American Crossword Puzzle Tournament 2026
Apr. 16th, 2026 10:05 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
fauxkloreI have other things to write about, but I should attempt catching up on actual stuff I did and can let the rest wait a bit. So here is my write-up of the 2026 American Crossword Puzzle Tournament. Note that I will include some spoilers about puzzles, but I will put those in comments encoded with rot13. In order to decode them, go to rot13.com and copy the encoded text into the top box.
This was my 10th ACPT (non-consecutively) and, significantly, was the last time it was being held at the Stamford Marriott in Connecticut. I’ve missed several over the years, largely because my life is a schedule conflict. For registration, they opened it early to people who had competed before and, when they did open up registration for rookies, it sold out in 1 minute and 37 seconds! (There were more rookies who got in off the wait list, but many people were disappointed,) Next year (and, presumably, subsequent years) will be in Philadelphia, which is a city I like a lot, at the Sheraton (which is a hotel I’m not crazy about, but I’ll live.)
Anyway, I took the train up on Friday afternoon, which was fine. I had a quick salad for supper and settled into my room. You can tell that the Stamford Marriott has enjoyed hosting the ACPT, as they gave each of us this thermal mug as a gift.
I picked up my contestant packet and settled into a seat in the ballroom for the evening activities. The first game was “Mental Geog-ing” which was advertised as “a wordplay quiz on place names, solved in pairs.” But, actually, it was just two teams, i.e. the room divided in half. I had a quibble with the answer to one question, which called for naming three European countries which have a “z” in their names. Will Shortz didn’t allow Azerbaijan, saying that it’s in Asia. Well, most of it is, but part of northern Azerbaijan is in Europe.
After that was the usual competition where you chose from four puzzles. (I you have time, you can do more than one.) I’m not a fan of Split Decisions, so ruled that out. A quick glance at the Cryptic (which is usually my favorite) suggested it was a hard one, so I saved it for later. The Spiral was going fine, but I got bogged down on a couple of answers. I should have started with the Quote Boxes, since I was able to solve two of the three of them quickly, but ran out of time before doing the third. Oh, well.
I should note that a lot of my reason for going to the ACPT is socializing. I looked for several friends and found some, but with over 800 people there, it was impossible to find others. I gave up quickly on the wine and cheese reception, as it was just too crowded and noisy. I did catch up with a few others later in the weekend, but never found some other people who I know were there.
The actual crossword competition started on Saturday morning. I opted to go to one of the overflow rooms downstairs, which was a good decision as it was much less crowded than the main room upstairs. Puzzle #1 was by Ross Trudeau. This was straightforward and one could have solved it without getting the theme (though I did find the theme gettable and amusing enough). So far, so good.
Will Shortz said that some of the test solvers had found Puzzle #2 harder than the dreaded Puzzle #5, but I sussed out the theme fairly readily. Which surprised me since the puzzle was by Brendan Emmett Quigley, who I think of as my puzzle nemesis. I wasn’t particularly quick on solving it. I did take a minute to proofread, which was good, because I caught an error I’d made on one answer. Unfortunately, I didn’t keep proofreading, as I had a really dumb error in another place, essentially the handwritten equivalent of a typo. When the scores got posted, I kicked myself mentally for blowing what should have been a clean solve. (See spoiler below.)
Puzzle #3 was by Mike Shenk. I almost always enjoy his puzzles and I enjoyed this one, too. However, I have a bone to pick with him regarding one answer, which resulted in two wrong squares. (See spoiler below.) To be fair, I should have caught this from the crossings, but it still struck me as problematic.
I’d eaten a large breakfast so I just had yogurt up in my room for lunch. At that point, the scores had not come in yet, so I still felt pretty confident. (It turned out that there was a problem with the scoring system and I hadn’t discovered a workaround yet.) I went into the afternoon session believing (incorrectly) that I had solved cleanly.
Anyway, Puzzle #4, which was by Rena Cohen, was straightforward. I think this is another one where you didn’t really have to grasp the theme to solve this. I will, however, note that the revealer was not actually correct. (See spoiler below)
Then came Puzzle #5, something I have solved cleanly exactly once ever. Robyn Weintraub made a puzzle that lived up to its evil reputation, which is surprising since I used to think of her as a good person. I did, eventually, grasp part of the trick, but didn’t quite get it completely. And I certainly didn’t do so with enough time to finish the puzzle. My score on it was downright pathetic. (See spoiler below)
The afternoon ended with Puzzle #6, by Lynn Lempel. This was reasonably straightforward, with an enjoyable theme.
By this time, I’d figured out how to see not only the scores but, also, the scans of my puzzles. What worked for me was switching from looking using Safari to using Chrome. Other people had some success with clearing their browser cache or using an incognito window. Let’s just say that I was not in the top half.
Saturday night started with a celebration of the movie Wordplay, which is what made a lot of people aware of the ACPT. Not me, by the way. I had first learned of it from an A-hed in the Wall Street Journal. (That’s the human interest story that they put on the front page. The WSJ A-hed often has some of the best writing in American journalism.) I know that I did see the movie when it first came out, but I can’t remember where. There were various outtakes from the movie and director Patrick Creadon talked about its surprising success. That was followed by the presentation of the Merl Reagle MEmoRiaL Award to Lynn Lempel. I was particularly amused by her mentioning that she used to play jotto (a game that is, essentially, the forerunner to world). My 50th high school reunion is approaching and my friends and I used to play jotto (and some other games) in the library during our free periods. Yes, I was a word nerd back then, too.
Sunday morning finished off the tournament with Puzzle #7, by Sam Ezersky. I found this one pretty easy going and it did improve my standing somewhat. In the end, I finished 459th out of 803. That’s the 43rd percentile, which is the worst finish I’ve had. To compare, here’s how I’ve done each time I’ve competed:
2009 – 265 / 654 (55th percentile)
2012 – 241 / 594 (59th percentile)
2014 – 202 / 580 (65th percentile)
2016 – 171 / 576 (70th percentile)
2017 – 141 / 619 (77th percentile)
2018 – 254 / 674 (62nd percentile)
2019 – 220 / 741 (70th percentile)
2022 - 243 / 474 (49th percentile)
2023 - 301 / 774 (61st percentile)
2026 - 459 / 803 (43rd percentile)
Oh, well.
Anyway, there was a little more entertainment to be had, in the form of a live taping of a “Hello Puzzlers” podcast with A. J. Jacobs and Greg Pliska. They had several audience members come up to play a game in which they had to guess the most common clues for certain answers. That was followed by them grilling Sam Ezersky about his word list for the Spelling Bee, which often gets criticized for leaving out words many people think should belong. There was a cute story about someone sending a large box of raffia to Will Shortz to prove to the New York Times that “raffia” is a legitimate word. Here’s a photo of A.J., Sam, and Greg (from left to right).
After Puzzle 8 (the Championship puzzle) was distributed, I went up to my room to finish packing. (I had late check-out). Erik Agard was the big winner. I said goodbye to a few people and slowly made my way over to the train station. I stopped to take a farewell to Stamford photo, since I don’t expect to have much reason to go back there again now.
Next year in Philadelphia!
Spoiler for Puzzle #2: Bar pyhr jnf "Wrnaf phg" naq V vavgvnyyl jebgr "fyvz phg." Juvpu vf qhzo, orpnhfr gur jbeq "phg" jbhyqa'g unir orra ercrngrq. Jura V cebbsernq, V ernyvmrq vg unq gb or "fyvz svg." V pnhtug gung naq yrnearq gung Trbetvn B'Xrrssr unq gjb "S"f va ure ynfg anzr. V unir ab vqrn ubj V jebgr n "T" vafgrnq bs n "Q" va nabgure nafjre, ubjrire.
Spoiler for Puzzle #3: Gur pyhr sbe 47Q jnf "Fynatl 'haqrefgnaq?'" V jebgr va "pncvfpr" vzzrqvngryl. Juvpu vf gur pbeerpg fcryyvat bs guvf Vgnyvna jbeq. Zvxr Furax'f fcryyvat jnf "pncvpur." Tbbtyr fhttrfgf guvf na na Nzrevpnavmrq nygreangvir, ohg V'ir arire frra gung orsber. V unir frra gur rira zber nobyzvanoyr "pncrrfu" snveyl serdhragyl, fb gung zvtug unir yrq zr gb erivfvg gur pebffvatf gurer.
Ol gur jnl, "pncvfpr" jnf bar bs zl sngure'f snibevgr jbeqf naq V jnf nobhg 40 lrnef byq orsber V qvfpbirerq gung vg jnf, va snpg, Vgnyvna naq abg Lvqqvfu!
Spoiler for Puzzle #4: Gur gurzr nafjre sbe guvf chmmyr jnf "fcner gver." Gur cynvagrkg orybj gur gvgyr ernq "Rirel pne unf bar." Fcner gverf fgnegrq orvat cunfrq bhg va gur zvq 2000'f naq, nf bs 2025, ebhtuyl unys bs arj pnef ner fbyq jvgubhg gurz. (NNN pynvzf 60% naq Pbafhzre Ercbegf pynvzf 45% ynpx gurz.)
Spoiler for Puzzle #5: Lbh unq gb svther bhg gung lbh unq gb zbir n jbeq sbe na navzny sebz bar pyhr gb gur pyhr orsber vg. Sbe rknzcyr, 69N jnf "Nfgebabzvpnyyl, vg'f n yvggyr" naq 71N jnf "Oebja orne va gur pbzvpf." Zbivat gur jbeq "orne" tvirf lbh HEFN ZVABE (na nfgebabzvpny yvggyr orne) naq PUNEYVR (Oebja va gur pbzvpf), erfcrpgviryl. Vg'f irel pyrire, bs pbhefr, ohg punyyratvat gb svther bhg.
This was my 10th ACPT (non-consecutively) and, significantly, was the last time it was being held at the Stamford Marriott in Connecticut. I’ve missed several over the years, largely because my life is a schedule conflict. For registration, they opened it early to people who had competed before and, when they did open up registration for rookies, it sold out in 1 minute and 37 seconds! (There were more rookies who got in off the wait list, but many people were disappointed,) Next year (and, presumably, subsequent years) will be in Philadelphia, which is a city I like a lot, at the Sheraton (which is a hotel I’m not crazy about, but I’ll live.)
Anyway, I took the train up on Friday afternoon, which was fine. I had a quick salad for supper and settled into my room. You can tell that the Stamford Marriott has enjoyed hosting the ACPT, as they gave each of us this thermal mug as a gift.
I picked up my contestant packet and settled into a seat in the ballroom for the evening activities. The first game was “Mental Geog-ing” which was advertised as “a wordplay quiz on place names, solved in pairs.” But, actually, it was just two teams, i.e. the room divided in half. I had a quibble with the answer to one question, which called for naming three European countries which have a “z” in their names. Will Shortz didn’t allow Azerbaijan, saying that it’s in Asia. Well, most of it is, but part of northern Azerbaijan is in Europe.
After that was the usual competition where you chose from four puzzles. (I you have time, you can do more than one.) I’m not a fan of Split Decisions, so ruled that out. A quick glance at the Cryptic (which is usually my favorite) suggested it was a hard one, so I saved it for later. The Spiral was going fine, but I got bogged down on a couple of answers. I should have started with the Quote Boxes, since I was able to solve two of the three of them quickly, but ran out of time before doing the third. Oh, well.
I should note that a lot of my reason for going to the ACPT is socializing. I looked for several friends and found some, but with over 800 people there, it was impossible to find others. I gave up quickly on the wine and cheese reception, as it was just too crowded and noisy. I did catch up with a few others later in the weekend, but never found some other people who I know were there.
The actual crossword competition started on Saturday morning. I opted to go to one of the overflow rooms downstairs, which was a good decision as it was much less crowded than the main room upstairs. Puzzle #1 was by Ross Trudeau. This was straightforward and one could have solved it without getting the theme (though I did find the theme gettable and amusing enough). So far, so good.
Will Shortz said that some of the test solvers had found Puzzle #2 harder than the dreaded Puzzle #5, but I sussed out the theme fairly readily. Which surprised me since the puzzle was by Brendan Emmett Quigley, who I think of as my puzzle nemesis. I wasn’t particularly quick on solving it. I did take a minute to proofread, which was good, because I caught an error I’d made on one answer. Unfortunately, I didn’t keep proofreading, as I had a really dumb error in another place, essentially the handwritten equivalent of a typo. When the scores got posted, I kicked myself mentally for blowing what should have been a clean solve. (See spoiler below.)
Puzzle #3 was by Mike Shenk. I almost always enjoy his puzzles and I enjoyed this one, too. However, I have a bone to pick with him regarding one answer, which resulted in two wrong squares. (See spoiler below.) To be fair, I should have caught this from the crossings, but it still struck me as problematic.
I’d eaten a large breakfast so I just had yogurt up in my room for lunch. At that point, the scores had not come in yet, so I still felt pretty confident. (It turned out that there was a problem with the scoring system and I hadn’t discovered a workaround yet.) I went into the afternoon session believing (incorrectly) that I had solved cleanly.
Anyway, Puzzle #4, which was by Rena Cohen, was straightforward. I think this is another one where you didn’t really have to grasp the theme to solve this. I will, however, note that the revealer was not actually correct. (See spoiler below)
Then came Puzzle #5, something I have solved cleanly exactly once ever. Robyn Weintraub made a puzzle that lived up to its evil reputation, which is surprising since I used to think of her as a good person. I did, eventually, grasp part of the trick, but didn’t quite get it completely. And I certainly didn’t do so with enough time to finish the puzzle. My score on it was downright pathetic. (See spoiler below)
The afternoon ended with Puzzle #6, by Lynn Lempel. This was reasonably straightforward, with an enjoyable theme.
By this time, I’d figured out how to see not only the scores but, also, the scans of my puzzles. What worked for me was switching from looking using Safari to using Chrome. Other people had some success with clearing their browser cache or using an incognito window. Let’s just say that I was not in the top half.
Saturday night started with a celebration of the movie Wordplay, which is what made a lot of people aware of the ACPT. Not me, by the way. I had first learned of it from an A-hed in the Wall Street Journal. (That’s the human interest story that they put on the front page. The WSJ A-hed often has some of the best writing in American journalism.) I know that I did see the movie when it first came out, but I can’t remember where. There were various outtakes from the movie and director Patrick Creadon talked about its surprising success. That was followed by the presentation of the Merl Reagle MEmoRiaL Award to Lynn Lempel. I was particularly amused by her mentioning that she used to play jotto (a game that is, essentially, the forerunner to world). My 50th high school reunion is approaching and my friends and I used to play jotto (and some other games) in the library during our free periods. Yes, I was a word nerd back then, too.
Sunday morning finished off the tournament with Puzzle #7, by Sam Ezersky. I found this one pretty easy going and it did improve my standing somewhat. In the end, I finished 459th out of 803. That’s the 43rd percentile, which is the worst finish I’ve had. To compare, here’s how I’ve done each time I’ve competed:
2009 – 265 / 654 (55th percentile)
2012 – 241 / 594 (59th percentile)
2014 – 202 / 580 (65th percentile)
2016 – 171 / 576 (70th percentile)
2017 – 141 / 619 (77th percentile)
2018 – 254 / 674 (62nd percentile)
2019 – 220 / 741 (70th percentile)
2022 - 243 / 474 (49th percentile)
2023 - 301 / 774 (61st percentile)
2026 - 459 / 803 (43rd percentile)
Oh, well.
Anyway, there was a little more entertainment to be had, in the form of a live taping of a “Hello Puzzlers” podcast with A. J. Jacobs and Greg Pliska. They had several audience members come up to play a game in which they had to guess the most common clues for certain answers. That was followed by them grilling Sam Ezersky about his word list for the Spelling Bee, which often gets criticized for leaving out words many people think should belong. There was a cute story about someone sending a large box of raffia to Will Shortz to prove to the New York Times that “raffia” is a legitimate word. Here’s a photo of A.J., Sam, and Greg (from left to right).
After Puzzle 8 (the Championship puzzle) was distributed, I went up to my room to finish packing. (I had late check-out). Erik Agard was the big winner. I said goodbye to a few people and slowly made my way over to the train station. I stopped to take a farewell to Stamford photo, since I don’t expect to have much reason to go back there again now.
Next year in Philadelphia!
Spoiler for Puzzle #2: Bar pyhr jnf "Wrnaf phg" naq V vavgvnyyl jebgr "fyvz phg." Juvpu vf qhzo, orpnhfr gur jbeq "phg" jbhyqa'g unir orra ercrngrq. Jura V cebbsernq, V ernyvmrq vg unq gb or "fyvz svg." V pnhtug gung naq yrnearq gung Trbetvn B'Xrrssr unq gjb "S"f va ure ynfg anzr. V unir ab vqrn ubj V jebgr n "T" vafgrnq bs n "Q" va nabgure nafjre, ubjrire.
Spoiler for Puzzle #3: Gur pyhr sbe 47Q jnf "Fynatl 'haqrefgnaq?'" V jebgr va "pncvfpr" vzzrqvngryl. Juvpu vf gur pbeerpg fcryyvat bs guvf Vgnyvna jbeq. Zvxr Furax'f fcryyvat jnf "pncvpur." Tbbtyr fhttrfgf guvf na na Nzrevpnavmrq nygreangvir, ohg V'ir arire frra gung orsber. V unir frra gur rira zber nobyzvanoyr "pncrrfu" snveyl serdhragyl, fb gung zvtug unir yrq zr gb erivfvg gur pebffvatf gurer.
Ol gur jnl, "pncvfpr" jnf bar bs zl sngure'f snibevgr jbeqf naq V jnf nobhg 40 lrnef byq orsber V qvfpbirerq gung vg jnf, va snpg, Vgnyvna naq abg Lvqqvfu!
Spoiler for Puzzle #4: Gur gurzr nafjre sbe guvf chmmyr jnf "fcner gver." Gur cynvagrkg orybj gur gvgyr ernq "Rirel pne unf bar." Fcner gverf fgnegrq orvat cunfrq bhg va gur zvq 2000'f naq, nf bs 2025, ebhtuyl unys bs arj pnef ner fbyq jvgubhg gurz. (NNN pynvzf 60% naq Pbafhzre Ercbegf pynvzf 45% ynpx gurz.)
Spoiler for Puzzle #5: Lbh unq gb svther bhg gung lbh unq gb zbir n jbeq sbe na navzny sebz bar pyhr gb gur pyhr orsber vg. Sbe rknzcyr, 69N jnf "Nfgebabzvpnyyl, vg'f n yvggyr" naq 71N jnf "Oebja orne va gur pbzvpf." Zbivat gur jbeq "orne" tvirf lbh HEFN ZVABE (na nfgebabzvpny yvggyr orne) naq PUNEYVR (Oebja va gur pbzvpf), erfcrpgviryl. Vg'f irel pyrire, bs pbhefr, ohg punyyratvat gb svther bhg.
