What did people think of my game demo? (Developing 5)

Hi, my name's Mark and I am making a video game about magnets. So in the last episode of this series, I released a minimum viable product: a super short, 30 minute game demo, made in order to see how my game feels as a complete product, and to get some feedback. And boy did I get some feedback. I got detailed Discord messages. I got comments on Patreon and Itch.io. I got loads of survey responses and endless bug reports. I got full Let's Play walkthroughs over on Youtube and people played the game over screen share on Zoom. I was also lucky enough to have the game played by the designers of some kick-ass games like The End is Nigh, Room to Grow, Celeste, and It Takes Two.

Plus I also got some fan art, there was a 10 minute speed run of the game, and Untitled Magnet Game is now on How Long to Beat, which is pretty hilarious. But what did people actually think of the demo, and what's next for the development of the game? Stick around to find out.

So the very first thing that pretty much everybody commented on

was the feeling of the character. How does it feel to move and jump and pick up carry and throw the magnet? And the answer is… not great! Not great at all. There were lots of complaints about the slow acceleration, the floaty jump, that weird annoying little hitch of inertia whenever you hit the ground, the over complexity of the aiming and throwing system, and the complete lack of options in terms of rebinding controls and accessibility.

I think the only thing people actually liked about the character was how it looks and the satisfying snap of the magnet when it hits a metallic surface. Now I wasn't exactly blindsided by this feedback. I knew for most of the development that the character just didn't feel that great to control. But I kind of had myself in a bit of a bind: I had made a bunch of levels and so if I radically changed the character I'd need to go back and change all of those levels, and I had neither the time nor the energy to do it. Not a great excuse, but that's just what happened.

But whatever the case this does give me a very good, let's call it… action point for future development. I need to radically overhaul and improve the character controller. Basically throw out pretty much all of that code and start from scratch. I just need to make a character who feels really good to control.

The next subject of feedback is what I'll call design language. This is how I, as the designer, communicate to you, the player. Perhaps how to play the game or the rules and logic of the game world. And the feedback here was also not terrific. There's a lot of stuff in the game that is a bit unclear. These look like platforms, but they're actually screens on the background layer. The complex aiming system is hard to explain with just simple icons. And, perhaps worst of all, the green and orange buttons are basically indistinguishable with certain types of colour blindness. Ah, sorry about that one. There's also a lot of stuff in this game that is kind of inconsistent: for example these one-way platforms block the magnet, until later they don’t. And how come you can stand on the magnet when it's attached to a wall, but not when it's on the floor? And these green buttons seem to change their functionality from level to level with no good reason.

And this is just kind of bad design. You see, when people play games they build up a mental model of how the world works. Like, okay, if I jump on a Goomba it dies. And if those rules are then randomly, arbitrarily changed the player has to throw out their mental model and can no longer confidently predict how the world is going to react to their actions. Which just doesn't feel good. So that is another action point: I need to focus on clarity and consistency. I need to put myself in the shoes of someone who doesn't know how the game works under the hood and design the game with that viewpoint in mind.

Next up, let's talk about the levels - which were kind of a mixed bag. The stages that I spent a long time working on were generally well received, and the stages that I threw together in a rush at the end of the deadline were seen in a less favourable light. Fair enough. Now, I'm not going to go into every individual stage in the game, but there were some key universal takeaways that I saw in pretty much all of the playthroughs of these levels.

Let's start with the most important one: in a previous episode of this series I talked about my challenge of defining the game's genre. Should this be a primarily puzzle based game like Braid or The Swapper, or more of a platformer like Super Meat Boy or Celeste? In the end I decided to go for a game that is primarily about puzzle solving, so you get levels like this one which are all about using logic and deduction to solve the problem at hand. But then I could throw in some light platforming levels like this one, that's all about speed and dexterity, as a sort of palette cleanser.

Now this seemed like a good idea at the time, but then when I actually watched people play through the game I realised that this had some pretty dire consequences. Perhaps the best way to sum it up is with this piece of feedback from Matthew of the YouTube channel Video Game Story Time: MATTHEW: “My instinct is, because there's puzzles where sometimes it's just dash through, and sometimes there's puzzles where I need to actually think and engage my brain, I'm having a hard time spotting when a puzzle involves me thinking things through and when it just needs me to dash in and try my best first time”.

And he's not wrong, and he's not the only person who fell into this trap. I saw a number of players try to solve puzzles using their platforming skills, thinking that the solution was just to run a bit faster, or jump a bit higher - not realising that they actually needed to use logic, deduction, thinky skills. And, look, I do think it is possible to marry action and puzzle gameplay. As I saw in a lot of the comments for that earlier video, there's a game like Portal which is primarily a puzzle game, but you need to have FPS smarts, skills, and fluency to get by. But I'm not Valve, I’m Mark, and this is my first ever proper game - so I think what I'm gonna do is just try and focus on doing one thing well. I want to focus on the puzzle design, I don't want players to have any doubts in their mind about what sort of skills they need to bring from level to level. So I'm going to ramp up the puzzles and cut down on the skill and execution

side of the game. I'm also going to massively restrict the player's jump distance because right now he's giving Super Mario a run for his money.

And here's the thing: when people did actually come at these puzzles as puzzles, and solve them, it was really fun to watch. Especially when they made a noise like this: DOTWO: “Ah! Okay…” That is the “a-ha moment” that every puzzle game designer is striving to hear. That moment when the player stumbles upon the clever lateral thinking twist at the core of the puzzle. It is a great feeling to be able to give the player and I want to give it over, and over, and over again.

Unfortunately I didn't always achieve this in my MVP. Levels like this one were just a bit too easy and players just kind of stumbled into solving it without even really thinking about the solution. And some levels were just way too hard: a lot of people bowed out of the game before ever finishing this one, and so they never got that cool a-ha moment. But this does give me, at least, something to focus on. Another action point: make players say a-ha. And I think I'm going to get there over time as I get more experience building puzzles. Plus, as I do more play testing. What I've realised is that puzzles are essentially impossible for me, the designer, to test: I already know the solution. Plus puzzles are often about luring players into assumptions and traps and you can only see if those work if you give the level to someone else to play. So going forward I'm going to be using my Discord to let players try out puzzles and give me feedback,

so I can iterate and evolve those designs. More information on that in the future.

Now there were a few things that were really holding my puzzles back from true greatness. One of which is an over abundance of stuff on the screen. It is not uncommon in Untitled Magnet Game to go into a room and just be bombarded with lasers and moving boxes and electromagnets and more. It makes for puzzles that are messy and convoluted and a game that just feels a bit overwhelming. Now I know exactly why this happens: I would always start the puzzle with a really simple and elegant design, but then I'd start to find ways that the player could very easily skip and bypass the puzzle. So I need to put in an extra door here and then a laser there, and before long things have gotten completely out of hand and the puzzle now looks like this.

Now I did get some really good feedback on this from Oliver Granlund over at Hazelight. He says that when situations like this emerge, often the problem is a gap in your toolbox as a designer. There is some kind of game mechanic that you are missing that would solve that problem in a very elegant and simple way, so you just need to go off and build it. He also suggested using the camera to frame small parts of the puzzle at once to avoid overwhelming the player with too much information on screen all at once.

So that's another action point: focus on puzzles that are simple and elegant, and use more tools and perhaps camera framing to help cut down on information overload. Another source of frustration was times in the game where you just kind of have to wait for the game to align itself. It's especially common with these moving blocks, where you just have to wait for the cycle to reset before you can even try solving the puzzle again. That's just kind of tedious and annoying and it discourages experimentation with puzzle solutions. So there's another action point: don't do that.

And finally on the level design: well for pretty much every puzzle in the game, it turns out that there was a way for players to essentially skip or bypass the whole puzzle by using some kind of bug, glitch, or oversight. And I want to state that I think it's fine for there to be ways to overcome a problem in a game in a way that the designer did not necessarily intend. In fact I think it can be cool. That is, provided the solution is interesting, it should require some level of skill and mastery, and it should be made perfectly obvious that this is not the intended solution, but a clever way of getting one over on the designer.

For example, there were a few players who managed to throw the magnet up into the air and then at the moment where it blocks the laser they could whip past and get out of danger. That is really cool, so I'm happy to leave stuff like that in the game. But if the player is kind of fudging into the edge of a collision zone to grab a magnet on the other side of the block and sailing across the level in a really glitched and awkward way… that's got to go.

PLAYTESTER: “That doesn’t… yeah that doesn't seem right…” So another action point is that I need to improve my bug testing and bug fixing processes. Going forward I kind of have this philosophy that I'm calling the airlock, where you can kind of imagine a workshop where you make the new game mechanics and ideas - and then there's the game itself. And in between there is an imaginary airlock, and nothing can pass through that lock until it has gone through really rigorous testing in different scenarios and stuff.

I'm never gonna catch all of the bugs, but the idea is that nothing can go into the game unless I'm reasonably confident that it can withstand a high level of scrutiny.

So all of this feedback is stuff that, in general, I was expecting or anticipating. I knew the game had plenty of bugs, I knew the character controller didn't feel that great, I knew some of the puzzles were a bit rushed, I knew that I broke some of my own rules. This was all stuff that I was aware of but just didn't really have the time to fix. But there was one piece of feedback that I received that I was not expecting and it did give me pause. It first came from the user YK on my Discord who said “it doesn't feel to me like the magnet is the core of the game”, and for a game that is supposed to be literally entirely about magnetism, that's probably not good.

So I made sure to talk to people who were giving me this feedback to get to the root of the problem and it turns out there were a few different sources. So some people thought the game felt like a pretty standard puzzle game with a bunch of different game mechanics: like lasers, moving boxes… magnets. When really the magnet should be the core of the game, not just one other game element! And to their credit there are some scenarios in this MVP where the magnet is not involved at all, which is a pretty huge oversight on my part. So to fix this I think we need to install a big core pillar of the design: this is a game where you overcome problems through clever use of magnetism.

And pretty much every single scenario in the game should involve you using the magnet in some way. Others thought the magnet felt a bit rudimentary, a bit too simple to be a core mechanic. And this is something I had always planned to address. Right now in this MVP there's one simple type of magnet: a red one. But eventually I wanted to add a red one and a blue one, and a way to switch between those two different polarities.

Giant magnets that can attract stuff towards you. A magnet you can recall back to yourself. And perhaps other ones as well. So maybe I need to introduce those ideas a little bit sooner so players can quickly understand that this game has more to it than just throwing a red magnet around. And also some players thought that the magnet was a bit boring: like just a standard inanimate object. And perhaps the solution to this is something that Nintendo does a lot, where it makes its game mechanics more crucial by turning them into characters.

Think about Cappy and Navi and Ezlo and Fi and the Baby Luma, and loads more. Perhaps I could do something similar for my game give the magnet a face, a personality, perhaps even dialogue and little cut scenes? I don't know, but anything to make it feel more like a character. Whatever the case, that is my final action point from the MVP: make the game way more magnet-y.

So this was my first time getting real feedback on a game that I had personally designed and the experience was… okay, pretty interesting! For one thing it was just super valuable. I learned so much about what to do next from all of this feedback. I know what works what doesn’t, and where I should be putting my attention next. Seeing people enjoy the game gave me huge motivation to make more of it, and seeing people get frustrated at certain things told me exactly what I should cut out. And with these action points I have a really strong direction for where to take the game next - far stronger than anything I had when I was working on the game in isolation.

It also wasn't as painful as I was perhaps expecting. Sure it was a bit cringe-worthy at times to watch some of the videos, and it was a bit of a hit to the old ego also. But I think because I was expecting a lot of the feedback that I saw, and pretty much agreed with everything people said, it wasn't as painful as it could have been. I guess the worst parts were seeing people glitch their way through some of those puzzles, which is why I released a quick version 1.2 release with a few bug fixes. And hearing the same piece of feedback over and over again., which is nobody's fault - nobody knows what other people have already said. But to me, I just end up hearing the same thing repeatedly, and it is a bit tiresome.

But more than any of that, it just felt really cool to see people playing this thing that I had made. When I wasn't cringing myself inside out I had this huge smile locked on my face as I watched the playthroughs of people enjoying the game. It's an amazing feeling: like I have made a game, and it's on the internet, and I've got hundreds of players. That's not nothing! Here's what Oliver, the guy who sent me this MVP challenge several months ago, had to say about the outcome: OLIVER: “It’s super impressive. Like, that you got this far and released it that in itself is a huge accomplishment. Like, seriously, you should you should feel that. And try to ride that for a while, right? It's out there and you made a game!”

But, okay, enough patting myself on the back. What's next? Well for starters I'm going to essentially scrap the whole MVP. Like, it's served its purpose. I can perhaps salvage some code, some art, some level layouts. But it's kind of like a Jenga tower of bugs and mistakes, so it's gotta go. I'm gonna essentially start the game from scratch. That being said I have now released the MVP on Itch.io to the public if you'd like to check it out, as a sort of historical artifact of this developing process. But now I'm gonna move on to making the game for real. I guess starting with that first crucial action point: the character controller. It's time to make a robot who actually feels nice to play. I will show you that in the next episode.

Oh before you go, this whole developing series has been nominated for a Unity award, which is hilarious. It would be beyond amazing if you could take the time to log into the site and vote for this series. How about this: if GMTK wins this Unity award, I will finally make that “basics of Unity” video I teased in the first episode of this series. Sound good? Thanks for watching see you next time. Beep.

Why did I do that?