What it's like to launch a game on Steam

The final episode of Developing

Hi. My name's Mark, and I made a video game about magnets.

So, this is it. The final instalment of Developing. Mind Over Magnet has been on sale for three or four weeks now, and so now I can tell you everything.

I'll tell you what it was like to launch a game on Steam. I'll tell you exactly how many copies of the game I sold. I'll tell you about all of the bugs that came up in the first week. I'll tell you what the most negative reviews of the game said, and I'll tell you what I learned from making this game that will help me if I ever make another video game. I will tell you it all.

But before we get into that, it's time to talk about the elephant in the room.

Okay, so a big goal with Developing was to give you a super transparent look at what it's like to make your own video game. And so I've tried to show you every step of the process, from learning the game engine, to making your first prototypes, from running feedback sessions, to wrangling with design problems, from hiring a composer, to releasing the game on Steam.

I've tried to show you everything so that you can see what it might be like if you too were to make your own game and release it on Steam.

But there's a few aspects of my development story which simply aren't going to be applicable to the average game dev. And that's because by virtue of having a rather popular YouTube channel, I've been afforded a number of advantages that probably aren't going to be applicable to you. For instance, if you run into a programming problem when you're working on your game, you probably don't have a huge Discord full of people willing to help you out at the drop of a hat, or direct access to the devs at Unity.

And if you need playtesters for your game, you probably can't summon hundreds of them with a single tweet.

And most pertinently for the discussion today, if you were to make a bog-standard 2D puzzle platformer with basically zero external marketing, you can't be expected to sell the same number of copies that I have.

I say none of this to put you off from making your own game. In fact, I really hope that you are inspired and influenced by my story to go out there and try and make your own games. But I think before we dive into how the game actually did, it's worth putting that asterisk on these numbers. Putting a caveat on these results to say that some, if not most, of the sales come by virtue of me having this channel, and not the quality of the game itself. And so if you were to make your own game, you need to set your expectations realistically.

Is that cool? Cool. Let's get into it.

Launch Day

So in the last episode of Developing, I hovered my mouse cursor over that big green release app button, and then on November 13th at 5pm, I clicked it.

Now a thing about Steam is that Valve can be a little bit slow with the numbers. For instance, the wishlist stats are only updated once per day. So sometimes you have to wait an agonising 24 hours before you know whether your Next Fest demo or your YouTube devlog have had any impact on the numbers.

But Valve does give you a real-time look at exactly how many people are online and playing your game right now. Which, while they're not exact sales numbers, does give you an idea of how popular your game is. So I decided to park my bum on that screen and watch as the numbers changed. And so I saw 12 people were playing Mind Over Magnet, and then 50 people, 100 people, 200 people, uh... 150 people.

Yeah, like I say, this is not the sales numbers, this is just how many people are online and playing and so that number is going to fluctuate up and down.

However, I then accidentally pressed the back button on my browser and went back to the main Steam stats and realised that - oh, right, unlike the super delayed wishlist numbers, Valve does actually tell you pretty much up to date how many people have bought the game! And so I could see that 800 people had bought the game in the first couple hours. And then 1000 people. And 2000 people.

And then the rest of the evening was just a bit of a blur. The game ended up on the New & Trending tab on Steam. It was in the top 25 of the top selling games on the platform. Because Valve had OK'd the game as Steam Deck verified the day before launch, it had a prominent position on the handheld's front page. And most mind-boggling of all, the game was in that massive carousel of big images at the very top on the very front page of Steam. Getting the game hundreds of thousands, perhaps millions of eyeballs.

And so by the time I went to sleep, the game had sold 5000 copies, or 10% of its 50,000 wishlists. And by the end of the first week, the game had sold 10,000 copies, or 20% of those wishlists. And by the recording of the above video, the game had sold 12,446 units.

Which is absolutely incredible. Like beyond anything I was expecting or dreaming of, to be honest. So thank you so much. If you were one of those 12,446 people, thank you so much for showing the game an enormous amount of support. You are truly amazing.

Money

Now wait a second. Hold up a minute. Because it might be tempting to take that sales number, and take the game's price tag, do a quick calculation, and think "holy moly Mark, that is a lot of money!"

And yeah, it is a lot of money. But it's also not the full story. So let me explain where that money actually goes.

Now it's a little bit gauche to talk in actual dollar values, so let's say that the game made 100 Monopoly monies.

Now, the first stuff comes at the point of sale. Stuff like sales taxes in various countries, credit card fees and chargebacks, and refunds, which I'll talk about later. That for me ended up being about 10% of the income, so we're down to 90 monies.

Then Valve would like its cut. The company wants 30% of your revenue as a little thank you gift for directing millions of people to your game and running all the infrastructure and all of that good stuff. So that's 27 to Valve, leaving us with a slightly lighter load.

Now this money will end up in the GMTK business bank account at some point, I imagine. It's not in there yet, but fingers crossed. But that's also not the full story, because we do need to think about operating costs. You know, how much did this game actually cost to make?

Now GMTK and Mind Over Magnet is a bit of a strange beast in that the game didn't really cost me anything. The game was funded by the incredible GMTK backers over on Patreon. You could make the argument that spending time making this game meant I had less time to work on YouTube videos and thus less YouTube ad revenue and fewer opportunities to recruit new Patrons. That is certainly true, that is what happened, but it's a little hard to calculate and a little murky, so I'm just going to ignore that.

But what I can't ignore is that I had to pay my contractors, including Zach, who made the amazing soundtrack for the game, and Grayson, who made the dope 3D promo artwork, plus there were some other fees here and there like some Unity assets. And so without revealing exactly how much I spent on these chaps, I can say that I need to lose another 5 Monopoly monies.

And then if you are lucky enough to have made a small chunk of change, well, you know what's going to happen. The tax man is coming. Again it's hard to calculate exactly, but living in the UK and paying corporation tax, that's going to be another, I reckon 15 monies to His Majesty's Revenue and Customs.

And so we are left with, I think if my calculations are correct, 43 Monopoly monies.

Which is still an incredible amount of cash. This will fund some more video games perhaps, or just pay a few mortgage repayments. Still a huge amount of money, but it is a little shocking and a little sad to see this compared to the 100 that we started with.

And so if you take nothing else from this whole Developing series, I'd say this. Consider all of these hidden costs and taxes before you put a holiday to Barbados on your credit card.

Bugs

Not that you're going to be going anywhere anytime soon, because an influx of new players means an influx of new bugs. And as soon as Mind Over Magnet was out, people were swarming onto the Steam forums; onto the GMTK Discord, which has a public room for Mind Over Magnet bug reports; and into my email inbox to tell me all of the bugs, problems, and glitches that I had missed with my QA team during our last few months of checking the game.

Now fortunately there were no massive game-breaking problems that affected great swathes of the game's sales base. I mean, probably the worst thing that happened was that achievements didn't unlock on the Mac version of the game, which was a bit of a boo-boo, but not the biggest deal in the world. So I could take the first four days of the game's launch to just relax, recover, digest, and celebrate the game's launch.

But come Monday morning, it was time to fire up the old spreadsheet and start listing out those bugs and start fixing them.

So first on the chopping board was to get Mac achievements working. They're now all good. There were about 20 levels that people had found various problems with. Most of those could be fixed by just slightly moving things around to avoid cheesing and unintentional solutions, though one level, 3-11-Spiky, I actually ended up completely redesigning simply because by the time I was back in Unity with a fresh perspective I thought of a much better way to design that puzzle.

I mean I could probably redesign every level in the game to be honest, but that would be ridiculous.

Some people were also having trouble with the game not showing the right controller symbols if they were playing with a PlayStation or a Switch controller, so I just put in an option where you can just pick exactly which controller glyphs should be showing if the automatic picker isn't working properly. And I also added in another feature which was to replace the world select screen with a level select screen. So once you finish the game you can go back and play any level from the game without having to start a world from scratch.

But there were a few bugs that I actually decided not to fix, simply because they were being used as speedrun techniques. These were the sort of things that 99% of normal game players would probably never run into, but speedrunners could use physics glitches and bugs and my poor programming skills to make Uni run faster or jump higher or to bypass puzzles or to skip past bits of the game.

I decided that it wouldn't really hurt the game to keep those bugs in and it would make the speedrunning community much, much happier and more interesting. And so while I was tempted to watch the speedruns to look for bugs that I could fix, I decided in the end to leave those things be. And even beyond that I decided to keep the release version of the game available, thanks to Steam's beta drop-down doohickey, so that it's always available for speedrunners and also for preservation so you can always see what the game was like when it launched.

And so with a patch, and then a much smaller patch to fix the problems I made in the first patch, the launch window support for Mind Over Magnet is now, I think, done. Which is a relief.

Though I know that my time with Mind Over Magnet is not over yet. There will be more bugs to fix. I would like to do a console version next year, at least on the Switch, but no immediate plans so please don't hesitate to go buy the Steam version right now. And maybe one day I will wake up and think "You know what I want to do today? I want to make some Mind Over Magnet levels." Or maybe I want to make the, perhaps, inevitable level editor for the game.

Reviews

So we've talked about the launch, we've talked about the numbers, we've talked about the money, we've talked about the bugs. Let's talk about the reviews.

Now Mind Over Magnet has actually had pretty darn good reviews. The game is sitting at about 88% recommended on Steam, with about 700 reviews, which is pretty awesome. Though I admit also probably slightly biased by my fanbase. Though also biased in the opposite way, thanks to the few GMTK haters who will never pass up an opportunity to tell me how much of a phoney I am. Love you guys.

The game has also had a few professional reviews in the press, including an 8 out of 10 from DualShockers, and an 8 out of 10 from Sports Illustrated, of all places. Those are very kind.

And on Backlogged, the game is sitting at 3.3 out of 5 stars, so somewhere between 6 and 7 out of 10, which I think is probably a bit more realistic. But all of those reviews are amazing. Really, really happy with them, to be honest.

In terms of positive sentiment, I think people just generally thought it was a fun video game to play, which is nice, and found that at least some of the puzzles produced that satisfying "aha" moment I've been chasing this entire development process.

People also praised the general feel, the polish, the presentation, the juice, the controls, the sound design, just how it feels to play as a video game. And of course, there was no end of praise for Zach's awesome soundtrack, which is available to purchase on both Bandcamp and on Steam.

Now in terms of the more negative reviews, the biggest issue was definitely the length of the game. Mind Over Magnet is about two hours long, and for some people that just left them wanting more, for others it left them wanting a refund. And a few people did take that opportunity, but nowhere near as many as I was worried about. Only about 2% of people have refunded the game on Steam, and of course some of those will be technical issues, others just didn't vibe with the game. So the number of people who actually played through the whole game and said "That's too short, I want my money back" is actually very, very small. So that's fine.

Now there's nothing wrong with a short game, of course. Some of the best games ever made are very short. But it's all about getting that price right. And price is something I went back and forth on a lot throughout this game's development. In fact, I was changing the price and tweaking the launch day discount within the last few hours before the game went live.

Ultimately I decided to just look at other games of a similar length and price the game in line with those.

Though I made sure to always give myself enough wiggle room for all of the inevitable sales, including the launch day discount, the various Steam sales they have throughout the year, and just about any other event where I can drop the price of the game. Burn's Night in Scotland? 10% off. The King's Birthday? 50% off. You know, we can just keep dropping that price. That's built in to the game's cost.

But whatever the case... money, value, these things are all very subjective and personal so I have no qualms about reviews criticising the game's length or price. Totally get it.

Another frequent criticism is that the game is too easy, simple, and straightforward. Now as you've probably seen in the previous episodes of this series, that is something I have struggled with and wrangled with throughout development. In the end, I am proud that the game is playable and enjoyable for a very general and casual audience of people. It's great that I can just give the game to anybody, and they'll probably be able to play at least the first couple worlds. And it has been great for people to tell me that their children have been enjoying the game. That is just really heartwarming and life-affirming, and so that does feel good.

However, I do regret that I didn't have time to put in content that was better suited to a more hardcore gaming audience. And while I certainly played with some of that stuff throughout development, it just didn't end up in the final game, and that does make me a little bit sad.

And then the other criticism I got was that the game is a bit bland, safe, uninspiring, unimaginative. Which... that one did kinda catch me by surprise. But the more I think about it, the more I can kinda see where those people are coming from. I mean, this game is not gonna blow your socks off. It is not the most innovative game of the year, or even of November. It is a pretty bog-standard, by the numbers, my first video game. And also as a solo developed project for the most part, it's not a very personal experience. You don't get a lot of me in this game.

So while I'm a little sad that people saw it in that way, I can at least understand where they're coming from.

And you know what, that is kind of the thing with all of these more negative reviews. That I agree with all of them, and I saw pretty much all of them coming. And to be honest, that was perhaps the hardest thing about releasing this game. Knowing full well that I wasn't 100% happy or confident with the game, but still putting it out there for people to buy, and play, and rate, and review.

It's like that bit in The Great British Bake Off where someone gives Paul Hollywood a raw cake, knowing full well that it's a raw cake, but still having to watch him eat it and complain about it. It just... the cringe is real, and that's how I felt for most of the first week of the game's launch.

Lessons Learned

But ultimately I came to the realisation that this was never the point of Developing. It was not about trying to make the best game ever made. (Though if I had done that by accident that would have also been fine.)

But no, this was about trying to find out what it takes to make a game. And it turns out it takes a lot.

Making this game, releasing this game, was hard. It was hard to make it look good. It was hard to make it feel good. It was hard to design all of the puzzles. It was hard to write all of the code and fix all of the bugs that I am aware of and have decided to fix.

As someone who typically works on a YouTube video for three weeks, it was hard to work on the same project and keep that motivation going for three years. As someone who typically likes to work alone and not show their stuff to anyone until it's absolutely fully done, it was hard to let others into the creative process and to show the game off at various stages of completion.

Basically, it is a miracle that this game even exists, and so the fact that it doesn't completely suck is just a nice bonus on top, if you ask me.

But okay, if the point of Mind Over Magnet was to learn about game development, what did I learn? Well, let's take the classic Game Developer post-mortem layout and look at what went right and what went wrong. Starting with what went right.

What Went Right

The first thing that comes to mind is the tools. This is basically anything I built that doesn't actually end up in the game, but helped me to make the game more quickly.

I'm specifically thinking about all of the mechanics in the game, like the drill and the block and the electromagnet. I can just drag them into a scene in Unity, they snap to a grid, and then I've got sliders and checkboxes to edit their statistics. By the end of development, it was basically effortless to put together or to change or to iterate on one of the levels in this game. And so while this stuff certainly took time to produce, it saved so much time down the line and I don't regret any of the time I spent building these tools.

Another good use of my time, I think, was working on the juice, the presentation, the polish, the game feel. I don't think it's too big-headed of me to say that this game feels pretty good to play, and it's got nice animations and the sound design is good, and it all just comes together and feels like a polished, real video game. I think these elements really tie the game together and make it feel better than it actually is, if that makes sense. So again, I do not regret the time I spent working on particle effects and transition shaders and sound effects and stuff like that.

Another big win was playtesting. Now I'll admit it took me a long time before I was able to show the game to anybody, but once I did, I didn't really stop. Over the course of several years, I constantly and consistently showed the game to friends and family members and strangers on Twitter and people at events and meetups. And basically no aspect of the game was not improved by listening to feedback from other people.

Mind Over Magnet would simply be significantly worse if I just ignored what everyone thought of the game. That should be obvious, but bears repeating.

I'm also pretty happy with accessibility in the game. Of course there's more I could do, but the game does have good colourblind support. It has full controller and keyboard remapping. You can play the whole game with a keyboard and mouse or with a controller and they both feel good. And it's got that simple backgrounds option where you can swap out the detailed and distracting visual backgrounds with a simplistic grid pattern.

The real smart move was getting this stuff into the game at the right point. Because while I think a lot of game devs get to the very end of the process and then think, we should probably put some accessibility options in here and then have to retrofit those ideas into a spaghetti nightmare mess of finished game code, I put it in as soon as those features were first being developed so it was always there, baked into the code from basically day one.

I'm also really happy with the hint system in this game. I've seen so many playtesters and other players get a hint for a puzzle and then kind of go, "Oh wait wait wait, I get it now. I know what to do." And kind of still feel that thrill of satisfaction and eureka moment despite the fact that I just kind of told you what to do.

And I think that's because the hints are actually hints. They just kind of push you in the right direction without ever telling you exactly what to do. Okay, some of the hints do tell you exactly what to do, but most of them don't. And if you are still stuck, you can just skip any level in the game as long as you've been stuck on it for about three or four minutes. So the game never has a total brick wall and there's no reason to look up a guide, leave the game, or stop playing.

What Went Wrong

Now in terms of things that went wrong, I think the biggest one was rushing too fast into production.

You see, game dev is split into two aspects. You've got pre-production where you're figuring out what the game is going to be and running all sorts of experimentations and explorations with prototypes and art tests.

You're basically deciding all of the major answers to the questions you have about your game. You know, what is this game? What does the player do? What does a level look like in this game? What is the overarching structure? What's the core loop? What does the artwork look like? So on and so forth.

And then you transition into production where you actually make the thing.

Now the nice thing about pre-production is that you can be really fast and nimble about making changes because they're easy and fast and cheap. You can try ten different solutions to a design problem and it doesn't matter because you're just going to throw everything away anyway. You're not committed to anything.

But if you do what I did and move too quickly into actually producing the game before deciding the answers to those most fundamental questions, you're going to end up repeating yourself, spinning your wheels, having to throw away lots of work. Basically, what I'm trying to say is you shouldn't have to change all of the art assets in your game within the last few months of development because you hadn't previously figured out how far the camera is going to zoom out. That is, and I believe this is a technical term, a dumb dumb decision.

Another thing that went wrong, and this is after actually entering production, is that I never really had a good plan for how to make the game. No roadmap, no milestones, nothing like that. I had a vague idea of what the game looked like in my head but I just ultimately kept working on the game with the vague hope that it might, by itself, coalesce into a final finished video game.

Spoiler alert, it didn't.

That is, until I finally sat down with a notepad or with a Trello board and planned out what is this game going to be like. How many worlds does this game have? How many levels per world? What are the main story beats? Stuff like that.

It was only when I put together that plan that I could actually start moving in a good direction towards the finish line.

I also focused a lot of time on stuff that really didn't matter. Like okay, in Mind Over Magnet, if you put the mouse cursor over one of the options in the menu and then switch to a controller, it will continue moving from where that highlight was. That took like a day to figure out and is not that important in the grand scheme of things, and I wish I had spent time that I spent on doing stuff like that on improving the underlying gameplay or making better puzzles or making the game feel good. Stuff that actually does have a real impact on what people think of the game.

Another thing that went wrong is that I picked a surprisingly difficult genre for my very first proper game. Now a 2D puzzle platformer might seem quite simple, but it's actually deceptively challenging. A 2D platformer is already pretty hard, because making that character feel good to control is an engineering feat of animation and programming and other skills.

But it's the puzzle design part that was really difficult. It's just hard to make puzzles. Puzzles are like a craft in of themselves, and I started to realise that some of the world's greatest puzzle game designers like Jonathan Blow and Alan Hazelden and Stephen Lavelle have spent years, like their whole careers, just learning the craft of puzzle design. And then I come in and I'm like, "Yeah, I could probably do that in a couple months while also working on every other aspect of the game. No problem."

And so it shouldn't be a surprise that the puzzles in the game took a long time to come together, and I'm still not really that happy with basically all of the puzzles in the game.

Oh, and on that note, it was probably a dumb idea to also have a heavy physics element in a puzzle game. Puzzles are very much about the designer having real control over every aspect of the game, and then when you throw in the chaotic nature of the universe into your game, suddenly there are a lot more problems to worry about.

Working Solo

And then there's the thing that I'm not going to put under what went right or what went wrong because I'm not quite sure how I feel about it. And that's the whole concept of working solo.

I made pretty much all of Mind Over Magnet by myself, but that meant that I had to stretch a pretty limited skillset, let's be honest, across a really wide spectrum of different disciplines. I had to spread myself across animation, sound design, art, writing, game design, level design, production, marketing, and probably a hundred other jobs I've forgotten about.

And then if you compare that to Zach, who made the soundtrack, all he had to focus on was the music. Putting aside the fact that he's also making his own video game, but that's neither here nor there. In the context of Mind Over Magnet, Zach could focus a hundred percent of his energy and efforts on just making the music.

And so I was left wondering, like, what would Mind Over Magnet be like if I could just spend three years exclusively working on puzzles or game design and let somebody else focus on programming or artwork or writing?

The game would definitely be better. It would almost certainly have come out more quickly. But I'm not sure if I would have enjoyed the experience as much.

I like working by myself. I like having full creative control over the process. I like being able to work on the game whenever I want, and I wouldn't want to have to split those Monopoly monies any further. Still, if I do ever make another game, I will think about which aspects of the game I can delegate.

Future Games?

But will I make another game?

That is a question I've been thinking about a lot lately, and it's the question that people keep asking me. "Mark, are you going to make another video game?"

Now, I am definitely going to enter Game Jams. I am certainly going to make more interactive video essays like Platformer Toolkit. But the thought of making another full, big, multi-year Steam release...

I'm not sure I've got that in me. I'm not sure if I'm ever going to do that.

But if Mind Over Magnet is the only game I ever make, I'm okay with that. I would be okay with that, because despite its flaws and its shortcomings, despite the missed opportunities I had with it, despite all of the problems I can see, despite the bugs and all of the stuff I wish I'd done differently, I'm proud of it. It is a cool video game.

And when I look at the game, I don't see the flaws anymore. I see a video game, a real game that went from idea to release, and has now been played by thousands of people, most of which did not refund the game, and the game has made significant amounts of money for me, which is, you know, part of why we do this as game devs, right?

And it is now the objective truth of the universe that for the rest of time, I made a video game, and that cannot be taken away from me.

But here's the thing. I could not have done it without you. So I want to say a massive thank you to everyone who has followed this series, has watched me slowly lose my mind, and have supported the game.

I want to thank you if you were there in 2021 when I announced my intention to make a video game, if you were there in 2022 when I released the first really bad minimum viable product for the game and started getting feedback from you for the first time, if you were there in 2023 when I took the game to the Game Developers Conference in San Francisco and got real encouragement and a real motivation to get this game done, or if you were there in 2024 when I hit that big green button and released the game on Steam.

And as for 2025, I do not know, but I'm excited to find out along with all of you. Thank you so much for reading.