2005 Google U.S. Puzzle Championship
Fortran IV writes "Registration is open for the 2005 Google U.S. Puzzle Championship, Saturday, June 18. Two winners will join Team USA at the 2005 World Puzzle Championship in Eger, Hungary (tourist info here if you read Hungarian). If you're the type who plays 12 simultaneous chess games in your head while debugging code and memorizing logarithm tables, you might have a chance of teaming up with last year's champ Roger Barkan (previous Slashdot coverage). If you just like puzzles, register here for the most intense (and fastest) 2-1/2 hours of the year. For a faint shadow of the real thing, take the practice test, which Barkan can probably complete in about 8 minutes; for a true challenge, the complete 2004 test is still available."
Kim Peek can't compete, he's able to do a few tasks extremely efficiently, very very well, and his brain houses an enormous amount of archived data but any challenge outside his narrow range of useless natural abilities and he falls flat on his face. He wasn't born a creative genius, just a genius.
--
Random Signature #1
Generated by SlashdotRndSig via GreaseMonkey
2004:
- instructions: endeavor
- test: xcode6fire
2005:
- instructions: hello
- test: kronos
Sorry - you fail already.
Did you not read the bit where it said "Password (see notes below)"?
And below there was a link to follow to get the password. Clever password it was too - very apt for a challenging puzzle; a nice reference to those who came before them.
The password for the practice test is kronos.
There's a link to the password below.
Not *too* painfully obvious.
One building a size larger in front of another removes it from view. So to see 3 buildings, possible combinations are:
1243 - 4 blocks 3
1342 - 4 blocks 2
2341 - 4 blocks 1
1324 - 3 blocks 2
2314 - 3 blocks 1
2134 - 2 blocks 1
The "4" requirement makes this one pretty easy, as you can see from the above list that only one possible combination has a "2" in the third slot.
Scissors are on the allowed materials list.
I competed myself last year; I submitted 8 correct answers and no wrong ones, finishing well out of the top 50%; I eventually solved 18 of the 25 puzzles, but only over the course of several days.
I believe that the people who solve these in 2-1/2 hours are doing nearly all the work in their heads, whether it's a rolling block puzzle or a crossword, then simply scribbling down the entire solution at once. A fantastic memory--swift, accurate, and strongly visual--is a definite advantage in this competition (an advantage I don't have).
More than a little mathematical background isn't unhelpful either. For one puzzle I did solve last year, #19 "Point Pairs", it's helpful to know more Pythagorean triplets than 3,4,5. I did it rather quickly (that is, in under an hour) but it was one of the 5 least solved puzzles last year.
What little advice I can offer:
I figure by 2030 or so my 6-digit UID will be something to brag about.
for the question that asks you to find the matching mirror image, tear out the original, flip it over, hold the page up to the light, and find the one that doesn't match. . .
Google didn't start this competition, they just started sponsoring it a few years ago.
There is a checkmark on the registration page that you check if you would like to receive notices of employment opportunities.
Since they give the directions to the puzzle in advance, most or all of them could also be coded in advance.
Oh yes? For the 2004 test, what kind of process would you create from the directions, "How many circles are either shown or implied by the diagram?"
Or take number 19, "Corral". From the directions and the example, how would you anticipate that the actual problem used a hex grid instead of a square grid, requiring a very different approach?
And even for the ones where the instructions are fairly clear, just how many of them could you really code (and debug!) in 24 hours? Number 22 for instance, where your program has to allow for competition-time entry of several different factors (the puzzle background, the totals, the circles for horizontal and vertical pairings).
For now, the human brain is still the most versatile (and most rapidly reprogrammable) problem-solving computer.
I figure by 2030 or so my 6-digit UID will be something to brag about.