chat prefs...
|
12:16 am
mommanine
seemed like a pretty easy "hard"
12:39 am
jawilke
very easy "hard"
1:16 am
Python Taz
someone that does these in two minutes, equals lazy cheaters! plan and simple.
1:16 am
Python Taz
someone that does these in two minutes, equals lazy cheaters! plan and simple.
1:16 am
Python Taz
someone that does these in two minutes, equals lazy cheaters! plan and simple.
1:18 am
Python Taz
it's a brain teaser not use autofill or some cheat program.
8:08 am
raggedyandy3478
neone else get stuck?
8:08 am
sprite
not so far, doing all the puzzles I missed during my vacation.
8:08 am
dan2bit
I did for a minute - found a singleton in the rightmost column
8:09 am
raggedyandy3478
im doing todays, and im stuck, which one is the singleton?
8:10 am
raggedyandy3478
sorry, what is a singleton, thats the better way of phrasing it
8:11 am
sprite
I guess it is a cell for which there is only one value left?
8:38 am
Andy
done after a very long interruption.
8:38 am
Andy
used greens. Bran fuzzy. Shouldn't have had that martini last night.
8:40 am
lk911
nice...easiest 'hard' in a coupla weeks!
9:27 am
UnikeTheHunter
22. Going green.
9:28 am
UnikeTheHunter
Okay, there's a pair I didn't see w/o greens.
9:30 am
sprite
doing August 8's now. August 7's was tough!
9:34 am
UnikeTheHunter
Done. Easy with greens. With a little more persistence, I should have done it w/o greens.
9:34 am
sprite
well done anyway
9:38 am
drwho
Difficulty score 22, well below the average score of 31.38 for hard puzzles. I thought yesterday's hard was easier. Look for a hidden pair.
9:38 am
sprite
dr. who, was August 7 a real tough one?
9:39 am
rosybifocals
this could be a silly question, but how do you rate the puzzles?
9:41 am
drwho
I didn't think so, and my program only scored it a 33. See:
http://www.pauls-pc-repair.com/sudoku/in
dex.html
for an explanation of scores and a link to a table of scores I have generated for the IronSudoku puzzles.
9:42 am
UnikeTheHunter
August 7 was tough. It had an x-wing not too hard to find It contained a number "which doesn't belong." I can only explain that by saying a bunch of pairs was complete without that number. And I'm not even sure if my logic is correct, but it worked out that I could eliminate the stray number. Even so, I had to guess later on.
9:42 am
sprite
*sigh*
Must be really out of training then... I did a lot of master rated puzzles with pen and paper... but they are much easier than the ones here.
9:43 am
drwho
Many sites limit their puzzles to ones which can be solved by a set of well known solving techniques. Apparently here that is not the case for the Expert puzzles.
9:44 am
rosybifocals
oooh... very interesting. thanks!
9:45 am
sprite
I think so too. I have these booklets from the tazuku series, they rate from 4 to 6 stars and they rarely give me real problems. But the experts here can.
9:46 am
drwho
If you're looking at my web site you might be interested in trying the online rating program. It generates slightly different scores than the program I use here.
9:47 am
sprite
you gave me that link before and I did check it out. Very interesting site. Perhaps I should check some of the 6-starred puzzles from the booklet and see how they turn out in your rating program.
9:50 am
drwho
I give a number calculated by almost the same method I use to rate puzzles here, plus I rate it Easy, Medium, Hard or Expert by the same method Dr. God uses here.
9:51 am
sprite
I think I will check out a few of the socalled tough puzzles from the booklet. I am really curious whether I am better with pen and paper than on line, or that the puzzles here are tougher (the expert ones anyway). I think the latter.
9:55 am
drwho
If I wanted to sell a book of Sudoku puzzles to as many people as possible, I think I would limit the difficulty so I wouldn't scare away potential customers.
9:58 am
drwho
Speaking of publishing puzzles, my next project will be to write a program to generate Sudoku puzzles.
9:58 am
sprite
I am not sure about that. They have booklets with 1-3 stars, 1-5 stars, 3-5 stars, 4-6 stars and so on. So I would think that the highest levels should be really challenging, to keep it interesting for the more experienced player. They will want them as regular customers.
9:58 am
UnikeTheHunter
Let me know how that goes.
9:59 am
sprite
nice drwho! Keep us informed!
10:01 am
drwho
Ok, I won't be able to do it in JavaScript (which is how the online solver and rating program is written), but I may be able to put a CGI interface on a Linux program and make it available online. I still have a lot to learn about web programming.
10:02 am
sprite
sounds like a challenge in itself.
So... caught up with the puzzles, on to today's :)
10:05 am
drwho
My online solver can almost do it now. Just tell it to solve a puzzle with no clues. Your browser will think my program is running too long, but if you let it run long enough it will give you a filled in grid with a proper solution. The trick is to figure out what to use as clues for that "solution", that and to put in some randomness to the solutions it generates.
10:10 am
UnikeTheHunter
Does that produce only unique puzzles?
10:12 am
drwho
No, it doesn't actually produce puzzles (no clues!), just a solution (all the numbers filled in).
10:14 am
sprite
but I can see how that will lead to creating solvable puzzles. It would suffice to let it create a complete puzzle and then go back enough steps to make the puzzle challenging AND solvable.
10:14 am
UnikeTheHunter
So if you take that filled in grid and eliminate a bunch of numbers, then can you tell if this is a unique puzzle?
10:14 am
drwho
The first step in generating a puzzle is to generate its solution, then figure out what to give as the clues and then verify that the solution is unique.
10:18 am
drwho
Verifying uniqueness is simple, generating the clues, or deciding which numbers not to eliminate, could be tricky.
10:19 am
drwho
It is conceivable that I could give all but 4 numbers of the solution as clues and still not have a unique solution (anybody heard of unique rectangles? :-).
10:19 am
sprite
just what I was thinking. What clues do you give to make it possible to solve the puzzle, yet tough enough to be interesting... and there would the rating help, I think. Which solving techniques are needed.
10:20 am
sprite
I used unique rectangles pretty often. They are purely based on the assumption it is a GOOD puzzle, one with only one solution :)
10:21 am
sprite
*use, not used
10:21 am
drwho
Right now, I don't have any idea on how to select the clues with the goal of producing easy or hard puzzles.
10:23 am
sprite
it might be connected with the rating. I am really looking forward to hearing how you tackle the problems, drwho!
10:26 am
drwho
When I get it working I will at least put up an explanation of how it works on my web site. I hope also to integrate it into my online solver so it will be able to solve puzzles you enter or generate random puzzle for you to solve.
10:26 am
sprite
that would be really great. I hope you get it working the way you want!
10:32 am
sprite
today's wasn't bad at all. No greens needed, and no techniques other than the obvious.
2:56 pm
sandbox74
breezed through this one alright
3:09 pm
raerae7133
hmm. it was easy up to a point, but now I'm stuck
3:55 pm
ypsigirl
and yet still at newbie at 237, humpf!
4:37 pm
moedog
hi everyone
4:46 pm
moedog
lotsa - whaddayacallem? triples?
4:53 pm
moedog
ah - found a 9 and now i'm on my way!
10:15 pm
raerae7133
okay. back to try this one again