croco-puzzle

[motris]

I hadn't checked back on that thread - I feel quite flattered I was pacemaking Para for a while, he's definitely a solving tier or two above me (although I will add I've had a few careless results recently!)

Thomas - re kakuro, presumably Jason (usuually Ziti) isn't on croco yet ...but the H.Jo evidence seems pretty conclusive.

I also take that melon is Palmer. Seems to fit with his solving abilities!

Yes, melon is the same melon we know and love. I was trying to make sense of H.Jo's story about his name. I think it's something like because deu is a shortened part of deutschland either the site didn't like it or he didn't like it but these translation engines are sort of rubbish and I can't parse the crucial sentence well enough.

[deu]

Yes, I'm kirarin I knew the existence of croco-puzzle last year,
but hesitated to start since I had no knowledge of German.

Some puzzles are "classic" in WPC but not so well known in Japan
(Tents, Magnets, Star Battle, etc).
I started croco-puzzle to get used to solving such puzzles.
It is very exciting that many top-level solvers join this site!

About my name:
I usually use "deu", but I didn't like to use it in a German site,
since it can be confused with "Deutschland".

[drsteve]

I though Pyramid puzzles only required different digits in the bottom row. Seems I was wrong...

[motris]

The rule is actually a bit more complicated. If the row is white, it must have at least one repeated digit. If the row is gray, it has no repeated digits. Recently we have not had any white rows in these puzzles, but who knows when they will show up.

[drsteve]

The rule is actually a bit more complicated. If the row is white, it must have at least one repeated digit. If the row is gray, it has no repeated digits. Recently we have not had any white rows in these puzzles, but who knows when they will show up.

Cheers Tom. It cost me a 5 minute penalty today. But hopefully not in the future

[afcwarren]

If the row is white, it must have at least one repeated digit...

I thought it was exactly one pair of repeated digits. Thanks for the tip.

[motris]

I thought it was exactly one pair of repeated digits. Thanks for the tip.

I've certainly seen in the preisraetsels that you can have multiple pairs of digits (I cannot remember three or more of a digit but I believe this is allowed). The simplest thing to remember is this: If a row can be gray (no repeats), it is, otherwise it is white.

[jbclements]

Follow-up to earlier note on Firefox 4 incompatibility w/ croco-puzzle: I'm sad but not surprised to report that the release version of Firefox 4 appears to display the same issues as the release candidate; viz., that the second and subsequent puzzles after application startup are insensitive to keyboard input. It looks like some kind of focus issue.

John Clements

[drsteve]

OK, that was an absolute mare with the magnets. Could somebody please record some worse scores? Please?

[PuzzleScot]

OK, that was an absolute mare with the magnets. Could somebody please record some worse scores? Please?

I tried, but didn't have that much time to spare! you more than made up for it on Pills though!!
Welcome to the +/- club, where every day is one of each

[drsteve]

OK, that was an absolute mare with the magnets. Could somebody please record some worse scores? Please?

I tried, but didn't have that much time to spare! you more than made up for it on Pills though!!
Welcome to the +/- club, where every day is one of each

Well I had to do the pills quickly as I'd wasted so much time on the Magnets. Not sure how it was so quick mind you...

[ColinMacLeod]

Follow-up to earlier note on Firefox 4 incompatibility w/ croco-puzzle: I'm sad but not surprised to report that the release version of Firefox 4 appears to display the same issues as the release candidate; viz., that the second and subsequent puzzles after application startup are insensitive to keyboard input. It looks like some kind of focus issue.

John Clements

I use FF4 at home without any issues. I'll double-check my version number and Java version. Windows Vista Ultimate, 64-bit.

[drsteve]

OK, how the flippety-flip can people do Heyawake that fast?

Presumably there must be a logical step that I'm missing, and I was taking time to mark the empty squares, but what's the secret? (After today's puzzle has closed, of course)

[detuned]

ditto that hitori haha. Granted, I was interrupted for a minute or two whilst solving it, but I still ended up making a guess. These hitori are pretty foul compared to what you get on nikoli, I can't get any intuition going with them (as I still can on, say, the heyawake ).

[drsteve]

To be honest, these are the two Nikoli puzzles that I basically avoid - find them a bit tedious. I find the croc hitori easier than nikoli - often there is only one next logical step, but it generally doesn't involve a vast logically leap. The croc Heyawakes all seem to have this one square box theme and very low numbers - guess I need to practise them a bit more.

[PuzzleScot]

There appears to be 2 ways of doing hitori.
Plan A) Solve logically - look for values between duplicates that must be in, pair + single in one row/col, etc. Slow, but sure.
Plan B) Guess, and see where it takes you!

I just redid it using method 2, and was done in 1'49 (including a 30 second check at the end!)
Chances are the wrong path would have shown up in 30s or so, so a 2 min time still easily possible after clearing.

But Heyawake in under a minute?!? My 2nd attempt was still at the 5 minute mark. I don't see a Plan B method for these...

[damn I hate those fehlversuchen, but they're a good lesson on checking your answer!]

[drsteve]

[damn I hate those fehlversuchen, but they're a good lesson on checking your answer!]

Heyawake is the worst for checking though. In the past I've done if carefully, hit submit, both on croc and Nikoli, only to find one mistake in the middle of the grid.

Now, I just hit submit and trust to luck. Worked today, might not work tomorrow.

[sknight]

I really find the Hitori on croco-puzzle sort of annoying, but that's also a question of basically unlearning the things that I have learned to look for on Nikoli. Nikoli goes out of their way to minimize blind searches for "easy" contradictions. The (presumably computer-generated) ones on Croco-puzzle are all about obvious contradictions. You'll see for example a 4 in between two 6's, so you circle the four. Then you find another 4 in the same row to black out, and circle a 7 next to that. Then you find a 7 in the same row as the other 7 and so on. It's just a chain of bouncing around the puzzle, with very occasional nods to keeping the black squares from spanning.

[motris]

Heyawake is hard for checking, and worse for Fehlversuche. I can "tweak" a Heyawake, but only if as on Nikoli.com I can occasionally be safe if I don't catch a third room constraint breaking. I don't feel it safe to tweak on croco-puzzle. So today, when I messed part of it up not once, but twice, I had to keep scrubbing all the way over and lost a ton of time. It's still a quicker puzzle than in drsteve's experience but you have to learn some of the basic patterns that pop up on croco-puzzle, particularly how to work around 1xN rectangles since the generator is in love with them.

[willwc]

OK, how the flippety-flip can people do Heyawake that fast?

Presumably there must be a logical step that I'm missing, and I was taking time to mark the empty squares, but what's the secret? (After today's puzzle has closed, of course)

Don't know that there's much of a secret on that puzzle, other than keying on some common patterns with low-numbered boxes. I also mark empty squares when I solve on Croco so that I can be sure when I have a complete solution--it might cost a few extra seconds, but it's well worth avoiding the penalty for leaving one square out in the middle of the grid, which you also seem to be familiar with. For what it's worth, I tend to solve the opposite way on Nikoli, as there's no penalty there for incorrect/incomplete solutions and I'm already a bit behind the curve in terms of mouse speed/accuracy.

[ronald]

Do we know who ChiliKing is?

[detuned]

OK, how the flippety-flip can people do Heyawake that fast?

Presumably there must be a logical step that I'm missing, and I was taking time to mark the empty squares, but what's the secret? (After today's puzzle has closed, of course)

Don't know that there's much of a secret on that puzzle, other than keying on some common patterns with low-numbered boxes. I also mark empty squares when I solve on Croco so that I can be sure when I have a complete solution--it might cost a few extra seconds, but it's well worth avoiding the penalty for leaving one square out in the middle of the grid, which you also seem to be familiar with. For what it's worth, I tend to solve the opposite way on Nikoli, as there's no penalty there for incorrect/incomplete solutions and I'm already a bit behind the curve in terms of mouse speed/accuracy.

Ditto marking empty squares on croco. For what it's worth I think I've gotten fairly good at seeing when there are 3 empty rooms in a row, and the fehlvers I have had on previous heyawake have been because I single clicked in a shaded square instead of double clicking a cross. I also find it slightly disorientating that the shaded squares don't quite fill a cell. Makes it a lot less easy to spot when you are potentially closing off a region...

Whilst we're on the subject, for the smaller heyawake at least, I'm trying to learn to do it sans blanks on nikoli. There's no way I'd be 4 seconds behind H.Jo on a heyawake there! (5 today )

Obviously wait til it's finished, but I'd be interested to hear how anyone else solved that Gebietssummen....

[dickoon]

Obviously wait til it's finished, but I'd be interested to hear how anyone else solved that Gebietssummen....

With great difficulty, a second attempt from scratch which went a lot more smoothly than the first go round and a stupidy-stupid Fehlversuch right at the end.

(Specifics after it has closed out, obv.)

[nickdeller]

Obviously wait til it's finished, but I'd be interested to hear how anyone else solved that Gebietssummen....

30% logic and 70% bifurcations. If there's a forced logical path through it, I'll also be interested to hear it!

[PuzzleScot]

Obviously wait til it's finished, but I'd be interested to hear how anyone else solved that Gebietssummen....

42 minutes wasted on a false assumption. 18 minutes when redone from scratch. The more I stared at it, the more logic I could see, but damned hard, yes.
Other than the slowly emerging logic, I echo Nick and Chris' comments. Stinker.

Addendum:

Obviously wait til it's finished, but I'd be interested to hear how anyone else solved that Gebietssummen....

If there's a forced logical path through it, I'll also be interested to hear it!

I've written a logical solve path for this one, as it was irritating me. Will post here when no longer current