October 15th, 2004
|07:50 pm - Magnetic Battletents Snakescraper|
The previous three posts are placeholders no longer; have pasted in the longer entries I had written at the time, so go back and look at them if you're interested. Will write up today's report at my leisure and post it in once written. In short, all puzzles over except for the top 13 who go into the play-offs tomorrow. Last we saw had Britain in 11th place, 22 points behind 10th. Happy, tired, puzzled out, in love with Meg. :-)
(Go back and read my last three entries again! Loads of new material!)
The title comes from a persistent joke that the way this year's puzzles would be more difficult than last year's was through combining the myriad complications that are sometimes used to increase the difficulty. If you want another challenge, try unscrambling the original puzzle titles we have combined to make the mother of all puzzle names. (For another challenge, work out what this puzzle would actually consist of, based on the name...)
Rather happier dreams brought about breakfast (after twenty minutes later sleep than yesterday) and the second day's puzzles. Breakfast brought the completion of the marking from day one: individual scores resulting in positions for our four solvers which did not really represent what they thought they were capable of, plus a fall for the British team back to a frankly disappointing fifteenth place. After the day's first round we received the papers back for checking; we discovered a couple of copying errors where they had accepted our answers as correct but not actually included the scores for them in the grand totals that gave the British team another 70 points.
We started with part nine today rather than part seven, on the grounds that we had had a team round to end yesterday's play so starting with the morning's team round (as opposed to finishing with it) would minimise the amount of chair rearrangement required. Thirteen puzzles among four solvers in 26 minutes, the first twelve generating numerical answers used in the thirteenth. This featured some of my best solving of the contest; I was responsible for two complete solutions, one for a puzzle involving numbering a complicated arrangement of thirteen weights in order to make all five pivots balance exactly, the other for a puzzle involving placing digits 1 to 9 into a square so to make all the inequalities accurate. (In practice, the trick here was to find the winding path which had to be filled in so that 9 > 8 > 7 = 7 > 6 > 5 > 4 > 3 > 2 = 2 > 1. Similar observation and logic made the rest follow.)
This was a fantastic round for both my teams, the one for whom I was solving and the one of which I was captain. Both teams completed seven puzzles within the short time limit, performances competitive with almost all our opposition. (Heck, it was a good twenty points better than the USA team.) This brought Britain up to eleventh and within 22 points of the top ten - a cheering performance to inspire confidence.
Round seven followed round nine, naturally enough: 91 minutes of puzzles about pentominoes (think Tetris blocks with five pieces instead of four) and arrows. Ingeniously designed puzzles that were a great deal of fun to solve, though I only managed to get two of them to work. A significant minority of WPC puzzles have a thematic sting in their tails and the answers can be guessed without working through in full; a puzzle I started at the end but ran out of time on was a large, time-consuming wordsearch (using the Croatian alphabet, including the bigram NJ as a single character) where the letters not used in the grid when you drew arrows through the many town names hidden formed another town name. That final town name was, inevitably enough, OPATIJA, where we were staying. Did I think of guessing that without skipping the puzzle-solving? No.
Finally for the morning, round eight, with three "optimiser" puzzles. These have no absolutely correct answer, just challenge you to produce the best answer that you can. One involved drawing a line through a square to pick up numbers along the way, with heavy restrictions on movement and diagonal motion costing you points; a second involved putting letters into a 5x5 box so that specified names of Greek gods (in their usual Croatian transliteration) could be spelled out Boggle-style; the third involved the placement of robots into a grid to meet rules about where heads could be placed and whether things could touch each other or not. I had one good solution, one poor solution and one illegal solution where I misunderstood the question completely. These are very hard to mark; as I type, all the rounds except this one have been marked and we're waiting for this last set of results.
Lunch next, which was very similar to all our other meals (cheese and ham, soup, salad and a few chips), but unusual in that it was shared with four-time champion Wei-Hwa Huang. Much of our conversation revolved around puzzle hunts and weblogs, oddly enough. (pchou, he remembered you as captain of the Puzzle Fighters team!) Wei-Hwa had also been playing the grand piano in the lobby the night before for relaxation. Wei-Hwa's piano is slightly less accomplished than his puzzling; far better than mine ever was and by no means bad, but not his greatest asset. We also enjoyed looking at some giant - literally poster-sized! - paper puzzles produced and sold in Japan, where even the pattern of information given on the blank grid was quite attractive. It was estimated that a beginner would take 277 hours to solve the puzzle and an expert 11 hours. Both estimates look plausible to me, considering the sheer quantity of puzzle.
Round ten in the afternoon had 39 minutes for thirteen Mastermind puzzles - you know, the old peg game where you try to deduce a hidden code and your guesses are marked with black and white pegs according to accuracy. There were some numerical puzzles and other themed verbal ones: a very neat one with the names of participants, one with forenames, one with town names, one with the names of past and present Manchester United players and a thirteenth puzzle which worked out as 13WPC (for "thirteenth World Puzzle Championship"). I guess-recognised the final one and calculated out the first numerical one, but my score of two out of thirteen dropped me further behind.
The eleventh round was the last of the marathons: 91 minutes, 13 puzzles, all original formats, lots of highly imaginitive new puzzle ideas in there. I particularly enjoyed "Atomic Alert", where you had to discern the positions of three-equilateral-triangles-touching-at-a-point hazard symbols secreted within the honeycomb grid, and a battleships variant where lightbuoys dotted around the ocean shone light around like a chess queen to illuminate some or all of the ships. There was also an utterly fantastic Latin Squares variant: imagine, if you will, a 9x9 grid filled with the letters of the word DUBROVNIK so that each letter appears only once on each row and each column. Furthermore, split the grid up into nine blocks, each of nine cells, such that each block contains one of each letter. So far, so standard. (You can get lots of examples of this style of problem on the Web, within very easy Googling.)
However, the clever bit was that the same grid was partitioned into nine blocks four different ways, obeying the rules in each case, so only by working out what information was to be gained from each grid, and working out what the different shapes of the blocks of cells led you to conclude about their contents, could you place the letters inside the grids. Large amounts of comparison from grid to grid and holding considerable amounts of information in your head required: way too much for me to be able to do, but I was able to appreciate the very impressive intricacy and ingenuity in the design. Possibly the single best puzzle of the contest.
Finally, round twelve required teams to place 49 of 52 playing cards into a 7x7 grid, in a sequence we had to work out, so that the totals of the ranks in each row and column were as specified, and the totals of the values assigned to the suits were also as specified. This was tremendously difficult: only two teams managed to identify all three missing cards correctly and even guessing one was impressive. (Working out what the three values might have been and what combinations of suits were possible wasn't too tricky, but working out the logic in order to cross-reference the two was terribly hard.)
Finished! We decompressed somewhat after the final round of puzzles, though the guys are so enthusiastic about puzzles that (hearteningly) their favourite method of relaxation involved answering the puzzles that they didn't try within the time limit. This was also the first time that we had a continuous 90-minute block in unpressured circumstances and so could buy a wireless access voucher, so this is the point at which I uploaded my longer entries for the previous three days. Happily, the 90 minutes can be used in a number of sessions, so I'm writing this offline and will only go online briefly once it's written up.
Dinner, then card games to relax. There being five of us, and not having any particular five-player games in mind (mmm - if I'd been thinking about it, I'd have borrowed the laptop and gone to Pagat) we struggled to decide what to play for a while, until I improvised a version of inductive-logic game Zendo played with playing cards. In short, one master devises a property that collections of playing cards may identifiably either have or not have, then the other players build examples and counter-examples in an attempt to determine what the rule is. There's a little more to the mechanics than that, but that's the principle. The Dutch team dropped off a handful of packs of playing cards, but these were all in Dutch, with Kings, Queens and Knaves signified by H, V and B respectively. (The mnemonic we used that stuck rather quickly: King Henry, Queen Victoria, Prince Baldrick.)
Some more results have been published, but we're just waiting for the results of the optimisers - and, theoretically, the results of any further appeals and protests - before we'll know the final positions of the team competition and the top 13 qualifiers for the play-offs. It does look like the British team will finish in 13th place, but perhaps we'll catch up 20 points on Poland to take twelfth from them. :-) Full results as soon as they have been officially confirmed. (We have tried to anti-protest ourselves another 30 marks fewer by pointing out that we had earned marks on one question which was marked as incorrect, but after review, they would not take the dubiously-awarded marks away from us.)
Since then, I've come upstairs and got this all out of my brain for, ooh, almost an hour and a half. There seems to be quite a party going on next door, with all the conversation in a language we don't speak (not even the language of puzzles, sadly!) and our attempts to displace them by playing obnoxious MP3 files at high volume have not yet been successful. Ah well; downstairs now to look for the results from the optimisers and post this. Lie-in tomorrow! We'll be skipping breakfast in the morning and have no need to attend the 9am meeting about the rules for the quarter-finals, but we'll be sure to get up and get ready to watch the quarters, semis and finals. So long as our Internet access still works, we'll try to get the official results to you as soon as possible. Watch this space! Better still, also watch ericklendl!
(Edited briefly before posting: WE DID IT! WE OVERTOOK POLAND! Subject to confirmation, we finish twelfth - not the tenth I had hoped for, but both an absolute record in terms of closeness to the top and in terms of proportional position because we've never been the very top of the bottom half before. Top three teams, in order: USA, Germany, Hungary. I finished 83rd out of 90. Hurrah!)
Current Mood: happy, tired, puzzled out