« Back to Discussions

Small griddlers not solvable by SLS

Combination View Flat View Tree View
Threads [ Previous | Next ]
toggle
Small griddlers not solvable by SLS
Answer
2/4/04 5:45 PM
I'm looking for griddlers which require MLS, that is not solvable my SLS only.

I did not found any 2 color (B/W) griddler ever requiring MLS. Tried griddlers up to 60x145.

I did found several MLS multiple-color griddlers. The ones I found are quite large. I'm looking for griddlers as-small-as possible, requiring MLS.

Thanks in advance

Re: Small griddlers not solvable by SLS
Answer
2/5/04 10:11 AM as a reply to jvloenen.
>I'm looking for griddlers which require MLS, that is not solvable my SLS only.
>
>I did not found any 2 color (B/W) griddler ever requiring MLS. Tried griddlers up to 60x145.
>
>I did found several MLS multiple-color griddlers. The ones I found are quite large. I'm looking for griddlers as-small-as possible, requiring MLS.
>
>Thanks in advance


Naturally, you won't find one in Griddlers because the checking system uses SLS. However, half a year ago Malcolm posted the following message
------------------------------------------
I tried to post this puzzle

xxx....xxx
...xxxx...
xxx....xxx
...xxxx...
xxx....xxx
...xxxx...
xxx....xxx
...xxxx...
xxx....xxx
...xxxx...

But the auto-solver insists that this puzzle cannot be solved.
------------------------------------------
If you want to read more about it go back to June 1st to 6th, 2003.
The Subject is The Auto Solver doesn't have the Human logic

Hagit

Re: Small griddlers not solvable by SLS
Answer
2/6/04 12:01 PM as a reply to jvloenen.
>>I'm looking for griddlers which require MLS, that is not solvable my SLS only.
>>
>>[...]
>
>Naturally, you won't find one in Griddlers because the checking system uses SLS. However, half a year ago Malcolm posted the following message

I know about the checking system. I thought other puzzles could be added by hand after checking.

Now I found puzzle 1206 (triceratops), which is not solvable by SLS only I think. Can anyone confirm this?

I try to give the point where I cannot go further, using SLS only
(Using _ for unknown yet, # for black and - for white)

__________________--
__________________#_
_________________-##
__________________-#
__________________-#
________________---#
_________________--#
_________________--#
_______________-####
________________-###
________________-###
_______________-####
_______________####_
_________________---
________________-___
____________________
____________________
____________________
____________________
____________________

No single SLS can bring me further than this.

Re: Small griddlers not solvable by SLS
Answer
2/6/04 5:45 PM as a reply to jvloenen.
It looked a little confusing by your chart. But I think I know where the problem might be. Your line #9 has 4 blacks at the end. (If I'm reading it right.) It should only have 2 blacks. Try changing that.

Re: Small griddlers not solvable by SLS
Answer
12/17/04 5:00 PM as a reply to jvloenen.
[quote="jvloenen"]
Now I found puzzle 1206 (triceratops), which is not solvable by SLS only I think. Can anyone confirm this?


Are you still looking for these puzzles? I think you are right about this puzzle not being solvable by single line solving and now it's possible to show how far I got. At this point I had to start solving by what I like to call 'trial and error'. It wasn't very hard to go on, but I did have to start looking for places that only could be white.

So here it is:
Attachment

Attachments: 1206 by Single line solving.jpg (53.3k)

Re: Small griddlers not solvable by SLS
Answer
12/17/04 7:23 PM as a reply to jvloenen.
I would have started (or continuted) 1206 with the 2 top lines.
The 1st line has: 5
The 2nd line has: 2, 2
Now for the "trial and error" work, starting from the left.

I place 5 black squares on the grid and realize that it is wrong because the longest clue in the 2nd line is 2.


I move the 5 squares to the right. It is wrong again for the same reason:


I skip a few squares because I know now that I'll get the same "error". My next trial will be:


My next move is from right to left, as follows:


According to those findings, my final conclusion is:


Since this is not the first puzzle I solve, I don't really try those moves. I know that I'll get the "errors" in the 2nd line. Hence, my first move will be the last shown in the picture.

If you combine "my" findings to Jeltje's above - the puzzles is in your hand. :-)

Hagit

Re: Small griddlers not solvable by SLS
Answer
12/20/04 7:18 PM as a reply to jvloenen.
Triceratops is SLS solvable. If you look at row 16 (1 3 3), you can fill in a black at row 13.

Hagitm is right that you can continue in the left corner with the technique described, which is not SLS. (I do use this technique myself too, if i'm stuck however.)

SLS technique does look at one line and tries to fill in whites or blacks (or colors of course). It does not do MLS ("fill in a color and start filling another line to see if it blocks, in which case the first fillin happens to become false"), which can be seen as a form of backtracking.

Still I wonder if there are griddlers, 2 colors preferably, which are solvable by some technique (not "guessing"), but not solvable by SLS.

Or can it be proven that: if there is one and only one picture that can be translated into a griddlers, it is always SLS solvable?

Re: Small griddlers not solvable by SLS
Answer
12/20/04 11:48 PM as a reply to jvloenen.
[quote="jvloenen"]Triceratops is SLS solvable. If you look at row 16 (1 3 3), you can fill in a black at row 13.
You are right about that one, I never noticed it and I looked at that line several times :oops: And from that on it is solvable all the way with SLS.

And 'clever' me, I thought to look for the most difficult puzzle (highest points I mean) in this size, to find one that's not SLS-only, but what a surprise, the first one in line turned out to be... right, Triceratops!

I'm quite sure though that there are puzzles that need MLS to get solved. I've used it so many times and I think the solving program knows those 'tricks' as well. The ones I've used it are mostly 50x50 puzzles though. It might be fun to know what's the smallest size you can have, that needs MLS!

Jelga

Re: Small griddlers not solvable by SLS
Answer
2/3/05 1:41 AM as a reply to jvloenen.
It is possible to create puzzles that require MLS to solve, but the site's checker isn't suppose to admit them. Sometimes I can't find the one line that is the key (like the one square in the triceratops picture) especially on big puzzles with lots of little clues, but when I make progress with MLS I realize where the critical line was that would have allowed progress with SLS.

The color ones I think are always susceptable to the computer's version of SLS but because of the color even SLS is harder. The computer probably marks a square as not available to a color when it either places another color in that square or when a pass says the color can't be there. It can keep tabs on that for each color. Sometimes I'll do something similar for the color with the circle or x marks, but with 8 colors that is a partial solution and you end up having to reference multiple lanes at the same time because you don't have enough ways to mark the squares as unavailable to each color.

Re: Small griddlers not solvable by SLS
Answer
2/23/05 10:07 AM as a reply to jvloenen.
Now I understand why I didnt like to make the griddlers on this site. I tried the triceratops and the only fun in solving it I had at the point you described To do the trial and error in my mind, that's where puzzling starts for me The rest of the solving was just counting and filling in. Isn't it possible to open a special group of puzzles: the real braincrackers. It should be possible to make the validation so that a small griddler (never a multi) that is only solvable by MLS comes in a special group? We have Kiddlers, Griddlers, Triddlers, why not add Briddlers?
I would be very emoticon
Fineke

Re: Small griddlers not solvable by SLS
Answer
2/23/05 7:01 PM as a reply to jvloenen.
[quote="fineke"]... We have Kiddlers, Griddlers, Triddlers, why not add Briddlers?...

It will be a real challenge to create those puzzles I think. I'm already writing a small program to verify wether using MLS is the only way to solve a puzzle, but it might take a while to finish ;)

John

Forum Moderators: griddlers_team, elad, Ra100, chefmomster2, domi77, dreamtheater, elimaor, ElinaMaria, Jeltje, sslug, cosmictrombonis, raist.
Please read the Board-wide Policies before you start using this forum.