How I Made Word Play

Hi, I'm Mark Brown from Game Maker's Toolkit, and I am chuffed to bits to let you know that my new game, Word Play, is out now on Steam. It's a game where you spell words to score points, so you can unlock powerful perks that let you score even more points. It's a game about spelling, and strategy, and synergies, and special powers. It's got a classy art style, a boppin' soundtrack, it's got accessibility settings, and difficulty modes, and loads more.

I hope you'll check it out. Now for my first game, Mind Over Magnet, I shared the entire process of making it across 15 or so episodes of my YouTube series, Developing. But for Word Play? Well, there just hasn't been time. The game has come together so quickly that I've barely had a chance to tell you how I made it! But better late than never, right? So here it is. Here's how I made Word Play in just seven months.

Okay, so the idea for Word Play - which was essentially to combine the poker roguelike Balatro with a spelling game like Scrabble or Bookworm Adventures - the idea has been rattling around my head ever since I played Balatro at the start of 2024. But in December, I finally had a reason to try and make it for our secret, Patreon-only GMTK Game Jam. So I bashed together a game about spelling words, scoring points, and unlocking rule-bending modifiers.

It was called Wordy, and it was actually pretty dang fun. The response was great, I had a really good time working on it, it clearly had a lot of potential, and if I had any intention of turning this into a real game, well I had better get on with it before someone else has the same idea. And so, shortly after the jam was completed, I opened up a fresh Unity project and started working on a more fleshed out version. Now I decided that the best thing to do at this early stage was to not worry about things like perks, or rounds, or rogue-like elements - but to just build a really good spelling game.

Something where it was nice and juicy to spell words. Something that would work seamlessly across mouse and controller, and keyboards, and touchscreens.

At the beginning of January, I continued to build that basic word game - adding in features like a shuffle button to mix up your tiles, and a letter bag. But then, with the spelling game mostly done, I started adding on the rogue-like features like the shop, and the perks. I added the basic architecture for special tiles, and started making modifiers which are your passive perks - think jokers in Balatro, or boons in Hades.

By this point the game had 11 perks, though they were mostly just there to test different functions like having a modifier that can boost a tile's score, or the whole word's score. In February, work started to slow down - mostly because I was just focusing a lot of time on making videos for GMTK. I added a few features, like the ability to sell modifiers, but that was it. Then midway into February, disaster struck. I was scrolling through my YouTube subscriptions and saw a video on Second Wind with Yahtzee playing a game called Wordatro.

Which I didn't even need to click on it to know what that one was all about. YAHTZEE CROSHAW: Ah ha ha ha ha! JESS HOOPS: Oh, beautiful! It was another game that mixed Balatro with Scrabble. Damn. Now, I briefly considered cancelling the game, and I definitely regretted not showing this thing earlier, but ultimately I decided that I can't control what anyone else makes. The only thing I can control is how good my game is.

So I decided to keep working on it, and just make the game as awesome as I could make it. Oh, and then later that same month, I listened to a completely random podcast about screenwriting, and one of the hosts said... JOHN AUGUST: We have a game for you to try out right now. So it's called Birdigo. Birdigo, like the thing that flies. It is a roguelike deck builder where you're trying to make words. Which you might think would give me a heart attack.

Another one! But actually, this made me feel better. Because between Birdigo and Wordatro, and over the next few months, many more word-based roguelikes all having the same basic idea of Balatro plus words but going in different directions.... well it feels less like I'm competing with and, in some people's eyes, copying a single game, and more like I'm part of a growing genre. So crisis averted, and I even ended up chatting with the devs of both Wordatro and Birdigo and, well, I'll have more to share on that down the line.

Suffice to say, we're all good.

In March, I finally gave the game a proper name. Here's how that happened. You see, the game has, essentially, lives. But you use a life every time you spell a word. So it's a bit weird to lose a life when you do something correctly. That just feels weird. So I asked my Discord users for alternative words to "lives", and user Fly suggested "plays". Which sparked the idea for Word Play. Now, when you're making a word game, as it turns out, every single possible name has already been taken by some other game or toy.

But Word Play? Well, Word Play doesn't exist on Steam. So you know what? I'll take it. I'll snap up Word Play. Anyway, I started March by continuing to add more roguelike elements. So I added upgrades. They're your active perks which can change, improve, duplicate, or discard tiles. I added in all the special tiles - golden, diamond, emerald, potion, and dot tiles - which all have unique gameplay. I added in the special rounds, which are, well, special rounds with unique requirements like "your word must have six tiles".

I added modifiers that can multiply the score. And I added reroll to the shop. So now that I had made the spelling game, and coded in all the roguelike stuff, I could tie all of those elements together into a full gameplay loop - so that it could be played as an actual game. Which was especially important with the Game Developers Conference coming up at the end of the month. So now I could start getting proper playtesting feedback on this new version of the game.

I distributed it on the GMTK Discord, and took it on an iPad when I did talks at schools and colleges. That led to lots of changes based on user feedback and requests. I made it so you could rearrange the tiles on the board. And I made it so you get extra bonus points if you spell especially long words - because otherwise you barely get any points for dropping a single mega word. And then I took it to GDC, where I spent the entire week playing it with other developers and got fantastic feedback from the devs behind games like Patrick's Parabox, Spell Tower, and Alto's Adventure.

Which led to even more changes. Also I rode in a driverless taxi, and had my breakfast delivered by a robot. San Francisco! Also in March, I started reaching out to my collaborators on the project. I emailed Zach, who did the music for Mind Over Magnet, and offered him first dibs for the music on this game. And I DMed Jay on Discord. He made a spec video of the game with different sound effects, and they were so good I hired him to do the whole game's sound design.

I also added in Unity Analytics to help with future balancing. I created the Steam store page, though kept it private. And I brought the total number of perks up to 52. Basically, I went into hyperdrive. Looking back on my notes, I can't believe how much I accomplished in one month. I was racing through this game. Development was coming along so well. And then I broke my arm.

Yeah. April was a bit slow. I had to stop development for several weeks as I waited for my elbow to heal. Jay finished the sound effects, but it would be a while before I could put them into the game. But I was able to slowly get back to work. I added in a new special round where you have to play a mandatory tile, or else lose two plays. I added compound tiles, letting you play the letters I N G with a single tile. And I added in the first modifier that was based around the specifics of the word you spelled.

It's this one, which gives you a boost if the first and last letters are the same. These would prove to be very important in the game's overall sense of strategy and balance. But I think I'll talk more in depth about that in a future video. Oh, and this also brought the total number of perks up to 74.

May started out with disappointment. I managed to get in touch with the chap in charge of Apple Arcade, which is the iPhone maker's Game Pass-like subscription service. And so at his request, I sent over a build of Word Play on TestFlight. Ten agonising days later, Apple Arcade wrote back to pass on the game. Which was a huge shame, but hey, it was cool for them to play it at the very least. But okay, with that out of the way, my goal throughout the rest of May was to get a demo ready for Steam Next Fest in June.

Which meant fixing lots of bugs, adding lots of polish, and chucking in a lot more cool features - like the glass tile, which can be used to have a one-time copy of any other letter. And so towards the end of May, I was finally ready to reveal the game to GMTK viewers like you - with a demo, an announcement video, and a let's play of the game. Which was amazing. The response was overwhelming. 20,000 players in a week! The game got picked up by various YouTubers and streamers, including Yahtzee on Second Wind, him again, and the world's biggest Balatro streamer, Northernlion.

NORTHERNLION: I bet there's like some sick Italian wines you can spell with these letters, man. I also got an email from… uh, Apple. Yeah, not their Apple Arcade team this time, but a different group at the company, who reached out to offer some help, listing it on the main App Store. More on that in the future, because right now I'm focused entirely on today's Steam release. Plus, all of those demo players sent me so much great feedback and ideas, which led to all sorts of cool stuff - like being able to submit words over 10 letters, and being able to petition for more words to be added to the dictionary.

And so at the end of May, the game was in a very good state. And the game now has 90 perks.

By June, the basic underlying game was essentially done. It was absolutely a minimum viable product, or MVP. It had a fully functioning game loop, it had saving and loading, options, and difficulty modes. Still, I added a few more quality of life adjustments, like the ability to peek into your bag or at the board during the perk screen. Some accessibility features, like alternative controls, dyslexia-friendly fonts, and spelling suggestions.

And I even added the groundwork for people to mod in custom dictionaries and letter bags if they want to play with non-English words. But, with all of that done, and with Steam Next Fest taken care of, my job now was to simply make stuff. So I added new tiles. The mirror tile. The exclamation mark tile. The locked tile. I added new special rounds, like one where the maximum word length starts at four, and increases with each word you submit.

And I added perks. So many perks. I started by bringing the game up to 100, but then I thought of some more and took it to 120. Then I had a few more ideas, and it was like, why stop at 143? I might as well take it to 150, that's a nice round number. And over the rest of the month, I slowly added a few more until I really, really had to stop at a whopping 160 perks. And look, I've got even more ideas, but let's save those for future updates.

But at this point it was time to start doing the final quality assurance, or QA. I brought in a group of helpers - a mix of GMTK patrons and people who gave particularly insightful feedback about the demo, and had them run through the game over and over and over again, looking for problems. They found bugs. They found typos. They found perks that were overpowered, and perks that were pointlessly weak. They helped me balance the difficulty curve, suggested Steam achievements, and pushed me to add in features that I had been way too lazy to make previously.

To be honest, this lot were the real MVPs - the most valuable players. Without their help and guidance, the game that just launched would be buggy, unbalanced, and just significantly less fun. So a huge, heartfelt thanks to all of them. Cheers.

And then finally, it was July. It is July! And the game is out now. And so that is the story of how I made Word Play in seven months. And if you compare that to the development of my first game, Mind Over Magnet, which took me well over three years to produce, well obviously things went a lot more quickly. Now, of course, they're very different games. And I was also a very different developer back then. I literally had to learn how to use Unity while I made the game.

But it's not just that. I also thought carefully about how to develop Word Play while I was making it. To make sure the process was streamlined, and to ensure that this game didn't devour years of my life in the way that Mind Over Magnet did. Now I'll talk more about this in future episodes of Game Dev 101, which I'm looking forward to getting back to soon. But as the Cliffs Notes version, it really came down to two things.

Number one. I was ruthless about the game's scope. I would routinely make decisions through the lens of which choice is going to take the least amount of time. For instance, I could have made unique art for all 160 perks, like the jokers in Balatro. And I would totally accept the criticism that the game is worse off for not having that. But ultimately, using a more simple set of icons saved weeks of time. And it was the difference between the game coming out while I still had some fuel in the tank, versus causing the development to drag on for many more months.

And number two. I would develop the game in a kind of layered approach, where I would focus on one part at a time. For instance, in December, I focused exclusively on just making a good spelling game. Then when that was done, I focused on getting the roguelike game loop complete. Then I just made perks all day, every day. And then I just did bug fixing and balancing. Now of course, none of that is really true - I was still adding in features from previous layers as development went along.

But for the most part, I did just focus on making the bones of the game before turning my attention to the content. Which made development faster and more agile, as I was always building on the really strong foundation of the previous layer. And so ultimately, while I'm really proud of this game - and I think it is genuinely a lot of fun and I hope you dig it - I think I'm even more proud of the fact that I made a video game without losing my mind.

By taking everything I learned from Mind Over Magnet, and really planning out the development instead of just making stuff willy nilly, I was able to take this game from idea, to prototype, to playtesting, to QA, and to release in just seven months. So I'm going to leave it there, so you can actually go and play the dang game. It's out now on Steam for PC, Mac, and Steam Deck. There's a 10% discount for the first week, and there's still a free demo if you haven't tried it yet.

And now, well, give me a sec to fix all of the inevitable bugs, and then I'll see you later this month for GMTK Game Jam 2025. I will not be taking part. Do not let me. I cannot be trusted. Bye bye!