

If that's true, you're doing pretty well all things considered!!

The New Sudoku Players' Forum

July 15, 2019

July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019

Re: July 15, 2019