WPC 2004: My Story
Joseph DeVincentis
After years of trying, I made the US
team for the World Puzzle
Championship. My teammates were:
- Wei-Hwa Huang, who I first encountered online back in 1993 when
he seemed to post in all the few Usenet groups I was still reading,
which in particular showed him to have common interests with me in
puzzles, math, pinball, and roguelike games
- Roger Barkan, who I have known from the National Puzzlers' League, the American Crossword Puzzle
Tournament, and the MIT
Mystery
Hunt
- Jonathan Rivet, who I didn't know at all, though it is possible I
might have run into him playing games on BrettSpielWelt
The short story: The US team won the team championship. I placed 19th
in the individual. Jonathan Rivet,our other first-time team member,
placed 27th. Roger Barkan and Wei-Hwa Huang nearly tied at 4th and 5th
before the playoffs; Roger finished 3rd in the playoff and got to take
home a fairly large but conventional, cup-style trophy. Wei-Hwa
currently has our 1st place team trophy, which is even larger.
Germany had the top two in individual points but their other team
members were in the 40s and they didn't do as well on the team rounds,
leaving them 89 points behind at the end. Hungary placed 3rd, another
89 points behind. The Netherlands were about another 200 behind in 4th,
and Canada put in a strong showing coming in 5th. The individual
playoffs ended with Dutch former champion Niels Roest coming back from
a 9th seed to win, German former champion Ulrich Voigt 2nd, and Roger
3rd.
The British team captain blogged an amazing amount using the hotel's
internet access:
http://www.livejournal.com/users/jiggery_pokery/
He has more detailed score information (though note that some of the
scores changed in minor ways after what he recorded, due to corrected
scoring errors; one individual moved up a few points ahead of my
teammate
Jonathan RIvet, dropping him to 27th).
Repeating a tradition started last year, the puzzle
instructions were placed on the web the week before the event to
give teams even more time to work out whatever questions they needed to
ask. At least Nick Baxter (our team captain) and I attempted to solve
all the example puzzles provided, though some of them were unsolvable
as initially presented. The veterans called this year's puzzles "very
Dutch," meaning that they were almost entirely of the types appearing
in the Dutch magazine Brein Brekers.
When the WPC was held in Turkey some years ago, it was "very Japanese"
with features such as a whole round of Number Place puzzles. This year,
the picture puzzles some of us dread appeared in a minimal quantity;
just one Find the Differences, one Comic Story (put the comic strip
panels in order), one find-the-parts-of-a-picture, and a "jigsaw
puzzle" round which turned out to be edge matching, out of about 140
puzzles.
Tuesday
Everybody arrived Tuesday, with Wei-Hwa, Jonathan, and I meeting in the
Munich airport for our flight to Trieste, along with our "embedded
journalist" Derrick Schneider who promises to write up and sell a story
about the event. The four of us worked together to solve the puzzle of
packing our luggage in the trunk of the Accord driven by Croatian
captain(? some relation to their team, anyway) Valter Kvalić, then
crammed into the car
for the drive to Opatija. After checking in and visiting our rooms, we
gathered in the lobby and played a few hands of bridge with Ulrich
Voigt, while Wei-Hwa greeted other arrivers and introduced them to us.
At one point he introduced us as "here are our new team members,
Jonathan and Joe, and Ulrich" leading to an all-around laugh as we
considered the
idea of Ulrich on the US team.
The hotel was equipped with two kinds of internet access. There were
two coin-operated PCs which gave you 3 minutes per 1 kuna coin (1
dollar = approx. 6 kuna) and just allowed you to use a web browser.
There was also wireless access which was more expensive (100 kuna for a
card that gave 90 minutes of access time) but you could use it with
your own laptop, and upload stuff you had written offline, or download
stuff to read later, etc. I didn't use it, but some of the others did.
At some point late Tuesday afternoon, the packets became available, and
Nick, our team captain, played his role and got us our packets. The
packets consisted of a tote bag with the 13th WPC logo, T-shirt with
same logo, the instruction booklet (more about this later), a nametag,
an issue of Feniks (one of
the sponsors, a Croatian puzzle magazine,
3/4 of which was crosswords and the like in Croatian, but it had some
word searches and logic puzzles we could solve), and a pen with Feniks
logo.
Tuesday evering there was a party with champagne (and some other
horrible green-colored concoction) and snacks, and some music, and
introductions of the team members. Japan brought along the youngest
solver, at 15, and apparently also the shortest solver (not the same
one). There were an amazing number of people I already knew either from
previous meetings or from online:
- David Savitt (Shadow of the NPL) on Canada's team
- Alan O'Donnell of the UK who I know from some math puzzle sites
like those run by Ed Pegg and Ken Duisenberg
- Cihan Altay (who runs a quarterly
competition on a set of vaguely WPC-like, but harder, puzzles) from
Turkey
- Ulrich Voigt from Germany (uvo on BrettSpielWelt)
- and of course Wei-Hwa and Roger on my own team
Also during this time, Wei-Hwa noted that this downstairs area
where
the party was held had a small game area, with foosball, a couple video
games, a pool table, and a pinball machine, Tales of the Arabian
Nights, but the pinball was out of order, to my and Wei-Hwa's
disappointment.
After this was dinner in the hotel restaurant (all meals were buffet
style, except that drinks were only available served by waiters, and
overpriced, but Google took care of those for our team). Throughout
many of the meals
and in-between times, we had the opportunity to watch Wei-Hwa solving
puzzles in his Xeroxed Japanese magazine. He had one of these magazines
from Nikoli which is full of puzzles like we do in the WPC, and in
order to be able to re-solve the puzzles later, he had made a copy of
the entire thing and was solving it on a clipboard, throwing pages away
after he had completed them. He solved Fences puzzles at an amazing
pace. After dinner, it was time for bed for most of our team to get a
real sleep, as opposed to that kind of half-sleep you get during
flights.
Wednesday
Wednesday, we had breakfast in the hotel restaurant, which, like all
the
meals, had a full and varied buffet, with perfectly normal breakfast
food except for the inclusion of pizza. Also, at breakfast they had
normal drinks like water and orange juice available from dispensers,
unlike at other meals.
Later, we were loaded into 3 buses for the excursion which took us to
Poreč and Motovun. Poreč was another tourist town, but larger. They
showed us a Byzantine church (original building, with some
reconstruction) and then we had 1.5 hrs free time. A group of us from
the US delegation stuck together, as we wandered among the shops and
other attractions in
the area they dropped us off. We walked to the coast, where there were
a number of different boat tours available. As we approached one sign
offering a 1.5 hour tour, a man nearby asked if we had the time for the
tour. When we said no, he asked "well do you have money? Because time's
not important." Apparently he was willing to make a tour of any length
as long as we paid him, but we pressed onward. Wei-Hwa gave up in his
quest to find a Polo shirt with "Croatia" printed on it, and settled
for a T-shirt.
Later we saw what looked
like a toy store and some sort of electronics store side by side, with
something inside that attracted somebody from our group. We went in and
it turned out they were parts of the same store -- a
department store which went on and on. There was a grocery store area
(we picked up cold 1.5
liter bottles of water for about 5 kuna) and several other departments
on two levels (which we moved between by climbing a stopped escalator).
The third floor seemed devoted to furniture, so we skipped it. After we
exited through a different door than the one we came in, we ended up
near a casino whose sign we had noticed earlier. We stopped here so
Wei-Hwa could see if there was any pinball in the casino, but it was
pretty small inside and there was none.
We got back on the bus, and the next stop was lunch, at some restaurant
that seemed to be in the middle of nowhere. We had a great and full
meal (appetizers, soup, a main course, and the dessert was rather
similar to what we would have called doughnut holes in the USA, with
powdered sugar on top. I and some of my teammates sat at the table with
the entire Russian team. As was to be expected among this group, at
times when there was no food, several of us pulled out puzzles to work
on. When the dessert was served, one of the Russians seated at the end
of the table was engrossed in his puzzle and didn't notice -- or else
had eaten so much that he was intentionally ignoring it. One of his
teammates placed a doughnut hole on the table next to the top of the
magazine, and when he failed to notice, then put a second one there.
Then they moved one of them onto the puzzle he was solving which
finally got a reaction.
The next stop was Motovun, a reconstruction of an entire medieval town
at the top of a hill which the buses went most of the way up a narrow,
switchbacked road to reach. There was a steeper and narrower part
toward the top that we had to walk up. More chances to buy souvenirs
here, as well as local wine (we passed lots of vineyards on our trip)
and
truffles.
Finally we went back to Opatija, and had dinner, and then attended the
question and answer session for the next day's puzzles. The
instructions and examples for all the puzzles were given to us online
the week before, and then we got a printed, updated version in our
packets. One of the major changes was that a 17-puzzle, 26-minute
individual round was corrected to be a team round, where the four of us
together would work on those puzzles in the short time limit. It still
wasn't enough time for any team to finish. Also, some of the example
puzzles were broken in various ways, and most of these were fixed in
our updated packets. For instance, in one puzzle we were to divide a
grid into two pieces that could be rearranged to form a chessboard. The
first example had 70 squares, and simply didn't work.
In the Q&A
session, all these people from all over the world nitpick the English
instructions for any defect, missing rule, etc.
If you read Chris Dickson's blog, this is where the "not even
diagonally" bit came from. There was a whole section of pentomino
puzzles of various sorts, and everywhere they had been careless enough
not to state each of these things, somebody had to ask if the
pentominos could be rotated, mirrored, or could touch at a point.
In the late evening I broke out my Tichu cards. I think just one game
was played this night, with more on later days. Over the course of the
week we taught the game to a couple people, and played with some
veterans of varying skill levels.
Thursday
Thursday we had breakfast and then, before the competition, was a photo
session. A photographer took pictures of each team and all the teams
together. Then we went in and solved two rounds of puzzles.
This was the 13th WPC. Croatia integrated a number 13 theme throughout
the event. Among other things, all the rounds were timed in multiples
of 13 minutes. Also, they put us under intense time pressure, with only
a few opportunities for anybody to completely finish a round.
The first round featured 13 standard puzzle types, such as paint by
numbers, cross sums, number place, and lighthouses. I began on the
20-point Paint by Numbers, which worked out to form large letters WPC
with a small Opatija below it. Next I jumped to the Radar puzzle (a
grid of rectangular clouds, all at least 2x2, with battleships-style
clues telling you how many cloud parts are in each row and column. This
seemed overvalued at 25 points. I also did a cross sums for 25,
bringing me to 70. I started briefly on the 25-point number place but
decided it was going to take too long, and instantly decided the same
about the 25-point domino hunt on a double-9 set, leaving the lower
valued puzzles. I
managed to finish a zigzag, 1 of 2 minesweepers, the lighthouses, and
an End View (called Easy as ABC here) for a total of 100 of the 200
possible points. A few people actually got perfect 200s here -- perhaps
those who have practiced these standard puzzle types to an extreme.
Wei-Hwa had a bit more, I think about 130, but the rest of my team was
similar to me, and when the results from this round were posted later,
we as a team were 4th or 5th at this point.
Round 2 was a very important team round, with each puzzle counting 4
times normal value and 800 points possible for the team. Again we had
13 familiar or somewhat-familiar puzzle types, but this time only 39
minutes and
we could work together on them. We had agreed the night before to split
up the work in a certain fashion. I had taken the single
highest-scoring puzzle, the most unusual type, because I saw how it had
to be solved. This Common Letters puzzle asked us to arrange 13 names
in a grid so that the various numbered connecting lines showed the
number of shared letters between each the connected names. I knew
that I would need to find the numbers of common letters between all
pairs so I pulled out my graph paper, drew up a grid, and spent the
first 9 minutes counting common letters in all the pairs among 13
names. Then it took me about another 10 minutes to fill in the grid. It
took me half the time, but the puzzle was worth 4x30 = 120 points,
worth half of one solver's time. Wei-Hwa finished his assigned puzzles
quickly and also took a couple others, including one of mine, and we
ended up all checking everybody else's work at the end. We had enough
time to solve all the puzzles and check all but one of them, and were
one of a few teams to get the perfect 800.
After lunch we had 4 more rounds. Round 3 was a "medley" -- which
Wei-Hwa pointed out is usually called a relay in these events. The
puzzles were dependent on one another, so each used information from
the previous one. The first was a picture puzzle where you had to
locate 10 parts in the picture. Their coordinates were used to give the
clue numbers for two battleships puzzles -- standard, except for the
fact that we were told that there were multiple solutions, but only one
that worked for the next puzzle. In this case the dependencies worked
both ways. It turned out that one puzzle had a
unique solution and the other had many solutions. You had to work back
and forth from the next puzzle, a fill-in which needed black squares on
every square that had a ship part in both battleships
puzzles. This was very time-consuming. I finished but only made it part
way through the fill-in, ending with only 30 of 100 points.
Round 4 was another long one, this time with one puzzle type from each
of the past WPCs and one from the Croatian qualifier. A lot of people
found this one tough. I finished a 25 point hidden cars (a list of cars
and a grid of letters, where each row and each column has one of the
cars in scrambled order, with each letter used exactly once), and a 20
point I crossword (a fill-in puzzle where all the I's are already in
the grid, but all the black squares are also marked as I's. I got 14
points on an archipelago puzzle, which had a lot of similarities to one
which appeared in the Feniks magazine they gave us with the initial
packet. I also finished one of two Spokes puzzles and one of two
Elastic Bands (which were new to me). 71 points wasn't all that great
but my teammates did better on this round.
Round 5 was a "sprint" - 13 little arithmetic grids to be solved in 39
minutes, with bonus points for the earliest correct solvers. Each grid
was a 3x3 arrangement of squares to be filled in with the numbers from
1 to 9, with operations between the numbers in each row and column,
with the goal being to make each row and column come out to a specified
value. In this case, all rows and columns in the first grid were to
result in a value of 1, in the second grid 2, and so on. The 13th grid
was also missing the operations (each of which must be used at least
once), and the numbers in this grid were to be copied from
corresponding gray squares in the other grids. And we were also warned
in advance that the solutions for some individual grids were not
unique, but there was only one configuration which would allow the 13th
grid to completely work with 9 different numbers. Roger and Ulrich
finished first, seconds apart, although in the end Ulrich turned out to
have an error. The rest of the bonus went to other teams. Wei-Hwa made
the error of failing to use division in the final square, but otherwise
our team got all the puzzles correct.
Round 6 was the "jigsaw puzzle" round. Having been told we would get 8
nine-piece jigsaws that could be easily distinguished from one another,
and 39 minutes to solve,
we knew this wasn't going to be ordinary jigsaws but rather some kind
of tricky thing, probably edge-matching. And indeed it was. Each team
got a bag of 72 cardboard squares with fish, shells, clowns, and
various other things split in half over the edges of the pieces.
Wei-Hwa made the correct assumption that these would be cheaply
produced with the pieces initially printed together and simply cut
apart, so that the grain on the back of the pieces could be used to
orient the pieces, and the cutting could further be used to test the
validity of matches, and he was correct on all counts. We were the
first to finish,
getting the maximum bonus. After this round, there were copies of all
the puzzles waiting outside for us to pick up. There was also another
issue of Feniks distributed at some point, I forget when, and some
other magazines from other countries that got distributed via team
captains.
Then we had dinner, which we had been asked to wait until 8 before
eating, because another group in the hotel was eating dinner 7-8 and
this would avoid overcrowding the dining room. So we played a little
Tichu before dinner, and after eating went on into the next day's Q
& A session. At the Q&A session we were told that team round 9
was moved to the start of the day tomorrow, to minimize the number of
times they had to rearrange the tables from individual to team
settings. We took the time during/after the session to assign puzzles
for round 9. Another game of Tichu after the Q&A and off to bed.
Friday
After breakfast Friday we went into team round 9. This was a tough,
tight round; 13 puzzle types to be split among the team and solved in
26 minutes! The scoring was the same 4 times normal scoring from puzzle
2, so it was potentially worth 800 for the team for a complete
solution. I solved three easy-to-medium mathematical puzzles: Darts
(choose three numbers from a given set that add to a given total; four
of them to be solved), Animals in Equations (a 4x4 grid with one of
four animals in each space, with sums for all the rows and columns, and
the number represented by each animal to be determined; two of them to
be solved), and Arithmetical Progressions (a 7x7 grid of digits; along
each edge of the grid an arithmetic progression needs to be written,
with the digits of each number taken from the corresponding row or
column of the grid, and all the digits used). Wei-Hwa failed to solve a
hard Balancing Act puzzle (place the weights 1 to 13 in specified
places on a set of balance arms so that each arm balances) so we did
not do so well here, solving only 7 puzzles -- but nobody finished them
all.
Back into sequence with round 7, which was a set of various puzzles
each involving either pentominoes or arrows. I solved 6 puzzles for 85
points and the rest of the team had similar scores. Next came the
optimizers round. Three puzzles on which to try to obtain the highest
scores possible. I know from experience (even if not in this particular
setting) that in doing optimizers it is important to first try to get a
decent, complete, and correct solution for each puzzle; use the
remaining time to try to improve scores. Each puzzle had its own
scoring system, but these were added and then the top score assigned
100 points, the next score 99, and so forth. I scored 321 out of a
possible 368 points which was the 11th highest score for 90 points. I
think Wei-Hwa got the 100 here and the others also did well, but there
were horror stories from some who submitted incorrect solutions on
puzzle 1, where there was no possible partial credit for a solution
which broke the rule, thus falling far down the overall ranks.
After lunch, round 10 was another sprint, this time 13 mastermind
puzzles to solve in 39 minutes. 6 numerical ones (5-digit) on the first
page; 7 more with letters on the second page. The letter masterminds
had various sorts of themes; one used 6-letter city names. The last one
mixed letters and numbers and came out to a solution of 13WPC. Despite
the hint-ish nature of the second page, few finished all the puzzles. I
finished 8 of them for a score of 48.
Part 11, the last individual round, was an "innovative" one -- new
puzzle types or new variations of known puzzle types. There were 13 in
91 minutes, but this time worth 300 instead of the typical 200. I
solved 7 puzzles for 170 points, which I think was the 8th highest
score in this round, and this good round at the end boosted me a bit in
the rankings. One interesting one was Atomic Alert. In a hexagon-shaped
grid of triangles, we had to place 10 atomic alert symbols (each made
of 3 triangles, in either of two possible orientations), so that the
number of filled triangles in certain rows added up to the numbers
shown outside the grid, and none of the symbols touched, even at a
point.
Another puzzle from this round was Connected Magic Squares, which was
like a Number Place except that we had 4 copies of the grid which were
to be filled in the same way, but the sub-regions were drawn
differently in each grid. A disappointment was the last puzzle in this
round, a rolling cube maze where we had to roll the cube so that letter
on top of the cube always matched the letter on the page it was sitting
on, AND we had to spell out the name of sponsor FENIKS twice between
the (S)tart and (F)inish. We discussed this one on the bus during the
excursion and decided it was trivial if they gave us the pattern -- it
simply meant finding the unique sequence to roll the cube so that those
letters came up on top in that order and applying this to the grid. And
they did give us the pattern, with a lame grid that simply had the
letters FENIKSFENIKS across the diagonals so that any pattern of
rolling the die the minimal number of times would hit all correct
letters. They might as well have not
bothered to have letters in the grid at all.
Round 12 was a single puzzle to be solved by the team in 26 minutes. We
were given a deck of cards and a page with the puzzle. We were to
arrange the cards in a 7x7 grid so that a path could be followed A,
2, 3, ..., K in one suit, followed by some other suit and so on. Three
cards were to be dropped from the sequence, and the sums of the values
(with A=1, 2-10 = 2-10, and face cards = 1) and suits (clubs 4, hearts
3, diamonds 2, spades 1) were given to us in each row and column. Only
the
three unused cards were checked, and each correct one was worth 26
points. There were bonus points for the earliest correct solutions, but
nobody completely solved it, and only a couple lucky teams had managed
to get one right card and make the correct guess as to how to split up
the remaining rank and suit points.
After dinner, we managed to get 8 people together for 2 tables of
duplicate bridge. Results from various rounds came out as we played,
and by the end of the 8 hands we had all the scores and knew we'd won
the team title. I joined in a quick game of 10-player 6 Nimmt which I
came from behind to win with a clean last round. Roger and Wei-Hwa went
to bed to get a good sleep
before the finals while Jonathan and I stayed up later playing Tichu
(with Ulrich, which led to Nick dropping by and jokingly commenting
that we had a good strategy to keep Ulrich up late -- though it hardly
mattered since he had gotten the first round bye which meant he had at
least an extra hour before he had to compete). At another point
in the evening, somebody asked Ulrich if he thought it was unfair that
he wasn't going to get to do all the puzzles the next day. He responded
by pointing out that there were already many rounds when he couldn't do
all the puzzles, so this wasn't any different.
Saturday and Sunday
Saturday morning and early afternoon was the finals, which Chris
Dickson has described in far more detail than I ever could. But the
shortest Japanese team member made the playoffs, and his easel was
lowered upon request.
Then lunch,
and in the afternoon, Tichu and other card games, and many people's
final trips into town to buy whatever souvenirs they hadn't bought yet.
In the evening we went about a mile down the main road in Opatija to
the Hotel Kvarner where the farewell dinner party and awards ceremony
was held. Every team got to be announced, and go up and get handed at
least a packet of photos from the photo session, get a round of
applause, and yet another opportunity for people to take pictures of
the team. There were also some random prizes for the youngest solver (a
seashell about the size of his head) and best-scoring female (I didn't
see what this was).
The top three got team trophies and other prizes. It was reported that
the number 2 and 3 teams got puzzle books, though I didn't see them. We
got Croatian-English dictionaries, while we would have preferred puzzle
books, but at least they will help us decipher any puzzle instructions
in Feniks that we are unsure
of. And when you're a world champion, they
take a lot of pictures of you. I had Nancy Schuster take a couple shots
with my camera of us accepting our prizes. The top 3 individuals got
their prizes later, each a trophy and a shopping bag. The bags turned
out to have identical bizarre collections of items from the market,
including a ladle, a jar of salsa, and a package of some kind of food
seasoning.
There was lots of music during the party, before and between the
awards, and it continued on into the night, but after the last of the
awards, many of the teams started filtering out back to the hotel. I
went back around 11:30 and joined in a game of Great Dalmuti that
lasted until about 4 AM, when it was time for many of us to check out
of the hotel and catch our rides to Trieste or Zagreb for morning
flights home.