Squeezing Out Refreshing Takes From Squeezy’s Creator

Hi! Can you tell us who you are and about your work?

Hello! I’m Jeff Chen, the creator of imsqueezy.com—a daily game that I launched very recently. Aside from making Squeezy, I primarily work as a writer and have authored several works, like Bridge Crosswords and my book series targeted for kids and pre-teens, Ultra Ball. I’m also the Vice President of the Paramitas Foundation, an organization dedicated to supporting humanitarian and scholastic causes.

Squeezy works by creating new words by inserting given tiles in between existing letters of the given words. Each level results in a special word that is made up of the tiles inserted, and every level has only one definite answer. After finishing all the levels of a daily session, a fact linking all special words will be revealed at the very end. Give it a go on my website and see how well you do!

In your experience, how does someone get into crossword puzzle-making?

Well, to be honest, by mostly falling ass backward into it. Back in 2008, I had just finished working at a startup that my friend started in 2002, and we were working crazy hours and just burning out pretty badly. By the time 2008 rolled around, things had been going much better, and so I took the opportunity to exit and got some flexibility that way.

At that point in my life, I was trying to figure out what the hell should I do with my life. What was I going to be when I grew older? I didn’t have any answers. So I started trying a whole bunch of things, including writing, crosswords, philanthropy, travel… you name it, I’ve probably tried it. I even did some coding too.

And I had met my girlfriend at the time, who’s now my wife, and we started exploring some things. And she had always been a crossword solver her entire life. She was able to solve the New York Times crossword in about fifteen minutes. And on some of our first dates, I had just started doing crosswords out of curiosity. And we would sit down with two copies of the NYT crossword on a Sunday while having coffee and breakfast.

She’d be done quickly, and sometimes, even three hours later, I’d still be sitting there trying to figure out what the hell was going on. But that solving time got shorter and shorter the more we played, but she’d still always beat me by quite a long margin. Then finally, one day, I just decided to try making one of these because, like everybody, I saw the movie Wordplay and realized that they’re human beings behind this thing.

I figured, okay, well, I’ll try it. And I made one. It was terrible. Thankfully, my wife Jill is very, very sympathetic and empathetic. So she gave me a lot of support and I submitted it anyway, even though she gently and kindly tried to say that it was terrible. And Will Shortz wrote me back pretty quickly actually, and said, “Hey, thanks for the submission, I think you have a lot of promise.”

History behind puzzle creation

However, there are some rules with crosswords, like you shouldn’t make up your own words and stuff like that. I couldn’t fill it because I was doing everything by hand and graph paper, and I was trying to figure out something that would go well with SUN. Sun. Sun. Son. Then I thought maybe Apollo must’ve had a child, right? So I put that in, and of course, Will said that it would never fly and that I should try again. So I kept on trying, and he was really polite and encouraging the entire way. All throughout the submissions, from one to twenty, every single one of them was rejected, but he was still so nice and kept propping me up to keep going.

Finally, Jill had an idea for the crossword, and she wanted to figure out how to lay it out. Then I said no and that I’ve learned a lot, so I told her that she couldn’t do it that way and that it wouldn’t work. But she kept saying that we should just try it and see what happens. And I kept going no no no, really, until we ended up trying it, and, of course, it just serendipitously worked out.

Puzzle creation thought process

When it comes to creating puzzles, the biggest thing by far is to try and figure out a good idea, or what we would call the “theme” behind a crossword. It’s sort of like Connections, like, some way that links really good and colorful phrases together. You have to make and generate that’s pleasantly surprising or would make someone say, “Oh, that’s pretty cool!”

So you have to start with that and then the process of laying out a crossword puzzle then gets into technical details, which isn’t that difficult to learn but takes a lot of practice to go through and get good at. Once you get the right tools in place, it really doesn’t take nearly as long as most people would think. 

I think a 15x15 crossword these days can come together in two or three hours, once I get the idea, the theme phrases and if there’s nothing really difficult about a particular set of constraints or anything like that. So from weeks, I’ve learned how to do it in hours. But I’m not perfect, it’s still a learning process even after fifteen years of doing this. I recently heard from the editing team that they liked an idea that I sent in, but they wanted to change some of the theme entries and wanted me to put in some other ideas I could sub in.

Did you dabble into other things outside of puzzles?

Yeah, I remember something called The Game back in college, and it was based on a movie, I think. But overall, it was this 24 to 48-hour kind of activity whirlwind and you didn’t go to sleep. You got a clue with some sort of code that could be mathematical, foreign language-based, or anything under the sun, and when you solved it, it would tell you where to go next and it was so compelling. I loved doing it.

And after graduating, work sort of got in the way of things. But then I started getting more into these online puzzle hunts and the same sort of thing. There’s one called The Puzzle Boat, which a group of us look forward to every year, where it’s roughly 100 puzzles that all work together to get a set of meta answers. And those meta answers would all feed into each other to come up with a meta meta! You’d think that if we had like six people, we’d just crank through them. But one of the puzzles this year took me about six hours to figure out, and it was all about the way that bell chimes work. And it involved the switching of different pairs of bell chimes, and you had to go back and figure it out. I even had to write some computer code to figure it out, like, how would I determine the initial setting of these bells.

So, super, super fun and I’ve met a lot of really interesting people that way. One guy that I got to be friends with online that I’ve only seen on a Zoom call once. I’ve never met him in person, but he’s great, and it turns out he’s the director of The Last of Us, and a couple of years ago we were working on this puzzle together. We’re trying to figure out this thing out and it was just breaking us. Finally, he said that I’d try to work on it the next day, but that he had to catch a plane to the Emmy Awards, and another friend of ours said that I should come along with them. So I looked him up and sure enough, he’s Craig Mazin, who has written and directed The Last of Us, as well as HBO’s Chernobyl, which was amazing.

I just love the puzzling world, and it’s just fun ending up creating connections like that.

Who’s Jim Horne and when did XWord Info come in?

XWord Info was found by Jim Horne. Jim started it roughly fifteen years ago, and five years into it, he decided he wanted to stop. And by that time, he had made it such a valuable tool for a puzzle constructor like me. I was on there fairly frequently using the Finder Tool to look up specific letter patterns that I needed to fill. And when he announced that he was going to shut the site down, I realized that he only lived about half an hour away from me.

So we coordinated and got together, and we liked each other and had super fun. And he said, “You know, I just can’t spend the time doing this anymore. It’s just way too much of a commitment. And frankly, it’s just not that much fun for me anymore.” And a lot of people had come forth and asked if they could take over, but Jim always asked whether they knew SQL, C++, JavaScript, HTML and how to work a database?

Expectedly, a lot of people didn’t know anything about those things, including me, but I liked the idea of the challenge of learning it all and trying to figure things out. And Jim looked at me and said that we could try it. But I knew he was thinking that there was no way I was going to commit to it. But we got together, and I’m a recreational coder, so I just kind of hammer things away with total brute force. And he chuckled at some of the things I was trying to do.

So as time went by, I decided that I wanted to write some commentary just to kind of keep things positive and light in the crossword world, or as we’d call it, “crossworld”. And I just had a lot of fun doing that. I even asked Will Shortz whether I should write something that pointed out the positives of each puzzle because there’s already a lot of negativity in the world. And he said that he didn’t want that, and wanted me to give fair opinions that give both the positives and negatives. No one’s going to believe you or put any stock in what you say if you’re just pointing out the rosiness of the crossword. So yeah, I try to look at things from both sides.

I come from an engineering background, so I tend to think things through technically. So a lot of my blog commentary in XWord Info was about the technical aspects of putting together each puzzle. And it was really fun for me to do that for ten years, but when I hit the ten-year mark, I think I had gotten repetitive and I just didn’t have that much more to say. Aside from that, I had Squeezy as a project that I was more interested in spending time on, so I hung it up. That freed things up for me quite a bit!

But Jim and I still remain good friends and he’s been working on Squeezy with me, developing the server-side applications that we need to get the data to users every day. And it’s been really fun having this new project with him.

Status of XWord Info

In regards to running XWord Info, I’m not there on a daily basis anymore like I used to. Jim still uploads the puzzle that we get from the NYT team, but it’s more like whenever you can get around to handling them. So maybe every other week, once a month… we’d upload a bolus and try to keep it up to date. We do still maintain our word list, and I upload a big bolus of words once a quarter I think, just to make sure that we have kind of the most up-to-date, the freshest entries and also things that have fallen out of favor.

As for costs, we do have a support subscription model in place, but it’s just enough to cover the costs involved in maintaining it. There’s just so much database management and bandwidth that we have to pay for. So it’s very expensive to run, unfortunately, and we tried to figure out ways of either selling it or porting it over somewhere, but things just haven’t worked out. We’re hopeful that we can keep it running at this level for as long as possible, but we can’t make any guarantees.

How did Squeezy come to be?

I've had a lot of different ideas for games over the years, and most of them just feel like sort of derivative or not that interesting. And I've tried to develop a couple of things just with PowerPoint slides so that I get a rough idea of it all. And this particular one, I’ve always been interested in crosswords where you can alter the words themselves and turn them into different, valid words. It could become longer or shorter, but it would all still be valid, and there’s just something neat about it for me, like, it’s some kind of wizardry.

So I came up with this idea of inserting letters into words, and I talked to David Steinberg, who’s over at Andrews McMeel, and he really liked the idea and gave me some suggestions on improving it. I had given him a PowerPoint with a couple of examples of words that could transform into other ones when certain letters were inserted in them. And he noticed that the example I gave had six letter insertions. Those six letter insertions could be anagrammed to become a real word, and he suggested that I should go with that.

And they were interested in kicking it around and seeing if they could prototype and test it, that sort of thing. But they just didn’t have the bandwidth or the resources to. Then their programmer quit, so they just said that they couldn’t pursue it. So I sat around for a while and I thought to myself that I didn’t want to wait. I don’t want to pay somebody to do this, especially because I wouldn’t know what they would end up doing. Plus if I had to change anything, I’d always end up having to contact them and go through the trouble of explaining it.

So… why shouldn’t I just figure it out myself? And it turned out to be pretty difficult. I watched a lot of videos on YouTube. I started with Wordle clones, so there are a lot of people out there who noticed how popular Wordle was and then figured it out. I’d tune in on videos on how to make a Wordle clone in an hour using basic HTML, JavaScript and CSS, and I watched about half a dozen of those, and I retained only like 5% from each.

But I kept on watching some more, and picked up another 10%, then on and on, until I got to the point where I could start putting together some basic HTML. And I know enough basic computer programming to write some JavaScript behind it. And it took me probably two months to come up with my very first prototype, which I was so proud of. So I took it downstairs and I showed Jill and the kids, and they thought it was pretty cool. Then they had to set back some stuff and reset while fiddling with it, then I told them you had to reset the entire game to do so. And Jill was like, “Oh, okay… great.” And I was a bit surprised, because I learned all this stuff and created something, so how was she not impressed?

Then we went out to dinner a week later and kind of mentioned it with a laugh, and she said she’d never thought about how much effort it took to figure out something as simple-looking as that. And honestly, I understood where she was coming from. You’re just moving a tile from here to there, there’s no undo, and you can’t remove the tiles once you’re done. You don’t even have a reset button or anything like that.

So I went and made a second prototype that had a lot of these missing features. And by that time, I kind of had gotten up the learning curve and gotten the ball rolling so I could figure out a lot of logic behind the other features that were needed. And I kept progressing to the point I ended up creating results that had very mixed opinions from the people I showed it to. Jill was always supportive and provided me with different kinds of suggestions and insights, and I even had my identical twin brother have a go at it and he said that, while fiddling with it, he admitted that he would’ve never gotten that result and it just frustrated him.

Timeline of Squeezy

The first prototype for Squeezy was built around September and October, and the first daily launch for the public was on January 31st. Back in November, I sent Jim a screenshot I had mocked up saying I planned to launch Squeezy #1 on February 1, and he said, “Okay. If you want to work really hard over the holidays and bang your head for most of January, that could be realistic.

And I thought that I could do it. And I did, somewhat.

Visual theme of Squeezy

One of my wife’s Wordle group friends originally came up with the idea of calling it Squeezy, because you were squeezing in a letter into a word. Plus my kids are always saying easy, peasy, lemon squeezy all the time.

When I first started, the different levels were just easy, medium and hard, and then I thought that I should do something thematically and try things like breezy, wheezy, squeezy, queasy… stuff like that. And I put a lemon logo in as… well… lemon squeezy or something. Then my wife’s friend suggested that I should just go all in with the lemon thing. And there it is.

Technicalities behind Squeezy

I manually create all the puzzles myself, mostly the same approach I would come up with an idea with for a crossword puzzle. It doesn’t have to be as tight or as interesting an idea sometimes. And there are only three answers instead of four answers. So it’s pretty easy for me to come up with dozens, hundreds of ideas that I can kind of play with and call through. But then, at that point, figuring out how to break up the secret words and then what letters fit into where to make those secret words up also took some effort.

At first, it took me about two hours to create a puzzle, but these days, I’ve picked up steam and created some spreadsheets that help me figure out what possibilities are available for each letter or each pair or triplet of letters. So it probably takes half an hour to crank things out. Back when I launched, I had 50 games made, and now I’m at roughly 180, so about six months’ worth of games are stockpiled and ready.

There was also a difficult problem to think through, like what if somebody came up with a different yet valid solution to what I created? And Jim and I have spent a lot of time trying to figure out how to programmatically check for those things. He’s helped me with a couple of validation steps to make sure that there’s no varying answer. But it really is hard to predict.

As for the modes, there have been many calls to make Wheezy Mode to be a lot harder. So I made two modes. Breezy Mode is for those who want to casually play or folks who want to have an easy time. Wheezy Mode is for people who want a real challenge, and I always have an extra tile. That extra tile is meant to interfere with your thinking and mislead you.

How hard was it to promote Squeezy?

I’m terrible with social media, so it was a very big challenge for me. Thankfully, I’ve built up a big network of crossword friends, most of whom I’ve never even met. Going to the crossword tournament every year, back when I didn’t have kids, I was able to just establish a lot of friendships, and that was the biggest way I could spread the word about Squeezy. Surprisingly enough, my kids’ elementary school teachers also really got into it and started sharing it with some of the parents. So yeah, it’s been great and it just spread organically with a lot of local people and acquaintances.

Is there any advice you’d like to offer passionate people who want to enter game creation?

Definitely! I myself entered this path and I’ve encountered a lot of things I can give advice on:

1. Find the line between persistence and stupidity

See, I got Squeezy to the point it’s in now because I’m just very stubborn and quite boneheaded with some things. Finding that fine can be your unconscious guide to help you continue and keep moving forward. You’d be surprised how dumb ideas actually work out in the end, just look at how me and my wife conceptualized this thing—I even kept on saying no to her initial suggestions!

2. Keep swinging at things until you feel something that truly hits

When I told you guys how I stumbled into puzzle creation, I truly meant that I had no idea what I wanted to be or what I should be doing. You name it, I tried it, just to see if it would click with me. That also applies to game and puzzle creation. It all starts with how you feel it all out, and when you know it’s something solid to work on, everything just continues from there.

What’s the future for you?

Our biggest goal right now is to reach 50,000 daily players at the end of the year. Right now it’s around a hundred playing it daily, but we want to reach that goal as much as we can. I’d also like to have it featured in the NYT, and we do give them a nudge nudge at times, telling them that people are into it, but so far, no luck. But we’re open to talking to anyone interested in buying it or something. I’d really like to reach a point where I’m paid to do this, since my career in writing is similar to how I developed my crossword and puzzle skills—through persistence.

Do you have a favorite game to play?

Right now I’m actually really into Connections. I know Wyna Liu from the editing team at the New York Times, and I just love the way her mind works and makes these links that I just never would’ve thought of. I get to the very end and I have four squares left, and I’m like; “I don’t know, tell me, go ahead!” And I’m usually surprised, so yeah. Most of the puzzles on the NYT game site have one editor who’s kind of in charge of everything, and Wyna’s one of them. I’ve never met her in person, but we’ve exchanged a lot of emails and she’s super, super nice, so it’s been great to see her have success with Connections.

Where can we learn more about you and your work?

You can check out my website at jeffchenwrites.com. The books and things I’ve written are featured there, and you can also contact me via the About Me page. And I’m always open to different ideas and whatever I can do, and however hard it is that I have to work on something, especially when it comes to my creations, I’m absolutely willing to do it. So if anyone out there has suggestions, please drop me a line and I'd love to chat about anything that might be interesting.