Before I read this book, I vaguely knew that Elizebeth Friedman was a skilled codebreaker but figured I would never know the deets since her work was classified. But this biography pulls some impressive stories out of some recently-declassified material. (Very recently, like as part of researching this book. I (spoiler alert) am reading Code Girls, a book about WWII codebreaker ladies, and there's a bit about Friedman, but leaving out info re-discovered by this book's author.) There's good stories here, whether you like spymasters or codes or history.
It starts out with the part of the story I knew from The Codebreakers. Thanks to her background in Shakespeare, she got a job assisting on a project to help decode the works of Shakespeare. Some folks had decided that "Shakespeare" was really Sir Francis Bacon writing secret messages in a secret code he'd developed; they'd figured out the system but wanted help with the grunt-work of decoding Bacon's messages using the system. Being clever and skeptical, she figured out how Bacon's code worked and figured out that the folks around her had deluded themselves. And she got really good at thinking about codes along the way. And she met man who she'd marry, who likewise became really good at thinking about codes. He'd go on to found a team of codebreakers that eventually turned into the NSA; and thus was obscure for a long time because the NSA was super-secret. But eventually folks found out about the NSA's existence, and thus we found out that William was pretty badass, code-breaking-wise. He's famous albeit in the narrow scope of folks who are into the history of codes.
Elizebeth was famous for a while, but then she dropped out of the spotlight. She formed a team of codebreakers within the Coast Guard, then part of the Treasury Department. She trained folks and led them. (And training folks was not so easy back then; there weren't many books about codes back then; and the practical books that did exist were written by the Friedmans.) At first, they were breaking rum-runners codes; this was during Prohibition. She testified in court against some alcohol-smuggling gangsters. Magazines wrote profiles about her… But then folks figured out that maybe this publicity wasn't such a great idea. If every media outlet is telling gangsters "Ha ha, there are these folks who can solve your simple codes," then gangsters might start using tougher codes. So, like William, her work became secret. But why didn't we find out more about her work later? The Coast Guard's a lot less secret-y than the NSA, right?
Thanks to this book, I found out why Elizebeth stayed obscure: her work was classified—and before it was unclassified, J. Edgar Hoover lied and claimed credit for her work for the FBI. So folks didn't necessarily think to check "Hey did Elizebeth Friedman lead her team to catch South American Nazi spies during WWII?" because everybody "knew" the FBI had done that.
But she did lead a codebreaking team. And she was pioneering signals intelligence and analyzing the intelligence and and her team decoded messages that had been encoded on an Enigma machine which is pretty much the most impressive thing a team WWII-era codebreaking organization could do. (It helped that the Nazis used the same key for 100s of messages instead of changing keys… something that they felt comfortable doing because even re-using keys, decoding Enigma-encoding messages was arguably impossible.)
(Oh, and as I found out from Code Girls, when the Coast Guard got absorbed into the Navy, that probably aided obscurity: The Navy doesn't respect desk jockeys as much as it does folks on boats. That's a reasonable attitude up to a point… but this group of desk-jockeys saved plenty of sailors' lives.)
There's also some bits that will appeal to puzzlehunters. E.g., William isn't happy to be cracking codes on his own after having gotten used to working on a team with Elizebeth; folks who've gotten used to puzzlehunting on a team will nod in agreement. And in case that's not puzzlehunt-y enough, it turns out the Friedmans were Game Control on puzzlehunts back in the day:
They took these games [Holiday cards with secret messages for their friends] further by organizing live puzzle-solving events that were famous in their social group throughout the 1930s. Some of these "cipher parties" were scavenger hunts that sent guests winging through the city. Elizebeth handed you a small white envelope. You tore it open to find a cryptogram.The solution was the address of a restaurant. When you arrived, you ate the salad course, then solved a second crytogram to discover the location of the entrée.
Folks who have organized puzzlehunts will be doubly impressed here: not only did the Friedmans write codes for a city-wide puzzlehunt, they handled logistics with multiple restaurants. I'm imagining walking up to a restauranteur and proposing: So here's the deal, these teams of people are going to show up sporadically; each person will eat a salad, and then leave without ordering an entrée… This must have required some convincing.
Uhm, but there isn't much about puzzlehunts. Puzzlehunt-friends who aren't history buffs maybe don't rush to read this book. But if you are a history buff, then by all means do rush. (Or maybe you already read it? This book was a 2017 bestseller; I'm late to the party, as usual. Anyhow.)
I wrote a set of puzzles for Puzzled Pint. And I'd like to write some notes about what it's like for a puzzle-designer to go through the Puzzled Pint process. But I'm just now getting around to it because the summary is: Wow, it's really smooth; these amateurs are indistinguishable from pros. So… if you're a puzzle-maker who's wondering if you want to send in that set of puzzles you've got rattling around in the back of that drawer, you should do it. But this write-up is boring. I want to blog stories that make me look like some kind of hero, triumphing over adversity against all odds… But thanks to the efforts of PM Neal Tibrewala and a horde of Puzzled Pint volunteers, I encountered no adversity and the odds were the friends we made along the way. I bet Neal had to deal with some obstacles, but I never heard about it. Anyhow:
Years ago, seeking a theme for some example puzzles to with Scott Royer's Anatomy of a Puzzle puzzlehunting how-to guide, I chose #, the "octothorpe" typographical symbol. The inspiration: Shady Characters, a book (back then, a blog) about some of the quirkier typographical symbols. This book wasn't just about #s: it also had @, ‽, ¶ etc. I'd idly noodled with the idea of coming up with puzzles for every symbol in the shift-numbers area of a USA keyboard… but that idea evaporated when I looked at the ( symbol and came up with zero ideas. But but the ampersand seemed like it had possibilities. So that was rattling around in the back of my head.
I had this conversation with a puzzle nerd: they liked Puzzled Pint because it gave them a way to spend time with some not-generally-a-puzzle-enthusiast person they cared about. And then I had a darned similar conversation with another puzzle nerd. Puzzled Pint had fallen off my radar: a couple of times I'd schlepped across town so I could meet with puzzle nerds and demolish a set of puzzles… in about the same amount of time I spent schlepping across town; I'd stopped going, stopped even bothering to check to see if it was at a nearby-not-so-much-schlepping bar each month; Yes, I'm a lazy terrible person. Puzzled Pint had fallen off my radar, but then these conversations showed me Puzzled Pint was, y'know, a force for good in the universe bringing people together and stuff.
Remember when I said, of Hunt for Justice…
So I wrote three rough-draft puzzles. Yes, Matt had only asked for one puzzle. But I'd once again run into an old issue: when coming up with wacky puzzle ideas, I couldn't be sure which of them were wacky-awesome and which were wacky-not-so-awesome.
In this situation, I'll make three rough-draft puzzles. Three puzzles to give GC a choice. Rough-draft only because if I go to the trouble to make a puzzle really good, I can get attached to my ideas, and less ready to take feedback; Seeing as how I wasn't sure which of these ideas was best, it was too early to get attached to any of them.
… One of those rough-ideas was an idea for a word puzzle like a spiral word puzzle but with letters of an "answer message" smooshed in there. The H4J folks chose a different rough-idea, one that had a cool reason to include toys in their box-o'-puzzles. But after the H4J effort calmed down, I thought back to that smooshed spiral and…
There was this thought: the smooshed spiral could be a puzzle called Back & Forth. It could be one of a set of ampersand-themed puzzles. I had some other ampersand-y ideas rattling around from before, maybe it was time to start writing those down…
As I write this, the Puzzled Pint instructions for authors say
For first-time puzzled pint authors, we suggest writing just a bonus puzzle first. This will let us get to know your style, and you can become familiar with our puzzle editing process, which can be a bit rigorous.
I didn't read those instructions until after I'd written down a bunch of ampersand puzzles, whoopsie. So instead of sending in one puzzle, having a long-suffering puzzle editor look at one puzzle and give me feedback that I could keep in mind while writing lots of puzzles… instead of doing that, I eventually sent in a whole packet of puzzles without the benefit of feedback on one puzzle, whoopsie. Thus, this probably made more work for Neal, d'oh.
So… don't do what I did. But in the interest of jotting notes: here's what I did.
I had the back+forth idea. I wanted to use a puzzle that did some kind of &-et substitution. (The & symbol gets it shape from "Et", because "et" is Latin for "and". I am desperate to use my knowledge of Latin in puzzles or anything really so that I can fool myself into thinking that four years of studying a dead language wasn't a waste of time.) And pretty much any kind of "this and that" puzzle would work.
I needed a simple meta idea. Meta-puzzles can get complex: There's a temptation to use an especially gnarly idea for the meta so that teams' final memory of a puzzle set is overcoming this great challenge. But experience on weekend-long hunts shows how this can backfire: folks who have been awake for the past 40 hours might not be up for your especially-gnarly challenge, and thus their final memory of the hunt might be grumpily calling in to ask for a hint. Puzzled Pint has a different way this can backfire: folks who have been knocking back beers for the past couple of hours might not follow your especially-gnarly idea. Percolating through my memories, I thought of Patrick Berry's book Adventures in Puzzling. It had sets of puzzles with metas—and I remember thinking that the first couple of metas had been pretty simple. I pulled the book off the shelf and
ripped off th saw how one of the meta ideas could inspire something via combination with something ampersand-ish.
For that back&forth puzzle, the way to design it was to choose a word, then figure out how to extend it by a few letters, then extend it by a few letters more. A skilled puzzle constructor like Nathan Curtis can do this "by hand". But when I tried, I got stuck a few times—I ended up writing a computer program to show me a bunch of choices. Maybe someday I'll be good enough at this stuff so I can just eyeball this stuff, but I'm not there yet…
Anyhow, I tinkered with puzzles on weekends. Progress was slow: I spent more time than I'd like to admit wondering "What was I thinking when I scribbled these notes last weekend?" and by the time I'd figured that out, there wasn't much time to work on things before it was time to pack up again. But then I got laid off and thus had uninterrupted time. Priorities, right?
I came up with a set's worth of puzzles plus some extra puzzles. Remember when I said before "…when coming up with wacky puzzle ideas, I couldn't be sure which of them were wacky-awesome and which were wacky-not-so-awesome. In this situation, I'll make three rough-draft puzzles." I didn't make three sets' worth of puzzles, but I did write some extras. This gave GC some "wiggle room." Plus it got me not-so-attached to rough-draft ideas: There was no way all of these ideas were getting in, so there was no reason for me to throw a nerd hissy-fit if one puzzle got turned down.
(Did you know that many movie titles have ampersands? In your head, it's "Mr and Mrs Smith" but when you look at the poster or IMDB entry or whatnot you realize it was "Mr & Mrs Smith" (2005) all along. That's the gist of one puzzle that didn't make the cut.)
I wrote in to Puzzled Pint GC saying that I had a set of too-many rough-draft puzzles for someone to look over and tell me which ones to discard and which ones to develop into not-so-rough drafts. I guess behind the scenes the GC folks figured out who would deal with me. Soon I heard back from Neal Tibrewala: he'd be the "PM" for this set of puzzles. Puzzled Pint GC has PMs, project managers: they're sort of like editors who also coordinate with playtesters, do page layout, and somehow keep progress ticking forward even if a puzzle author doesn't know how PP does things. This was good news for me: I'd had a chance to hang out with Neal volunteering at DASH in Austin a few years ago. If you're going to hear feedback about something, it's easier to hear it from someone you've met. Well, as long as they aren't a big jerk or something, I guess. And Neal isn't that, so yeah, good news.
I uploaded versions of the rough-draft puzzles for Neal to look at. A few days later, he scrounged up a solving partner and suffered through the set. Soon, he had feedback: a set of puzzles to use, a few puzzles to discard. Plus feedback on individual puzzles. Soon there was a mail thread per puzzle.
In general, I'd made the puzzles too complex for Puzzled Pint. I'd put in a recursive use-the-same-method-on-mini-answers-to-get-the-real-answer layer that turned a simple-idea puzzle into a two-hour grind. In another puzzle, I'd used a constraint in picking mini-answers that was very impressive… but didn't actually make the puzzle any more fun and furthermore led to some of those mini-answers being esoteric choices… Uhm, yeah. In my head I knew Puzzled Pint is looking for puzzles that are tractable to the somewhat inebriated and I remembered the ever-applicable Advice to/from a Puzzle Snob… but it was all too easy to fall into old habits and that's why it's good to have other folks around, e.g., Neal pointing out how these puzzles played out in practice compared to the usual PP puzzles.
There followed a round of edits.
Neal showed puzzles to other folks in the Puzzled Pint organization. There was one puzzle that relied on "common phrases". Now that Puzzled Pint has expanded into other countries, some puzzles built on a knowledge of "common phrases" have backfired since it turns out that some of those are actually Americanisms. Fortunately, Puzzled Pint doesn't just have players in those countries; it also has volunteers. A team of those lovely volunteers played through some of the dicier puzzles to check them for Americanisms. (Here, I was glad I'd kept my notes around when constructing puzzles. We didn't just send those volunteers the puzzle; from my notes, there were also some alternate phrases, so lovely volunteers could point out ones that worked better.)
Anyhow, there was this period of feedback and editing. The puzzles changed a lot, but that was part of my plan, starting with rough drafts. (I knew they'd have to change. I'd given puzzles answers that didn't fit in the meta-puzzle. This made sense because I'd started with "too many" puzzles; there wasn't "room" for all of their answers in the metapuzzle. But it meant that some of them would need to change so they could have different answers to "fit".)
Neal came up with a bunch of puzzle ideas. Like, feedback came back saying that some puzzle-layer didn't make sense, I'd mail to say "Uh, I dunno, I guess I will fix it by doing this" and Neal would say "Sure, or you could do that" where that was something that I wouldn't have ever thought of but it worked much better. Thanks to Neal, there's variety in this set of puzzles: I definitely have my go-to bag of tricks; he pointed out some tricks that I wouldn't normally think of. Oh! And Neal came up with the logo. And the hints. And made the online hint page work. And… Neal did a lot. I'm probably only aware of a fraction of it. There was a lot of stuff that I never had to worry about.
Once the puzzles were in OK shape for the next round of playtesting, I thought I had a few months to sit back. The Puzzled Pint GCs have a list of upcoming themes so that puzzle authors can "claim" a theme and not worry that someone else steals their idea. (I didn't do this; Unsurprisingly, nobody stole ampersands as a theme. But if I'd chosen the latest hit movie or whatever, things might have gone differently.) It was a long list, so I figured there were a bunch of sets ahead of me. But Neal mailed me saying actually the ampersand set was further along than others he was herding. (Did I mention that being laid off is good for finding uninterrupted puzzle-writing time? It's true.) Sure thing.
At this point, more Puzzled Pint volunteers playtested. (I'm not sure exactly who; all feedback went through Neal who anonymized things and perhaps softened harsh comments. But some of my
spie friends in the PP organization messaged to say "Oh, you wrote puzzles!" and such.) This ferreted out some lurking red herrings and let Neal compute the difficulty ratings for puzzles.
Anyhow, a final set of revisions. And then quality control at Puzzled Pint caught some errors in the final set of revisions and there was the final final set of revisions. And then waiting for…
I watched Twitter for #puzzledpint tweets. (You might think that for this occasion I would schlep across town to a local bar… but no, introversion won out.) Later on, I eyeballed the standings: more than 200 teams worldwide played this set of puzzles. That was plenty.
For a puzzle constructor working on a one-time event, finding playtesters is a hassle. But thanks to PP being a recurring event, they've found a bunch of excellent playtesters. Oh gee whiz it's really nice. Plus, Neal was an excellent PM, always a step ahead of me. It was a pretty lux experience, recommended.