Lots of words and a little cross

As folks round these parts know, I spent last weekend in Brooklyn, at the 2008 American Crossword Puzzle Tournament, the 31st of its kind, and the first within an easy shot by train from where I live. So a solving did I go.

And to be honest, I thought I'd do better than I did, so it was a little humbling for me, but it also just about guaranteed that I'll be heading back down there next year in an effort to improve upon my showing. First, let's talk results: after 3 rounds, I was ranked #365, and I climbed a couple of spots to #363 after the second set of 3 on Saturday. Sunday's puzzle dropped me a few spots to #370, though, which is where I ended the tournament. I was a little perplexed, because I thought I'd done better on 7 than I apparently did. But oh well. I've thought a lot about my performance over the last day or so, but before I talk about where I went wrong, charts!

Here's a table featuring my puzzle by puzzle breakdown alongside that of the folks who placed 1st, 100th, 200th, and 300th:

Perhaps I say this just to comfort myself, but really, all that separated me from #300 was my performance on 2 of the easier puzzles. #1 was extremely easy, and #7 was a fairly typical Sunday-style NYT puzzle. Why did I do so poorly on them in particular? I'll get to that, but here's one more chart, featuring my scores lined up again a specific version of ideal scores. The way that scoring works is that you receive 10 pts for each correct word, 25 pts for each full minute you finish under the limit, and 150 pts for a perfect grid. Errors or blanks cost you 25 pts, plus of course you lose the 20 pts from the words crossing the rogue square.

Put another way, every 2 errors cost you a little more than 3 minutes of time bonus, and a perfect puzzle is worth 6 minutes of bonus. So here are my scores lined up with the score for each puzzle done perfectly, but with no time bonuses. In other words, had I gotten each puzzle perfect, and taken the full time to get them that way, my scores would have been those of the second row below:

By this measure, I didn't do too badly. A score of 8290 would only have been good enough to place 357th, and one person got that exact score. But I don't think I got any of them perfectly. I did keep track of my times, but it's hard to calculate from there how many mistakes I made without seeing the grids. The first error costs 25 + 20 points for the 2 words that are wrong, but the second error can cost 45 or 35, depending on whether or not it's in one of the words I'd lost with the first error. Ditto for the third error, but the fourth, if it overlapped with two other mistakes, could cost 25, 35, or 45. On top of that, you can't lose more bonus in error than you gain in time. So if I finish 10 minutes early, and earn 250 pts that way, but make 12 mistakes, only 10 of them cost me the 25 points--the other two just affect my word count. So I can't really reconstruct my scores with any accuracy.

What I can say is that I thought I did better on puzzles 1 and 7 than I actually did, I did pretty well on 3 (a 30 minute puzzle I finished in 8), and 2 & 5 (the 2 hardest by all accounts) kicked my tail. And I think that second chart bears this out. By my calculations (I did keep track of my solving times), If I'd gotten #1 perfect, I would have had 1180, 1240 for #4, and 2000 or so on #7. Another 800-odd points or so moves me up into the mid-200s. But I'm relying here on perfect grids, and I knew that a couple weren't, because I neither knew the name of an old San Francisco mayor nor that olid was a word for a disagreeable smell (that was puzzle 7). Take out those 150 point bonuses (and drop another 45 points for at least 1 mistake), and my scores on those puzzles become 985 (guess I only missed one on #1), 1045, and 1805, which are not too far off from how I actually scored.

No, the really big difference between me and the top honchos was the fact that I really struggled with 2 and 5. I finished 2 with only 30 seconds to go, and 5 I didn't finish at all. And here's the thing: they were the most clever puzzles. But I was so focused on speed and moving clue to clue that I didn't really sit back, take a breath, and figure out the clever ahead of time. #2 was a word ladder, with 9 clues corresponding to rungs on the ladder (first word VENUE, second word VENUS, third word MENUS, etc.). I figured it out fairly easily after the round, but during the round, I didn't think I had time. And #5 was a little more diabolical, with long answers starting with notes from the do-re-mi scale, but transposed. For example, the second long answer was MINTSTRIKE, a relatively meaningless phrase (Labor problems in Denver and Philadelphia?) until you realize that the phrase is actually "RENTSTRIKE" with RE transposed one step up the scale to MI. And so on through the puzzle. But in neither case did I really bother thinking about the theme, figuring that it would just come to me as I frantically scrabbled for individual words. When I came up for breath after each round, the trick was obvious. While the clock was running? Not so much. Instead, I obsessed over the crossing words, and got 4 or 5 of the long words that way, never understanding (a) how they fit together, nor (b) using that information to help me solve them. Inexplicably, I didn't even read the title of the puzzle ("Up-Scale" might have been--how you say?--a CLUE?!?!).

Ah well. I think that the key to my improvement, both in nearing perfection on the easier ones and completion on the harder ones, will involve not taking the clock quite as seriously as I seemed to on Saturday. I need to be a little more careful, yes, but mainly, I think that I need to realize that a minute or two sacrificed to the big picture would have netted me hundreds more points on the harder puzzles.

I did get to wish Will Shortz good morning, and everyone at the tournament was extremely friendly. It reminded me of a mix of other nerdy pastimes. It was something like the debate tournaments of my youth, but without the head-to-head competition (except at the top of the ladder, I guess). The sessions themselves were not unlike the SATs or GREs, and there was a little bit of costumery and localized celebrity, which put me in mind of a fairly mild con, I guess. There was a broad range of ages, and certainly more boys than girls, but all in all, it wasn't scary or anything.

I could have done better, but I guess I feel like #370 (out of about 700) is not a bad place to start. I'm planning on returning next year and improving my score, and now that I know what it's like, I think I'll be able to do so. As I told people around the office today, if there was another tournament this weekend, I'd go to it. I'm a little compulsive that way.

And I'm only a little ashamed to see that it took me this long to get to the fact that Friday night, I had dinner with Douglas and hung out with him for the first time in a few years. So yeah, good times.

Gotta run. I've only got 360 days to practice before next year's tournament, so better get to it!

Yes, that was a joke. Sort of. That's all.