Log in

19 March 2006 @ 04:22 pm
WSC Report Day 2 - Morning  
After about 5 hours of sleep (the most I'd get while in Lucca), we woke up, showered, and ate breakfast. As we had done at the WPC in Eger, Hungary last fall, Wei-Hwa and I killed the time before the competition started by playing the game "Race for the Galaxy", an as-yet-unpublished card game that has an incredible mixture of different potential strategies and a very good play balance, with the basic goal of the game being to settle the galaxy, develop technologies, produce goods to sell or trade, and/or develop a military to take over worlds. As I do not play many card games, I was very much still in the second phase of the learning curve for the game and still making some errors during play in implementing my strategies. Combine this with the fact Wei-Hwa has been playing the game during most of his free time recently, our matches tended to not be too close. Still, "Race" is quite an incredible game. Wei-Hwa and I would continually joke, as we'd often start a game at the same time a press reporter found us and wanted to ask questions, that we should use this opportunity to get an incredible amount of press for the game which will hopefully be released by this Thanksgiving. What do the world's best Sudoku solvers do for fun? Play "Race for the Galaxy!"

The competition started as promptly at 10 AM as I could expect in Italy, roughly around 10:05, with a speed round. In ten minutes, we would have 9 6x6 mini-sudoku puzzles to try to solve. Most of the time I find this mainly an exercise in how fast I can write numbers, which is rather fast, so I figured I would do well. As an aside, I find this the most intriguingly formatted puzzle in the SudokuMix magazine as all the puzzles have an H/M box to mark down your time. I worry that many of my solving times would be turned into 0 minutes if correct rounding is done and wonder why for the minis they don't go to M/S for the time.

After feverishly working through the puzzles, I was on the ninth of the set when I heard the first person declare he was finished. I kept at it and declared I was done in the fifth bonus spot. Ron Osher on our team declared third overall and stood to get a larger bonus than me, with Wei-Hwa finishing about 18th overall. Of course, we would not know if the people ahead of us had made errors, or even if we were correct, but for now I had a small 25 point lead based on the bonus over the competitor I feared most, my roommate Wei-Hwa.

Round 2 was the longest and would prove to be the most decisive of the rounds in the competition. It had 9 puzzles in 1 hour of time, including 2 classic sudoku, a diagonal sudoku, an odd/even and a 147 sudoku (both very simplified sudoku puzzles), a sum sudoku (probably my strongest puzzle in terms of time it takes me to solve versus the points it is worth), an extra-regions sudoku with two special shaded regions that also obeyed the 1-9 constraint, a pips sudoku (6x6) that used partially filled in pips as on the surface of a die to clue the numbers, and a consecutive sudoku where all consecutive numbers that touched (even across square boundaries) would be indicated. I worked through this round efficiently, beginning at the start, and going through the sum sudoku, then working from the back, and found myself with about 15 minutes left with just the extra-regions puzzle to go. This one started out for me sort of tough, but I eventually got it to work out and, with about 5 minutes to go, had time to check all my answers twice. I was really pumped - a perfect score on the longest round of the competition. Wei-Hwa would tell me during the break he finished with about a half-minute left but not enough time to check over. So, I still believed I was leading Wei-Hwa by 25, and now Ron had fallen somewhat behind both of us.

Round 3 was the one I figured would be the most interesting for new puzzle types going into the competition. It featured the "Mechanical Sudoku" which, when I first saw it in the instruction booklet, I thought would prove the most unique. The puzzle involved being given 9 3x3 squares that had to be assembled to make an intact, solvable grid. The numbers were written in a font such that rotations were clearly going to be possible (6 and 9, 2 and 5 were identical by 180 degree rotation). I predicted it would be given out on transparencies so that in addition to rotations, the whole piece could be flipped over to introduce more complexity. This is the format it actually did appear in. To help out the solvers, the 4 corner pieces had the parts that would be in the corner marked with a dot. You could take all the possible orientations of the corners, and then gradually begin to place them on the grid. Then, working through the middle pieces you could assemble the whole puzzle. Finally, you had to actually solve the sudoku. It was one of my two favorite puzzles in the whole event, and I'd like to thank Cihan Altay for writing it (he also happened to write my other favorite puzzle, which leads me to recall how great PQRST is and how much I miss it). Round 3 also included the other new type I found interesting, "star sudoku," where 6 big triangles with 9 smaller triangle pieces were arranged so that all rows and diagonals had to obey the standard 1-9 constraint. I messed up on solving this one the first two times (sloppy!!!) and so, while I finished this puzzle, I did not finish the last variant, a word sudoku with the numbers given in a neutral language, "Samoan". It wasn't a hard puzzle; I just didn't have time to finish it after messing up the Star Sudoku. Wei-Hwa completed all the puzzles in the round, so our internal score-keeping now had me down just 5 points to him, entering the last round of the morning competition.

The final morning round was another speed round, with 3 classic sudoku given to us with 20 minutes to complete them in. They were rated 10/20/40 points based on difficulty so I started on the 40 pointer, got rather far but slightly stuck, made a rookie decision that worked this time for me to go to the twenty pointer and finish it (leaving 8-9 minutes of invested time behind without a finished grid), then came back and finished the 40 pointer. I had about 3 minutes left then to try the first puzzle but could not get it all done in time. Tetsuya Nishio from Japan was the only one to declare the bonus, but it would eventually turn out that he did not have all the puzzles solved correctly. While all the rounds until this one seemed timed perfectly, I felt this round at 20 minutes was somewhat short. Bonuses are meant to be collected, and so I would set this round to at least 25-30 minutes (or make the time variable until about 3 people have declared finished). The organizers felt twenty minutes would be enough, and it almost was. I probably would have taken 22 minutes overall. I figured I had 60 of the 70 points (no bonus for the round). Wei-Hwa did the first and last puzzle for 50 points. So suddenly I went from behind 5 to ahead 5. However, I knew we'd be duking it out throughout the afternoon for the best US score. We did not know how others were doing but felt we were doing well enough to be near the top. By the afternoon we would find out how far ahead we already were (which, needless to say, instantly got us the attention of almost all the photographers and reporters in the room for the remainder of the competition), but for two hours we had time for a quick bite of lunch, and of course many rounds of "Race for the Galaxy".