What A $1,000,000 Game Looks Like In Unity (w/ Tom Francis)

I know what you're thinking. Trust me, I feel this exact way right now. You feel like you're making your Unity game wrong. You feel like the folder structure is a mess. You feel like the assets are thrown together in a scene. And you feel like maybe it's a house of cards just waiting to come crashing down. Tom Francis, the creator of Gunpoint, Heat Signature, and his latest hit, Tactical Breach Wizards, was generous enough to show me what Tactical Breach Wizards looks like behind the curtain inside of Unity. At first, I was insanely impressed with some crazy custom editor tools that he and his team had created. And by the way, these are incredible. You're going to learn a ton just looking at these editor tools. But I was also encouraged. See, his game, although breathtaking in play mode, it looks well like a normal Unity project like yours or mine when all the post-processing and the lighting and the camera effects are removed. It's encouraging and also fascinating just to see a wildly successful project in the

editor. And I just know you're going to be encouraged and learn a ton from this clip from the full length podcast. So stick around and learn from someone who is currently in the trenches making a full-time living making indie games. And by the way, if you want to make game art like you see in Tom's Games or my 2D or 3D games, I can show you exactly how to do it. So, I wanted to give you a chance to check out my 2D art course, which dives into everything I've learned as a Time magazine featured game artist. For the next several days, it's 25% off. Plus, you're going to get my 3D art course completely free, but only for this sale. There's just 50 seats available, and then the sale will be closing. If you enroll today, you'll not only get my 2D and 3D art courses, but hey, you'll also get my custom color templates, my brushes, and I'll also throw in my pixel art course completely free. Join hundreds of students on their game art journey and learn from an actual published indie gamedev featured by Rolling Stone, Time Magazine, and PC Gamer. That sounds really cocky to say, but I don't know. It gives me some credibility. So, click the link below before the sale expires or the seats

sell out. Okay, so this is Tactical Breach Wizards, right?
Mhm.
Let's go. I don't even Wow, what am I looking at? So this is a um sort of custom inspector for uh our stage list which kind of holds the whole structure of the game. So pretty early on we really realized we needed a different kind of thing to a scene. Um you know our game is has Unity scenes and then it has data on disk and a stage is basically a pairing of those two things. So this is a gameplay stage. This is the intro mission of the game. And in the configuration of it, there's um the the mission here is determining where it's going to find the information about the the actual level layout. And then the scene here is is what you'll actually see. You know, this is a piece of art basically that is uh pre-made. Um and
you know, throughout the game we have we reuse background scenes a lot. Um but obviously each mission is unique levels,

so we're loading different data into the same Unity scene. Um,
can I have you explain this one more time?
Sure.
Honestly, I I'm sitting here. I'm like, is it a It's like a It's a custom scene manager kind of thing.
Yeah. Um, so it's the stage list and um each entry here is a stage. Um, and
what do you what do you mean by stage?
Uh, they are a custom thing we created that pairs a scene to some data. So, in the case of this title card, the first title card we have, um, uh, the scene is incredibly simple. It's just a black screen basically. Um, and then the data tells it what, uh, what title to show. So, it will be like a prefab. Um, if I edit it, it's going to link to this prefab. So, it's just this is really the important data. It's just a scene and there's text prefab and it's going to load this bit of art into the scene that we're that already exists. Um, and for title cards, it doesn't make a lot of sense. We could have just made individual scenes, but for gameplay, it

makes a lot of sense because we have a ton of missions with all the same scene, but different level data.
So, this is just a a scriptable object.
Um,
yeah. Stage list scriptable object. Gotcha. Gotcha.
Yeah. Each stage is a is a scriptable object as well.
How long did it take you guys to build this custom scriptable object? I'm sure you can't remember, but maybe like a rough estimate.
Yeah. So, this is all um our programmer John Windows department. Um, I do some programming, but he's our sort of lead programmer. And this was something he kind of built as we went along. So, the first version of it was very basic. I think it was just for gameplay. So, it just paired a gameplay scene to the the name of the folder that will have all the level files on disk. Um, and then it became really useful as it went along, as the game became more of a fusion of of, you know, dialogue scenes and gameplay scenes and title scenes and, uh, mission end screens. Um, to kind of make it more flexible and, uh, give it, you know, allow us to make scenes that were not just level gameplay and and, uh, covered all the kinds of things that

we have in the game. And so, it's divided into these tabs. There's four tabs here because the game has four acts. Um, that's just because it it gets unwieldy to to view it all at once. Um, but it's really nice for things like we can just drag like a dialogue scene around and just move it to where it'll appear in the game. So, this is kind of sequence of things in the game uh will happen in this order.
Are you serious? I didn't realize it was the actual sequence. I thought it was just a way to look at your scenes basically. But
yeah, it's also how we build it. So, we can add Oh, I don't want to add a new act actually because that would be pretty major. But, I can add a new scene and then you just choose what type of scene you're adding.
Dude, this is so interesting. I have never seen something like this before.
Yeah, we don't really
Is this normal? Do you think this is in the game industry?
I mean, you've probably talked to more developers about that process than I have.
I I have never seen that before. Um, but there is something to be said about obviously persistent game objects from scene to scene and going from scene to scene and ordering scenes and triggering

new scenes. That's super. It's like a spiderweb. Um, and it's very messy and very complicated. There's no Unity solution that's built in currently
that helps you create that sort of hierarchy or path from scene to scene to scene. Um, so I think it's it's brilliant. I wonder I wonder if there's a tool though that's in the Unity asset store that allows you to do that. I don't know.
Yeah. So, uh, to add a new scene, add a new stage, uh, you can just click here and choose a type. I can make a new dialog scene or um uh title card. Perks are like when you actually spend your upgrade points and we can put those wherever we like. Um and if it's a gameplay scene, then when I add it, I've got to edit it. I uh click override here to choose the scene. And this is a list of all our scenes in the game. Dialog scenes are uh obviously cut scenes with people talking, gameplay scenes are backgrounds. Um and I can just choose whatever scene I like. So, this means you could also load, you know, a level into the wrong scene. Like any level in

the game could could appear in any tile set. And I'll show you in a minute like how our level editor kind of fills the level with art because that's all automated as well.
Yeah.
Um, and then uh that would be a new scene and uh it'll show an error for a second because it doesn't update very promptly, but I can then put that wherever I like and um uh move the campaign around. So, reordering things is really nice. This is really good for perks. Um because perks are where you get the upgrades. Um and we don't there is no strict hard and fast rule for when those happen in the game. We just decide when it's time to give you some upgrades. Um and we've had to tweak that a lot, you know, especially as we add new characters because the structure of the game is you start with two wizards and then eventually you get five. Um but we add those one at a time and all that time those first two keep leveling up. So, we had a problem at one point where those first two were so powerful by the end of the game that they were just uh absurdly good. So, we had to figure out how to like balance those.
Yeah. Well, that's so funny. You I was just about to mention the word balance

because one of the best ways that I found to balance my game is to have everything in scriptable objects. Um, it's really hard to balance a game when you have to go into the scene and dig around and find like there's too much ammo in this crate or this enemy in this prefab need I need to adjust their health so I'm going to go into the actual enemy prefab and adjust the health or adjust their attack power.
Yeah.
So, it's so much easier to have it all in one g uh like manager effectively
and do the balancing in the manager.
And again, this is all built by our programmer John Winda and it's it's really smart. It's stuff I would never build if I didn't have a programmer because on our first game I did all the programming myself and it was very much like anything like this would just be like oh that would be nice to have but I can't build it.
Yeah.
You know what's really striking about this project Tom is this is this might sound offensive. This is like neurotic how how clean this is. I'm looking at the syntax of everything. The only thing that strikes me as as like outside of your boundaries is that

folder called Tom's experiments. But everything everything else has like really strict like s syntax and naming conventions. Um I am like and even like your scenes the way your scenes were named the conventions are very uh precise. Um can you talk a little bit about that? Are you in charge of that? Are you responsible for that sort of
um I'd say that's a a team effort. Um, I'm uh I did sometimes do like a sort of organizational overhaul if it was bothering me about where things were and uh difficulty. It's nice to hear that because to us by the end of the project, it feels like a nightmare.
It always does. It's a it's a it's a house of cards, you know, even if you've got like divine gamedev abilities, it's always going to feel like a house of cards, you know. But look at that. Wow. The naming conventions are so impressive. I'm just so kind of jealous of your ability to keep things clean cuz I I I I just inevitably can't. At one point it just kind of goes to hell for me in my projects. But this is awesome.

Let me um
let me ask you a question. So your forte as a game designer or I would say a game developer is game design, right? That's your big thing is design.
Okay. You just won an award at IGF by the way. Uh what was that? Excellence in design. Was that it? Yep.
Okay. So, question because dude, I'm not a game designer. So, I'm I'm a like a story guy. I'm an atmosphere guy. I'm an art guy. Like, that's and music. Like, that's my thing. In a way, I kind of approach making games like I do like making a film. Um, which is not the most profitable way to do things, but that's just my skill sets. You know, that's my talent. My question to you is, as somebody who is not a skilled game designer, my question to you is, is this level of organization crucial to a game with such intricate design?
Um, I don't know if it's crucial. Like,

we haven't made a game like this before. Like, this is our first game was 3 hours long and and all story driven and um kind of linear. Our second game was completely procedural and kind of as long as you want it to be, but maybe 20 to 100 hours depending on sort of how how much you rinse it. And this one is like 15 hours long and it's a story campaign. And we had to that was a real learning experience cuz I thought, well, we've done this before. We've done a we've done a story campaign. We know how to make a just put things one after the other and and decide what the players experience is going to be to a large extent. But we'd never made one 15 hours long before. So, uh that's kind of why it's in these four acts. like each of these acts is about what the size of our original of our first game, Gunpoint. Um, and once you've made that much game, you kind of have to reinvent it for the next bit. So, that's when organization becomes quite uh critical and yeah, just managing the players uh power level over the course of a long campaign like this. You know, we're not we're not a true RPG. You're not earning experience by doing extra side quests or anything. Um, we are kind of a puzzle game at times.

It's a kind of puzzle tactics game. So, there are
Yeah, exactly. There are times in the campaign where we need to make sure you can do this. Like the player has to be able to do this. Otherwise, they're just stuck. They can't do anything else. The game would just end. So, we had to be very careful about knowing what the player's abilities were.
Cool. Let's go ahead and uh I want to click I can we click a play stage button like I kind of want to see what the art looks like in a scene. Is that possible?
Yeah, for sure. Uh let me do the first like one in the Rushwater Police Department. This is um kind of when the game starts to get going is that there's a prologue intro mission that's buried on rails. Um and then this is still tutorially we're introducing a new wizard
but um
uh at the same time we've got to oh sorry because I've changed this the
I know
I added like a dummy stage
man
I was demonstrating that you could move it anywhere and now I can't remember where I put it.
Yeah.
Well, I'm looking at the Unity version. We're in 2019, so it's almost like we're in a little time capsule here trying to

open up a project from effectively 2019.
Yeah, I I haven't uh fired up from the editor in quite some time. This is not sort of like a the game can't run. It's like a warning. Um because uh we needed to delineate kind of, you know, this is the gameplay mission, but actually everything that comes before that mission, you know, the dialogue before it and the location card before it are kind of part of it. And when you go back and replay missions, we need to know like where do we start you from and when have you finished? So we have a system for that. And I I screwed it up by adding the extra mission in there.
So let's see if this works now. Yeah. So this is in game. This is we're in a gameplay stage now. But there's a little script file to to tell it to play some dialogue at the start.
Um this dialogue is up. If you don't mind, let's open up it in scene view and sort if we can look around this scene and see it, you know, and compare it to the game view. I love seeing stuff like that.
Yeah. Uh, how do I get it to actually look at the scene we're in?

I don't know. That's a good question.
Uh, let's see.
Okay. Good, good, good.
It's just wild to me how and I hope you don't take this the wrong way because I and I said the same thing to the creator of Inscription, how unimpressive a scene looks in scene view, you know. Yeah, this is I haven't really looked at this much. Um, especially this scene. So, this is a a real credit to our artist John Roberts, uh, who did all of the art. Um, but especially you can see how much the pro processing is doing here.
Exactly. Yeah.
I was going to try and zoom in, but I can't zoom in during a a dialogue scene.
So, we've got particle emitters. We've got 3D models that to me, they don't look like assets. I mean, are are they custom?
Um, do you mean did we make them models?
Yeah. Yeah.
Yeah. Yeah. Yeah, it's all our own stuff.
So, would you say that this project as a whole is pretty much fully custom? There's no assets being used.
Uh, we do use some assets. Uh, I can

probably just see in the project actually there's I remember for like cloth physics we use a plugin. Um, for some some ropes and cables, uh, we use a plugin for that. Um and we also have a lot of like behind the scenes like uh integration um with things like Trello for like bug reporting.
Sure.
Um we use something called flat kit which is part of our rendering style which I believe is a is a plugin and um MK glow freeze like a post-processing suite. John would know more about exactly what we've used. Uh so that might be an incomplete list but yeah we we did definitely use some third
these shaders these shaders that give this tune shading. You're saying that's a um a flat hit shader.
Yeah, I think that's that's something that John um sort of you know took took a baseline and then customized from there. Um
because it is quite a specific art style u especially like
uh John and I happen to to agree on on rendering faces in the sense that we

like it a bit stylized like like to just sort of suggest yeah we just have the eyes and no other facial features. um cuz as soon as you do that facial features, the character has an expression uh and quite a specific expression and then 90% of the time that's the wrong expression for what's going on.
Whereas if you blank I think that the player can kind of fill it in.
Yeah. I mean, there's a lot that you can get away with as well with these isometric games, right? Um you don't have to show their faces and you don't have to show close-ups of most things, right? So, helps to helps to have that isometric camera. Oh, cool. I love that.
Oh, wow. The backgrounds are 3D.
Yeah. Yeah. So, when I'm assigning a scene file to a level, that's this is the scene file really is all this this backdrop stuff.
I see. Can you give me a the thought process of why you wouldn't just use a flat, you know, sprite like take a screenshot of it and just put it underneath?
The player can move the camera in game. So yeah like this. So
uh we could have done completely you

know blank backgrounds just sort of you know a kind of abstract thing.
Um
but uh I yeah it took us a while to kind of decide on on the exact level of fidelity and and convincing us because obviously uh that would be the abstract end of the spectrum. The more real end of the spectrum would be to draw this building under you like you know the lower floors we just don't draw them. Um, and these other rooms that you're going to go to. Um, uh, we played around a bit with how much of those we show. We want to give you a sense of the building, the layout that you are going in this direction. Um, but at the same time, if there's detail in there, then that's distracting. Um, and I especially wanted to make sure the player knows you don't have to worry about anything that's not this room, you know. Um, the next room will be a little bit bigger than this and actually not that much bigger. Um, but uh, let me actually skip ahead a bit and get a more representative level. This is this is linear, right?
Yeah.
Um,
yeah. You don't choose what level to do.
Okay. And no procedural generation either.

No, we we toyed around with that for a while. Like we built a sort of prototype procedure generator, but it was going to be
um a lot of work to make it reliably fun. And we didn't really need it. Like actually part of the reason we didn't do it is um we brought on a level designer, Steve Lee,
um to help out with level design. And so he did a chunk of the campaign, but especially he did all of the optional extra missions. And having those be handdesigned is um you know, it was a lot better than doing procgen stuff. And another reason for proc not to do procgen is that this game we're quite careful about ensuring you have a good story reason to be, you know, breaking into buildings and taking out the people inside because this is a pretty hostile action you're doing. We wanted to make sure that it's always against a, you know, a powerful opponent who is is, you know, oppressing uh somebody
justified.
Yeah. And so Progend doesn't really it solves a problem that that we're not going to have, which is we're not going to want you to do a ton of gameplay without any story. Um because all of

that is is difficult to justify. That's why the optional missions we do have um they don't have story attached to them. Um, but uh they also technically take place in a dream. It's kind of like a wizard hallucination dream where you sort of like uh it's to do with uh Xan's powers. Um uh and it's a total fudge and we kind of make fun of it in the game, but um it enables us to just have a ton of missions without having to write a story for each one of them.
Exactly. Exactly.
Uh the way we make levels um is not in Unity itself, but in our own level editor. Um,
and uh, this is total overkill by our artist John. He made his own custom tile set just for the level editor. Kind of looks like a sort of um, you know, a VR room or um, uh, yeah, very nice and clear. You know, it shows the orientation of of each thing. So,
wait, this is real time. Like you How did you get to this?
Uh, I just hit L. This is I'm in development mode, so you can
So impressive. Wow.
This level editor is in the game. So,

uh, players can mess around with it. Um, and when you want to, uh, edit the room, you can just, uh, I can add some extra cover here, uh, like this.
And then jump straight back into the game and the art will just sort of figure it all out.
That is amazing, Tom.
So, super clutter.
Um, and yeah, making new geometry is is pretty easy, too. You can just, if this level should be bigger, we just add floor. We don't have to add the walls cuz a certain point it's just obvious where the wall should go. Like obviously you want walls separating any um any bit of the outside with the inside. Um I can add some uh windows along here. Can add some more enemies uh if I remember where to find everything. It's been a long time since I made a level.
Now do you have to use the level editor as a as the does the team have to use it? Yeah, we use this for everything. Um,

we uh for our first game, Gunpoint, I built all the levels inside Game Maker, which is what I was using for that game.
And then at the end, I didn't know if I was going to have a level editor.
Um, so I didn't build it first in case we didn't have time. And at the end, we did have time. So, I built the level editor at the end.
And after I'd made it, I was like, this is so much nicer to use than what I've been doing for 3 years. So, I'm like, next time I'm building level editor at the start, and then I get to use it. Would you recommend someone like me, like I make I like making like story driven first-person shooter games, kind of like Bioshock type games. Would you recommend a level editor for those kind of games?
Um, yeah. I mean, I've never built those those kinds of levels. Well, I guess I did in my like modding days in school. Um, where like we messed around a little bit with making like Quake levels and things.
Um, but it's been so long that I don't know what modern FPS level design is like. With our game, it's so the bones are so simple. Like we wanted it to be big and chunky and like there's no Yeah. You know, awkwardly shaped objects. Everything just takes up the full square

meter or nothing.
So the building of the the logic is very simple. And I do remember seeing a um uh a video about Bioshock um that talked about how much of a pain in the ass it was in that game where they built and and arted a level and then they decided like for gameplay reasons a flow needed to change and you got to just move this whole Yeah.
like change this whole corridor and all of the art and clutter you've added has to be basically deleted and then recreated after you do it.
We went through that process, dude. I I think we went through three different level iterations for 12 different levels.
So that's more than 30 times a level was built. And what was crazy is me and my team member Felipe, we we have this problem. We're very visual people. So we have to see a level set dressed to know if it's working. And that is such a pain because like we would try in gray box and we just couldn't

we couldn't fully realize the level and even like some even some like like game design stuff we didn't even know if the game design was good with just gray boxing. We'd have to
address it to realize if it was actually fun.
Um and I knew that was a problem because it's such an expensive way to make games. So, I'm curious with this, you know, like um are is your mentality that gray boxing and just very simple forms should be enough for you to realize whether the game is actually fun?
Yeah, I mean, we never play a level in this form. Um, we had we had at least a some kind of, you know, uh quality art tile set from the start. Sometimes it wasn't the one that we'd be using eventually for for this mission. Um, but we we're always playing with like shipping quality assets. Um, John is is incredible at uh everything he makes is ready to ship basically. We never have anything kind of rough and ready. We definitely have times when like we don't have all the elements that we would want

to put into the final game, but everything he makes looks great from the start. So,
um, uh, yeah, what I was trying to show here is that you can switch these levels between tile sets um, uh, pretty simply. So,
wow. I'll actually put this back in the police style sector so you can see the normal version of this level. So this is it normally. Um
yeah,
and then if I Oh, it's crashed. Let me Okay, I can just get past that. Um and then yeah, I was just messing around with the black site tile set. Uh black site ops. I have to change the mission the scene as well. So although you can set the scene in in the stage editor, you can also do it from within the um uh level editor and it just has to kind of reload. But then you got the same level in a different tile set. And it's crazy how different it feels, you know.
Oh, it changes everything. The background's changed.
It feels more spacious this level, even though it's the same geometrically. Uh

just the kind of light and airy, deserty feel. You can see the the background scene has all these, you know, sandbags and um military stuff. This is a kind of military installation.
Um, but I can also take us to uh let's do Kan uh Kan Street probably won't work too well because that's designed specifically for an outdoor environment which is uh trickier to you have to bear that in mind when you design the level. But for every other tile set in the game, every tile set works basically. So we're now we're going to go to Kan with the uh the chapel who are kind of religious institution.
Whoa, cool. So again, this is gameplaywise the exact same level, but it just looks and feels completely different,
man. Well, dude, thank you so much for hanging out. This was a lot of fun. I learned a ton. Hey, what do you want in the description uh of this podcast? Anything? Any link? Uh
oh. Um yeah. Uh actually pointing people to uh like our Steam profile page um and maybe our mailing list. That's I mentioned that that's where we get

testers from. So, if folks want to test our next thing, that's a a good place to go.
We'll do. If you want to test Tom's next game, the link is in the description. Thank you so much, man. That was a lot of fun.
Thanks a lot.
Yep. Cheers. And by the way, if you want to make game art like you see in Tom's games or my 2D or 3D games, I could show you exactly how to do it. So, I wanted to give you a chance to check out my 2D art course, which dives into everything I've learned as a Time magazine featured game artist. For the next several days, it's 25% off. Plus, you're going to get my 3D art course completely free, but only for this sale. There's just 50 seats available, and then the sale will be closing. If you enroll today, you'll not only get my 2D and 3D art courses, but hey, you'll also get my custom color templates, my brushes, and I'll also throw in my pixel art course completely free. Join hundreds of students on their game art journey and learn from an actual published indie game dev featured by Rolling Stone, Time Magazine, and PC Gamer. That sounds really cocky to say, but I don't know. It gives me some

credibility.