croco-puzzle

[david mcneill]

Very good Stefano. I have known that guy for quite a few years and he always seems to be driving in the fast lane. As far as puzzles are concerned anyway.

David.

[dickoon]

OK, I'm leaving today's Doppelstern unfinished, which is very annoying. It has defeated me after spending rather too long on it.

Tomorrow, please could someone post a logical walkthrough of it? I think it requires some sorts of logical step that I don't know how to do. I can identify about ten squares which must be empty, but I cannot logically definitely place even a single star. I've tried a lot of trial and error and produced nothing. I lose, and growl.

[Calavera]

Should be no problem...

[Calavera]

In the following RxCy means row x, column y.

Obviously the following fields must be empty: R8C4+6+8, R9C4+6+8, R3C2, R5C2, R8C2

Because of the lower right area R1-3C9 must be empty. Then R2C5 must be empty.

As there must be 4 stars in the two top rows R3C3-4 must be empty. Then R1C2+4 and R2C2+4 must be empty, too.

The pairs R4C2, R5C2 and R3C1, R6C1 both must contain one star. So the following fields must be empty: R4C1+3, R5C1+3, R7-9C1, R4-8C3

This should offer more than enough entry points for the rest of the puzzle.

This was also the way I used when solving it yesterday. And as I didn't click especially fast I was quite surprised that no one beat my time...


Nils

[ronaldx]

I labelled the columns 1-9 and the rows A-I

Usually in a hard Doppelstar, the hard part is placing the first star. So, concentrate on squares you can cross
I will list squares that you can cross through without explanation, in a reasonable order, split by reasoning.
Hopefully they are correct coordinates - please ask for clarification of any step you like.

The puzzle is still available if you log out (to play yesterday's puzzles)

Empty:
H468
I468

C2
F2

ABC9

B4

B5

C3,4

A2,B2
A4

C4567
E6

E457

B6

H2,I2

STAR - G2 (empty surrounding squares)

D1,E1

STAR - C1, A1, I3, E2, B3

[PuzzleScot]

Thanks also for the walkthrough - I bifurcated my way through it eventually.
Today's Buchstaben puzzle was evil too. Took a LOT of attempts to see the logical path. However, I was interrupted for 'about an hour' by a work emergency at 11:30pm!

[drsteve]

Thanks also for the walkthrough - I bifurcated my way through it eventually.
Today's Buchstaben puzzle was evil too. Took a LOT of attempts to see the logical path. However, I was interrupted for 'about an hour' by a work emergency at 11:30pm!

Ugh - four rubbish times in a row. This certainly is turning into an addiction...

[dickoon]

As there must be 4 stars in the two top rows R3C3-4 must be empty.

Thanks, Nils, but this is the bit that I don't get. Yesterday I didn't spot that R1-3C9 must be empty but it does make sense today, and given that part that I don't understand, everything else is clear enough.

I appreciate your help!

[dickoon]

As there must be 4 stars in the two top rows R3C3-4 must be empty.

Thanks, Nils, but this is the bit that I don't get. Yesterday I didn't spot that R1-3C9 must be empty but it does make sense today, and given that part that I don't understand, everything else is clear enough.

I appreciate your help!

ETA: Similarly, thanks, Ronald; I'm lost at C3C4, then lost again at C4567.

[Calavera]

As there must be 4 stars in the two top rows R3C3-4 must be empty.

Thanks, Nils, but this is the bit that I don't get.

Hope, there's enough time to check it out after my posting or that someone made a screenshot...

The 4 stars in 2 rows part is directly deducted from the rules. Now after R2C5 must be empty there are only two areas left that have available fields in the upper two rows. So all of their fields that are not in these two rows have to be empty.


Edit: Btw, what does ETA stand for? I only know it as a terroristic organisation and estimated time of arrival and it's hard to make a sensible explanation for usage of one of these...

[ronaldx]

Dickoon - I lost access to the relevant puzzle but I think the squares you mentioned are the same as Nils explained... sorry. If you have a screenshot I will explain further. Not sure why C4 is listed twice in my walkthrough, I hope that's not a coordinate error. [Edit: it is, it should read uhh DEFG3,F5. It was late ]

ETA in this case is '[post] edited to add' (but, I haven't seen that usage either )

[berni]

Hope, there's enough time to check it out after my posting or that someone made a screenshot...

screen.png (6.6 KiB) Viewed 74342 times

I'll try to explain this trick (called law of leftovers, see http://forum.logic-masters.de/showthread.php?tid=357 for a short introduction on this technique): In the two areas with red borders, there are 4 stars alltogether. But all this 4 stars have to be in the top two rows, because there are no other places where they could go there. Therefore the cells of the red borderd ares below the top two rows have to be empty.

One more pointer: http://www.logic-masters.de/Raetselport ... &chlang=en: Tipp 2 and 3 explain this technique (unfortunately I havn't translated it yet into english, so it's only available in german, but maybe the pictures are allready sufficent.)

[Calavera]

ETA in this case is '[post] edited to add' (but, I haven't seen that usage either )

Thank you. Never would have guessed that.

[dickoon]

Thanks, gang. It was indeed "Edited to Add", which would have made a lot more sense if I had indeed edited the post in question to add the comments rather than making another post with the comments in, but I clicked "Quote" instead of "Edit".

This needs more thought, but I have some starting-points to work from. I appreciate the help.

[berni]

I'd like to inform you of a change for the Arukone: From now on, it might be, that some cells remain empty. In the Ü-Rätsel this is true from first of April on, because the puzzles of a month are calculated at once. You can check this at this weeks Preisrätsel.

[detuned]

I was already solving the Arukone under that assumption!

[ronaldx]

Berni - can I request a change to the Tapa applet?

it would be extremely useful to be able to place a 'cross' through the clue squares - otherwise it's very hard to check the solution.

Thanks for your time
Ronald

[berni]

Berni - can I request a change to the Tapa applet?
it would be extremely useful to be able to place a 'cross' through the clue squares - otherwise it's very hard to check the solution.

Good idea. I added it to my todo list, but it might take some time until I'll implement it.

[GaS]

Hi all,

it's very strange my time in today's laser: it was 2min 51sek but it is registred as 6min 51 sek...

GaS

[detuned]

There are going to be some sickening times on that sudoku - my solve felt a little slow and clunky, as strange as that probably sounds.

I wonder if (and indeed by how much) the 1 minute barrier will be broken...

[motris]

It may just be a slow and clunky puzzle at points since I'd describe my solve in the same way.

My one complaint today would be the lack of functional note-taking in this sudoku applet, which is surprising since other apps like the Vergleichssudoku (or kakuro or japanese sums) at least let you put in pairs (or partial pairs) with the minus key which I like and make use of all the time. Having no such functionality on this applet was a slightly unwelcome surprise. It seems one can only do a # press to put in a candidate grid, type whatever numbers you want, and then a sloppy set of steps to remove it (unless a singleton where # will resolve it back to a placement) so this to me is mostly useless.

[berni]

it's very strange my time in today's laser: it was 2min 51sek but it is registred as 6min 51 sek...

The logs tell me, that you pressed "START" at 23:18:43 and you submitted at 23:25:36, which is 413 seconds. The applets tells me, that you needed 411 seconds, which sounds quite reasonable. Maybe, you remember the 2min 51sec wrong?

My one complaint today would be the lack of functional note-taking in this sudoku applet, which is surprising since other apps like the Vergleichssudoku (or kakuro or japanese sums) at least let you put in pairs (or partial pairs) with the minus key which I like and make use of all the time. Having no such functionality on this applet was a slightly unwelcome surprise. It seems one can only do a # press to put in a candidate grid, type whatever numbers you want, and then a sloppy set of steps to remove it (unless a singleton where # will resolve it back to a placement) so this to me is mostly useless.

Well, yes, I already read your complaints in you blog. A note why I implemented it this way: The a-b feature is meant to note ranges (not pairs). Whereever ranges make sense, they are implemented; but in a normal sudoku there is no use for ranges. Of course, whereever ranges are available one can (miss-) use them to denote pairs.

[GaS]

it's very strange my time in today's laser: it was 2min 51sek but it is registred as 6min 51 sek...

The logs tell me, that you pressed "START" at 23:18:43 and you submitted at 23:25:36, which is 413 seconds. The applets tells me, that you needed 411 seconds, which sounds quite reasonable. Maybe, you remember the 2min 51sec wrong?

Yes, maybe I remember wrong, the puzzle was so easy (also for me) that it sounds strange (to me) I solved it in more than 3 min.

Thanks very much for the attention and sorry for the inconvenience.
Gabriele

P.S.: sorry for my bad English

[detuned]

It may just be a slow and clunky puzzle at points since I'd describe my solve in the same way.

My one complaint today would be the lack of functional note-taking in this sudoku applet, which is surprising since other apps like the Vergleichssudoku (or kakuro or japanese sums) at least let you put in pairs (or partial pairs) with the minus key which I like and make use of all the time. Having no such functionality on this applet was a slightly unwelcome surprise. It seems one can only do a # press to put in a candidate grid, type whatever numbers you want, and then a sloppy set of steps to remove it (unless a singleton where # will resolve it back to a placement) so this to me is mostly useless.

I'd argue that so long as the sudoku is easy enough, then there are no need for any external notations. I try to do this for any newspaper puzzle I do, and I reckon I can sort of get by even with the hard killer sudokus. Although naturally this slows things down a bit. I don't think this puzzle was slow and clunky - the structure of the givens felt that puzzle was nearly "trivially" solvable - you could just look at the right bit of the grid, and instead of filling in cells you could fill in regions all at once.

I had no idea you could do that thing with pairs in Vergleichssudoku though. That will certainly make things easier than doing things in my head!

[motris]

I'd argue that so long as the sudoku is easy enough, then there are no need for any external notations. I try to do this for any newspaper puzzle I do, and I reckon I can sort of get by even with the hard killer sudokus. Although naturally this slows things down a bit. I don't think this puzzle was slow and clunky - the structure of the givens felt that puzzle was nearly "trivially" solvable - you could just look at the right bit of the grid, and instead of filling in cells you could fill in regions all at once.

I had no idea you could do that thing with pairs in Vergleichssudoku though. That will certainly make things easier than doing things in my head!

I agree in principle, and there was a time long ago before any good sudoku applets existed where I solved everything mentally, but now I'm lazy and am simply trying to replicate the way I solve on paper which with sudoku is just for competition where shortest time is the goal. I can't exactly do this with Nikoli's app, since the notes don't get on the grid lines which is 10x better for spotting things in the important rows/columns, but I've accepted "hold spacebar and hit number" as a pretty friendly interface where it pops up. "hit minus and hit number" is ok here where one can do it, but I guess I'm not using the intended function.

I think most of my interface complaints have to do with functionality I use in one applet and can't find in another. For example, I can't mark "takes a letter but I don't know which" in a Buchstabensalat, as I do in a Japanese Sums. A circled cell would be fine but I don't think I can do that, even though its helpful information. When I categorize puzzles I put those together as there is +/- cell information, then definite cell information, so encountering slightly different interfaces is off-putting. Having a KillerSudoku and Vergleichssudoku where I can do a thing and it works for me is ok but then Sudoku where the "fix" is not available is bad.

Anyway, these are nit-picks and I'm still really appreciative of all of Berni's work so don't take this as being very negative. I still most eagerly await solving each day's puzzles and seeing how I rank.

Regarding slow and clunky, I can't comment more explicitly as the puzzle is on-going, but there are 2 or 3 ways this pattern can resolve in the end. In both important directions, the most difficult form appeared. So you can be searching for awhile without making quick progress, at a stage you think you should be, and hence a pattern that should be an easy 1 minute puzzle (and could be on paper) has a few stall points to make it more of a 1m20s.