All the answer keys are simple and consistent this time. So we don't expect much manual overrides, but if anyone feels they didn't get points because of a formatting problem, they should post here or send a message to Grant.

While posting here, please don't post the answer, fully or partially - just the puzzle id is enough.

Alright, finished the test a little while ago with a score of 8/18 correct. Not a terrible performance, but of course I hope to improve in the future.

As for the puzzles, I'll leave specific comments to after everyone has competed, but the ones I solved were Classic 1-3, Easy Shikaku, Easy Even-Odd, Easy and Hard Cipher (Hard Cipher was quite good) and Easy Sum. I'm looking forward to doing the rest though, including the Hard Star Fillomino, which looks to be worth every one of those 20 points.

Thanks for the test Grant and Palmer. One question though: will you eventually say who authored which puzzles? I want to see if my guesses as to which was which are accurate.

Hi, I've made a really stupid and costly numpad-style mistake. I'm not sure whether you can do a "manual override" on such thing :)

PS. It's funny how nikoli made me to dislike several puzzle types. I never really thought that fillomino can be so entertaining/interesting, especially since in general I dislike puzzles with numbers/digits.

Yep, we figured it was a Numpad typo. You should already have those (hard-earned) points.

hey i think in the test the shape filimino is wrong...please check the top one

ok srry i dont no then.
thank u.

hey i got it just did silly mistake ..

Ok - I'm deleting your original post. Players may get confused.

That was fun. Sadly my sloppy handwriting cost me 7 points on the shikaku fillomino. Misread one of the digits. Obviously didn't think that digit could go there as it clashes with a given.

Great Test. Many puzzles that weren't super hard. I did enter my answer for the easy Sum Fillomino on the greater-than answer key (they're on the same answer page) Check for me?

Personal frustration on being 1 second slower than desired in typing the last answer. Oh well. This happens.

It's hard to rate this test. Normally I use an internal scale that starts at 5-6 (or lower) for my least favorite puzzles, up to 10 for my favorite. While that would work within everything here, it would not do justice to this test relative to others. I don't have any least favorite puzzles here - nothing that solved in an inelegant or unfriendly way. All I saw were quality puzzles and a lot of really creative designs.

In a few cases I think I could pass the Pepsi challenge and ID the designer, but even with these guesses I have both Grant and Palmer to thank for my favorite puzzles. A superstar team with an excellent test. Thanks.

Re: Para
mathgrant and I knocked our heads together on this for awhile, but in the end we are giving you those 7 points. Good job on your performance too.

Re: TroyS
Normally I'd say "I'll tell Deb and he'll fix it when he wakes up", but it appears to me that you would have had the answer wrong anyway. Look at the last few entries of column B. Sorry. Glad you enjoyed the test.

Re: motris
Thanks a lot for the glowing review and the blog recommendation. :D I think mathgrant and I may have an informal contest on our blogs to guess who wrote each of the puzzles after it's all done, so maybe you can try that out for fun. Great performance too. You may know what I mean when I say I thought you were in trouble 55 minutes in... but then I don't know the whole story. Your classic speed floored us both.

Edited by MellowMelon 2011-06-05 4:59 AM

I've changed in any case.

Re motris: I am so, so, so, so very excited that you liked this test so much! It means so much to me, someone whose only experience in the competition setting is your 20/10 test on which I scored a meager 777, having you, a grandmaster of construction and solving alike, say this about a test to which I contributed half of the puzzles. Honestly, though, even if I made the puzzles, Palmer is probably responsible for most of what made my puzzles come together along with his to be an actual test, and I think he deserves far more of the credit than I do. (I only allow myself to have top billing on the cover page of the IB because my name comes first alphabetically. :) )

If I don't post here much, it's not because I don't care about this test and what you guys think of it; it's because Palmer keeps saying the same things I would say, except more authoritatively. *laughs*

Edited by mathgrant 2011-06-05 5:09 AM

I think there is a lot to be said about how having a good co-author makes one's own work better; so I'd never single out recognition of one author here over the other as it's clear the combination of authors and styles worked better here as a test than either alone would have been.

You co-wrote Mutant Sudoku with Wei-Hwa Huang, so perhaps I should trust you when you say that. :P

I typo'd my greater-than (bottom) entry (should have been a xxxxxxxxxx before the last 2 digits). I'll take a manual override if they're still available :).

Edited by debmohanty at 2011-06-05 8:45 AM (since the answer key is partially revealed)

That one ended up being harder to decide on than Para's error. But you now have those points too.

Thanks! My puzzle for you ;).

Excellent stuff all around, as expected. Only thing that could've been better is if I was able to count to numbers above 3. :)

Message to mystery_boy

Please introduce yourself in the forum.

Message to cmd2do

Please fill-in your correct country and realname.

Thanks. Seems a bit odd to confuse those 2 digits, I know. But if you'd seen the handwriting you'd get it. Also got commented on it by judges during the WSC last year.

The possibility of having handwriting that can confuse those two was easy to accept. The issue of how plausible it is to have a wrong solution with a digit swap like that took a bit more thought. Not that we're accusing you of lying about the handwriting, but if anyone else submits the same solution...

More generally, shouldn't there be a consistent rule about mistyped solutions (those in the right box but which don't match the key)?

Everyone is essentially doing the test on trust since it would be easy enough to get a friend to help with a couple of puzzles, or maybe go to an internet cafe and view the PDF in advance from a different IP address (or even run a password cracker), so in principle I don't see anything wrong with allowing people to ask for typos to be corrected. If people want to cheat they can do so anyway after all, and it's usually pretty obvious if they do.

However in practice we all know it's faster to do the puzzle without entering the key so clearly getting the key correct and checking takes time, so isn't it unfair to penalise those who do take the time to check by awarding points to those who got it wrong? In general I'd have thought it would be better to require the key to be correct on the basis that typing it in accurately is "part of the test", however trivial a part.

If you do allow corrections, however, shouldn't they be applied consistently? E.g. allow a single digit only to be deleted/inserted/substituted if accompanied by a promise that it was a typing error not a puzzle mistake - that might be a reasonable rule for example. However some puzzle-based judgement as to how likely the person is to be lying (as is taking place here) is surely an awkward precedent to set. For full disclosure I had my single digit typing error rejected for correction, and I'm sure it was much harder to be sure of than those above, but I do think it's a reasonable question to ask generally. If it's done on "likelihood that player is lying", what is the threshold for that likelihood?

I'm always firmly in the middle of the results table and I really don't mind whether I personally get 4 points more or not, but wouldn't it be a good idea to have a consistent rule?