Game Design: You're Doing it Wrong.

News & General Discussion
1

If you’re not prototyping your games and playtesting them, you’re not designing games. You’re only thinking about designing games. That thinking tends to be wishful thinking. Imagining, if you will, that you have the design for a great game in your mind. So, you’re an imaginary game designer. That could mean a designer of imaginary games or it could mean someone who only imagines that they are a game designer. It makes no difference, the result is the same with the former as the latter: none of these great games are ever produced in reality.

Quick. Name three good game designers. What do they all have in common? Maybe they have a lot of knowledge about game mechanics, flow, presentation? But on a more fundamental level… what do they all have in common? Hmm? That’s right, they produce games.

I have been on the search for the Holy Grail of Game Development Wisdom for quite a while and it’s no secret that I haven’t found it. It’s no surprise why, either. It doesn’t exist. I have heard things said in a lot of different ways by a lot of different people, but I am convinced that the people who are successful don’t even know how to explain what it is they do that brings them their success, because it is so fundamental to their personality and the way that they see the world that they can’t possibly have any self-awareness of it.

Look at the two quotes I started this post with… think about what they are saying in a very casual way…

This is a point of instruction. If you program something, that doesn’t mean that it is going to be a part of your final game. You cannot be afraid to throw work out if it isn’t working. You cannot be a staunch defender of your work after coding up the first working version of your idea. It doesn’t make sense. It is actually counter productive. You aren’t learning anything this way. In fact, you’re hampering progress by insisting on keeping everything you write in the final game. The more you invest into this thing you are making, the more difficult it will be to scrap or steer in a new direction.

I spend a lot of time on here and on YouTube, reading threads, comments, articles… checking out people’s games and offering feedback… and I can say that I know you pretty well, aspiring game developer. You’re not likely to take any of this to heart because you know what’s best. At least, you think you do. But two years from now, when you haven’t figured it out and you’re thinking about giving up, hopefully you’ll remember stumbling upon this and you’ll go searching for it again. And if you do, then hopefully you’ll read it again.

Here’s a simple process that anyone can do:

  1. Build a basic version of your game’s core mechanics.
  2. Play it and Let other people play it.
  3. Learn what works and what doesn’t.
  4. Scrap it.
  5. Rebuild it, expanding what works and removing what doesn’t.
  6. Add more ideas.
  7. Return to step 2.

Repeat this process until your until your crappy prototype is fun. Then, and only then, should you invest any serious amount of resources into your idea.

That’s it. No cherry on top. This is the only thing that really works. You have been warned.

10 Likes
2

Does this mean you’re ready for us to try out your prototype? :slight_smile:

I agree with what you posted. Well on the prototyping and removing things that just aren’t working. I am not a big believer on releasing prototypes for others to check out very early on. I think that is actually a bad thing. But the prototyping trying things out in general. Absolutely yes. A lot of times just getting something up n running makes it painfully clear it just isn’t working or could be made even better.

It’s an interesting thing because each person kind of has their own development style. I get the impression that many successful Indies work on their games for a long time before actually releasing anything playable requesting feedback. In fact, sometimes I am not sure they even have done that and instead relied on videos and such.

For example, I got the impression that both the developer of Stardew Valley and Axiom Verge [those are links to articles / dev logs not the actual games] worked on their game for years relying only on their own goals, testing and judgement.

More so with Axiom Verge because the Stardew Valley creator blogged about the development over the years he made the game. Writing about what he added or changed and showing screenshots and videos. And people commented giving their feedback based on the screenshots, videos and features he wrote about. From what I can see though there was no demo or other way for people to actually play-test.

I think this approach makes a lot more sense overall. Because it is also building a fanbase.

Of course, it is possible both of these folks had a close friend of two check it out from time to time.

2 Likes
3

+1. Because “the value of prototyping cannot be overstated”.

At my new job, they do yearly game jams - artists, engineers, and designers spend a week working on throwaway prototypes. Last Monday, one of the prototypes was polished enough to go live - 52 Card Pickup. Big, medium, and small, we all prototype.

Gigi

3 Likes
4

Thing is there are plenty of games that don’t prototype for shit. For a game built off of a core gameplay loop, testing early and testing often just makes perfect sense. Fundamentally though, these are games with only one way to play, so not carefully designing the gameplay is not caring about the game.

There are, however, games that give the player enough agency that there is no longer a definite core to gameplay. You can’t reduce Deus Ex or Stardew Valley to ten percent of their total feature set and say you are getting their quintessential experience. The aesthetics these games try to evoke just is not prototypable in the same way. It’s only if you had a significant amount of the game in place (say 60%) before you could begin to accurately judge the value of any given aspect.

These are frequently games reliant on depth, on systems that feed into systems the feed into subsystems that feed back into the original systems and so on and so forth. Any one system is rarely that valuable, but as a whole they create oceans of possibilities which couldn’t be done otherwise.

7 Likes
5

Master Frog will never post his prototype and if he does he will do it tactfully with lots of ‘remember folks this is just a prototype, so you can’t critique it.’ But there is always a chance I could be wrong, unlikely but still like to give the OP the benefit of the doubt.

All being said, he’s right on how you should approach game prototyping… but at the end of the day you still want to end up with reasonably goodish graphics if you wanna attract attention, so at some point you need to start iterating that phase.

6

I think he will put out a prototype when he is ready. Prototyping in general is of course a good thing. Should be common sense really. I see it as just part of development. When I implement something I then test it for bugs, test how it “feels”, how it plays. And then tweak and test and tweak and test a bit more. I guess to me this is software development. I mean it’s just the natural way to do it. Do the minimal first, a proof of concept, maybe even 3 to 4 different ones, then move forward with the best (which fairly often comes about as a results of testing all of the others and maybe combining this and that).

Is there anyone who actually does not do this? I mean someone who literally comes up with an idea of a game or whatever “on paper”, makes it and then releases it without ever actually playing it and making some tweaks here and there? :hushed:

I do think there is little sense in spending hours, days, weeks and months just on generating the best graphics you can up front. It makes sense to knock out the simplest graphics possible. Then focus on implementing the systems and all.

BUT… at the same time we see people like the fella who made Axiom Verge who spent the first ~9 months on nothing but creating graphics.

Really, I think each person just needs to find what works best for them. AV certainly could have been prototyped with simple colored rectangles. But I think maybe he used a different approach… he knew what he wanted to make and he literally “just did it”. He spent 5 years just building that game before showing it at the Expo or whatever it was and getting funding.

1 Like
7

Great read, I’ve been doing that, and hopefully it’ll lead somewhere.

Can you help me out with tips on how to reach out to people for testing my game? I have good interest from friends, family and a few strangers on internet. So I know there’s an audience. I"m just not able to reach out to them.

8

I’ve really struggled with these points and the process of play testing.

I do think that there is value in having people play an initial prototype, or even test out some of your ‘horizontal slices’. But it’s not worth it if you’re looking for really meaningful feedback. This should be treated as a sanity check only.

You should not expect the play testers to really give you meaningful feedback or meaningful guidance. You can get input on your GUI, or if things “generally make sense” or maybe on the controls, but you will get zero insight into if the game will be good or not, or even if it’s on the right track.

That kind of input can be useful. But I think it’s extremely different from the kind of “most important thing you can do” - “playtest early and often”, kind of advice you hear quite often.

This kind of thing is not limited to game systems and subsystems feeding into each other, some games live and die on their thematic elements, on their atmosphere or narrative. Obviously the more important these kinds of elements are, the less value feedback will be before they’re largely in place. The amount of content that needs to be fleshed out will vary.

In my experience, I’m extremely glad that I prototyped this iteration before I started working on it. I’m glad that I gave it to people to play. But I ran play testing sessions every month after, and most of these pre-alpha tests were probably not worth doing, and never really gave me answers or assurance on the only question I really cared about: do you like playing it?

My current schedule puts me in early access release in 4 months. The last 2 months are dedicated to play testing and revising based on feedback. But I have stopped seeking additional playtesting until almost all the systems and a representative chunk of the content is in place.

NOTE: I really think that the emphasis on prototyping and play testing is not necessarily good advice on all projects and should really be tempered. I think that expecting too much from play tests and not understanding the limitations really hurt me at various points in my process.

3 Likes
9

I think you hit the heart of the matter. I made the mistake a few times of throwing out a very early WIP for feedback and the biggest issue was that people couldn’t really give any kind of deep or meaningful feedback on whether the game would be interesting, fun, etc. The only thing other people can give feedback on is what they have right in front of them.

Generally as developers I think we are looking for feedback on the direction we are going. Thinking people will see what is front of them and realize where it is going and provide feedback on that. In this way we look at the prototype as being a micro sample a sort of tiny preview of what the game is to become. However, the majority of (if not all) folks giving feedback will focus completely on what is in the current “demo”.

Which does make sense. Nobody knows what we have in our minds for the game. Even if we tell them our goals and such for it… it doesn’t matter. The feedback will be focused on what is actually in the current demo. Things that are missing that would be good to add (generally obvious stuff that we already planned to) Things that are in that should be improved (often things we know and even mentioned).

So… it’s not so much that having others test is bad or that prototyping early is bad. It’s just they are lumped in and seen as the same thing. If YOU prototype early and often then it is logical. As devs we have to do that and should be doing it automatically. But only throw it out for feedback from others when you really have something with some meat on it.

They are two separate things. It may be 2 weeks of work or it may be 2 years of work before there is enough in the game for others to be able to really understand what the game will become. Depends entirely on the overall scope of the game and how unique our design is. The closer it is to other games the earlier others can relate and give feedback. The more different it is the more we have to implement before they can understand it well enough to give the kind of feedback we’ll find most valuable.

2 Likes
10

I really agree with that.

It’s also worth noting that although the actual play testing wasn’t so valuable to my process. The goal of it was.

Keeping in mind that other people would play the game is - as stupid as it sounds - something that I think we can lose sight of. Although the feedback never really gave me what I was looking for, having the goal of something playable by other people was immensely valuable and really forced me to deal with little details or annoying little nuances that might have slipped through the cracks.

3 Likes
11

Very true. It could be we both were just looking for a different kind of feedback than what should be expected from such testing.

On the other hand… I think there is also a big difference in who is testing it. I have seen plenty of game dev blogs of others and their fanbase will be requesting features, saying they like where the game is heading or they don’t like where it is heading but would like it if this changed, etc. So it may be such testing needs to be focused specifically on the game’s intended audience. Or at least split into different tests. Like throw it out here for general feedback to nail the usability issues and then later throw it out to the fanbase for the actual game design feedback.

What do you think about doing something like that?

1 Like
12

I wonder if the Stardew Valley setup can still be prototyped, though? Each system is in essence a small game unto itself; thus, each can be prototyped to determine A) if it should get in, and B) what would make that system fun.

1 Like
13

I agree. Prototyping and play testing are powerful, extremely useful tools. But, especially as you get more ambitious, you really need to have a sense for what kind of feedback you’re going to get, and have very clear goals for the feedback itself.

I really think that this kind of thing is really one of the arts in game dev. Really experienced game devs will probably, at gut, have a sound understanding of what to expect and how to approach prototypes and playtests.

I still don’t really know enough about this stuff, but I definitely think that this is a far more complex and nuanced subject than we generally give it credit for. And for noobs, like me, even after being through many iterations of this, misunderstanding how to organize play tests and what to expect can be really tough, even counter productive.

1 Like
14

1 Like
15

To be fair I believe it was Peter Molyneux who said that he plays a game in his head for months, if not years before making it. I do the same, but the “it” I make is still the prototype.

2 Likes
16

In theory, sure, but remember that the value of the game as a whole isn’t determined by the value of all the pieces added up. Isolating a system and working on it alone can just as likely isolate the system from the rest of the game’s design, even more so if you start messing with the kineasthetics and the feel. Stardew Valley already pushes it with the mine/dungeon/combat parts being mostly a distraction. Expanding on that would just hurt what value it has on the rest of the game.

1 Like
17

I was just about to post this when I read your comment @hippocoder great timing.

It seems to me like much of this stuff comes down to time and clarity. The more time spent playtesting the game in your mind before ever doing anything externally makes a huge difference in the implementation and prototype.

The more time spent on the virtual playtesting (which really is designing the game) the more clarity you have of exactly what to build and put into the prototype.

So… it’s like there are two very different meanings going on here. One school of thought is to just come up with a very basic idea (or three) and immediately prototype it (them). Experimenting and trying to find the fun, if any, inside these game concepts.

At the other side, you have people who already have a very clear vision for their game. They’ve basically prototyped it for a long time in their own heads. When they build a prototype it is not so much about trying to find the fun as it is just bringing their vision to life. Their prototype is perhaps more to provide something concrete so other people can better understand their vision.

In the first scenario the person is basically not even sure what they want to build. So they prototype searching for answers. Guidance on what the game should actually be. They don’t have a clear well thought out design and are looking / experimenting to find one.

In the second scenario the person already knows what they want to build. That game already exists in vivid detail in their mind. They have played it a hundred times or more and modified it here and there improving it in the process. They prototype to create a real world playable version of their game. They are not looking for a game to build or trying to “find the fun” although they may stumble upon such a thing in the process.

So basically the one person is kind of stumbling along searching testing trying to find answers. The other person is already well prepared and has a game basically all designed in their head.

Yet when people discuss prototyping they do so in a generic “one size fits all” way and that is what likely causes the confusion and makes it more difficult for each person to relate to it.

Any thoughts, butcherings etc on that?

3 Likes
18

From the guy who literally turned his first day with Unity project into a game?

19

Pretty sure that was the prototype.

1 Like
20

Pretty sure Molyneux has always been the type to make shit up as he goes though.

1 Like