Did I complete my 30 day game making challenge?

Hi, my name's Mark and I am 
making a video game about magnets.

So in the previous episode of this series, I was 
set something of a challenge from a professional

game designer: to make a minimum viable product. 
Or MVP. Which is basically a fully functioning

game, with everything that makes a game a game, 
except the whole game can be extremely condensed.

And what's more I have to make this MVP under 
a very strict deadline. I've given myself

30 individual days to make a fully featured game 
with five complete levels. So how did I get on?

Well, you're gonna have to 
keep watching to find out.

Okay, so before I jumped into Unity I 
thought the best idea would be to make

a plan. How am I gonna spend these 30 
days? And so this is what I came up with.

I'd spend the first five 
days working on the character

and the magnet. The absolute basic stuff.

Then I'd have five days to make level one.

Make all the level architecture and 
the game mechanics and stuff like that.

Then I'd have 15 days to make levels two, three,

four, and five, with the assumption 
that I'll get faster as time goes on.

That leaves me five days for meta elements 
like a pause screen and saving and loading,

as well as polishing and bug fixing.

I think that's a pretty good plan and I can't 
think of anything too important that I've missed,

so let's crack on!

Spoiler alert: I, uh, I 
missed something important.

So it's day one and it's time to do 
something that I've been wanting to do

for a very long time: give my game some graphics. 
I am sick of my game being a white rectangle

jumping around some blue squares. It’s time 
to give my game some character and some life.

Now I was really pleased with the response to my 
pixel art in a previous episode of this series,

when I showed you my old Picross game. So I 
decided to do pixel art for this game too,

which has the added benefit of being pretty fast 
to produce. So in a matter of hours I had made

a character, who is now a boxy orange robot. 
A magnet with a slightly unnerving face. I

might remove that. Some key gameplay elements 
like an electromagnet and a button. And

some background decoration and some floor tiles.

Day two and I started a new project in Unity and 
started working on the character. I pulled across

a bunch of code from my prototype, but with the 
aim of refining it and rewriting it for this

MVP. For example the prototype never had any 
animation, so time to figure out how that works.

Which I realised… it was going to be kind of time 
consuming to make a full run cycle for this robot,

so I decided to cheat and just kind of cut 
its legs off and replace it with a wheel.

A bit cheeky but with some nice dynamic poses 
and some particle effect it looks pretty nice.

Day three and it was time to put the magnet into 
the game. Now I wanted to develop my own magnetism

system. I’ve been using the built-in point 
effector in Unity, which is cool but does have

some limitations. Unfortunately I'm not a strong 
enough coder for this to really work - my system

was good but had loads of bugs and weird edge 
cases, so I decided to cut my losses and scrap

most of it. But I could reuse some of the code 
to make a nice juicy system where the magnet

subtly tilts towards nearby magnetic sources.

Day four was all about magnet states. So I now 
have a red magnet that's attracted to blue stuff,

a blue magnet that's attracted to red stuff, a 
gray sleepy magnet that's attracted to nothing,

and a way to switch between 
them with a button press.

And then day five was all about picking 
up and throwing the magnet. And this was

another case of my pixel art making things 
more complicated. How do I make the character

pick up and hold a magnet, and make it look 
good with these pixel art arms? So I cheated

again: I cut the robot's arms off and just have 
these floating hands to pick up the magnet.

But, again, if I put some little 
arms on with a line renderer,

and use a springy hinge 
joint for holding the magnet,

I think it actually looks pretty cool. There's 
a lot of charm and character here. And so

in the first week I now have the character and 
the magnet moving around an empty environment,

and it looks pretty good. So I'm happy with 
that, and it's almost time for level design.

Though, first, I had to take a quick week off 
to work on the Metroid Dread Boss Keys video.

I am still trying to run a YouTube channel here.

Okay, now it's time for some level design and

the idea for the first level is that it kind of 
just introduces you to the concept of a magnet,

and has some nice simple scenarios and 
challenges to test you on those basic skills.

Now in order to actually design the 
level I made, what I consider to be,

one of the greatest purchases of 
my entire life. It’s a whiteboard!

This thing is amazing. This actually isn't even 
the one I purchased - I purchased this absolutely

huge one but it's screwed to the wall downstairs. 
But it was so good I bought another one just to

have next to my computer. It’s not just great 
for level design but also for to-do lists and

pseudo code and doodling ideas. I use 
it every day, it's the best thing ever.

So the way I actually did this was to draw 
out a small scenario on the whiteboard:

perhaps just a magnet to pull down 
a weight, which opens up a door on

a pulley system. And then take a photo 
of it, erase it, draw out another one,

and repeat. Then put all the photos onto my 
computer and arrange them in a way that I thought

flowed quite nicely, and had a nice 
sort of casual incline of difficulty.

I could then draw the full… this made much 
more sense before I realised how tight the

frame was going to crop in. I could then draw 
out the full thing on the whiteboard and, uh,

remake it in Unity. And so that's what I did for 
the first day: I made the level architecture using

the tile map, and I designed and drew and coded up 
the first mechanics, like the weight and pulley,

this crusher thing that can destroy 
your character, and an electromagnet.

I then wasted an entire day trying to 
make a really complicated enemy using,

like, a finite state machine AI 
system, which was absolutely overkill

for this project. So then kind of, again, cut my 
losses and just made a really simplistic enemy.

basically like a Goomba that can move left and 
right and you can bop it on its head to kill it.

By that point I had made most of the 
level so I kind of just went like…

‘done!’, uh, time for polishing!

So I used Unity's 2D lighting system to 
give the game a bit more depth and visual

interest. I had those floor tiles from before 
but it would be really tedious to put down all

of the floor tiles in the right positions. 
But here comes Unity with the assist again,

with its experimental rule tile system. You 
basically teach Unity how your tile map works

and now you can just draw it on and it puts down 
the tiles in the right places. It's an amazing

system. And I also played around with Cinemachine 
which is this built-in camera system that can do

really powerful stuff like have it track the 
player in a really nice way, and then subtly

transition to a zoomed out static framed shot of 
a puzzle room, before following the player again.

And then finally I just needed to add in a few 
more elements to make the level actually work like

checkpoints so you don't need to go all the way 
back to the beginning when you die, and a way to

respawn the magnet if you chuck it into the 
spikes. To illustrate this I used a chunky green

pipe as a sort of winking homage to my favourite 
video game series of all time: Flappy Bird.

So now I am on to the big 15 day sprint of 
making levels two, three, four, and five. But

over the weekend I had an idea for a puzzle. 
Now I'm gonna have to be a little bit vague

about how the puzzle works, not because I'm like 
super precious about my level design, but because

I need to get feedback on these puzzles at 
some point and that's not gonna work if you

already know the solution. So forgive me while 
I kind of dance around it a little bit. But I

drew the level idea out on the whiteboard, 
and then remade it in Unity. And then

realised that there were loads of problems 
and ways to cheat it in Unity so I had to

work on it over and over again, and kind 
of iterate on the idea until it worked.

But I am really proud of it. It was a lot of fun 
to make and I think it is a pretty cool puzzle. It

works on a lot of that stuff I talked about in my 
puzzle design video of using assumptions to lure

the player into realising that there's basically 
an impossible catch to the puzzle, forcing them

to think in a more lateral way to come to the 
actual solution. The puzzle itself is a single

screened puzzle chamber with lots of moving parts 
and complexity. I just really enjoyed making it.

And what's more, to make the puzzle work I had 
to design some brand new game mechanics like

a laser beam that you can block with the magnet, 
which ended up inspiring some new puzzles about

lasers. And I also had to make this plug 
and socket system where you can transfer

power from one block to another, 
which itself inspired some new

puzzles. And those new puzzles led to new 
game mechanics which led to new puzzles on

top of that. I mean puzzles were just 
coming out of my butt at this point.

But I did have a problem. You see, level 
one was just a single scene in Unity: a

long continuous stage with all of the elements in 
the world at once. And so I tried to do that for

level two by having the puzzle chambers 
all in a single scene and then, you know,

I’d put in, like, little corridors to 
link them up. But this was a bit unwieldy.

I mean that is a lot of stuff in memory at 
one time. My hierarchy in Unity was getting

way too messy. And, I think, worst of all 
it was really hard to move things around.

Like if I suddenly decide that actually the fourth 
puzzle room is quite easy and so should go in the

second place, it would be a nightmare to 
shift all of that around. It’s just it's

just not worth it. So I came up with a new plan: 
instead of having five, big, individual levels,

the idea is each level would be a series 
of small scenes all connected up via doors.

So a level might be three of these 
puzzle chambers, with a number of

interstitial levels that might teach you the 
game mechanics that you'll need in the puzzles.

Or provide some platforming as a sort of light 
palette cleansing in between the tricky puzzles.

It would be a lot more work but I think it is 
the right way to go and I'm pretty excited.

But I need to take time off from 
development again to make the

previous episode of Developing, which I 
mentioned in this video… how does time work?

And this is where things 
started to go pretty wrong.

So I came back to work on the game and I made 
one of these little interstitial levels. A

small tutorial, teaching you how the laser beam 
mechanic works. But then I needed to stop again

because time was running out for the year, so I 
needed to make the accessibility roundup for 2021.

And as soon as that was done, I had to 
transition immediately into making the

video on The Forgotten City, because I 
was just kind of running out of time.

So it's now been a full month since I last 
worked on the game, but I tried to put in

another day's work. I made some new levels 
that are basically replacing level one: you

know ,teaching you about picking up the magnet and 
throwing it. But then I had to stop for Christmas.

I worked for one day after Christmas but then 
had to stop for New Year’s. And now it's 2022

and I worked on the game for a day, 
and did a little bit of level design.

And I tried to work on the game for 
another day. I put in these little

screens that tell you which buttons to press, but 
that's literally all I did that day. And then I

had to stop again to make the Invisible Choices 
video… and this just isn't really working, is it?

You see, every time I stop development 
to go work on a YouTube video, I

lose a lot of momentum, and it's really hard 
to pick it back up again. Especially after

a month out, basically I took 
all of December to make videos.

I think what makes it even harder is that 
my systems for making the game are pretty

messy. Like my code and my tools in Unity… 
they're just a bit shonky. I've got loads

of little bug fixes and last minute patches and 
these things where it has to be set up in just

the right way for it to work, which makes even the

most simple and easy levels more tedious and 
time consuming to design than they should be.

So I've lost my momentum I've got tools that 
make things more tedious and time consuming

than they should be, and it's just sapped all 
of my motivation for this MVP. And I've done

20 days of game development at this 
point, but the thought of doing

10 more… like another 50 percent of that 
work again is, just, I don't think I can

do it. Like, it sucks, but I feel like 
I should probably just end it here.

But it's not the end of the world, because the 
MVP totally paid off! Like, the exercise did

exactly what it should have. By giving myself 
that deadline and that super specific goal,

I got myself out of that analysis 
paralysis rut. I made decisions,

I made content, I discovered what I really 
like about the game, which is making these

single screen puzzle chambers. I also came up 
with loads of ideas for how to improve the game

in the future. For example, having really robust 
and reliable and intuitive level making tools,

so I can churn out content much 
more easily and more comfortably.

And perhaps most importantly of all, I have 
made something now that I can give to people

to play to get some feedback on. Remember, I’m 
the only person who has ever played this game!

I need to give it to some other 
people to see what they think.

So it's time for operation “Git ‘er dun”.

Just spend a few more days putting together the 
last finishing touches to make the game work,

to make it playable, so that I can 
upload it and get some feedback.

So day 21. It's time to radically cut the 
scope of the game, from like 35 scenes

down to about 14. Much more manageable. And so 
I spent that day just working on the final level

designs that I would need to plug the gaps, 
and have it work as a full functioning game.

Day 22 and it was finally time to put in some 
audio, some music and sound effects. And,

whoa! I should have done this way earlier. 
This actually adds so much character and life

to the game: even something as simple as a button 
coming on and off, if you have a simple sound

effect with it, it just works so much better. I 
will remember this in future: put sounds in early.

And day 23, I crawled to the finish 
line by polishing up a little bit,

fixing some of the most egregious bugs 
that I knew about, and putting in a very

scrappy, lazy pause menu for resetting 
the puzzle and quitting the game.

And now I'm done, right? I mean, 
it's a bit messy, it's a bit broken,

there's lots of bugs and glitches that I 
know about. But I can't do any more. Like

it's over. It’s done. I've gotta just 
get it out there. It's the end. Right?

Well that was kind of the original 
plan: this video was going to be

about me making the MVP across 30 days. 
Or now 23 days. And then the next video

was going to be about me 
getting feedback on the game.

But I got a bit excited, you know! Because I 
finished the game it’s, it's playable, it's ready

for people to try. And som, you know, I gave it to 
my wife to play and her feedback was not… great!

MRS GMTK: Why am I up here? 
Mark, I think that's broken.

MRS GMTK: There's too many 
buttons, I don't like it.

MRS GMTK: I would say the jump is a 
bit… you go a bit all over the place.

MRS GMTK: Like I've got it now, so you 
don't need to hammer home the point.

MARK: It's quite tedious…
MRS GMTK: It’s really tedious.

This is really tedious, Mark.

MRS GMTK: *Annoyed cry*

MRS GMTK: I think I'm a bit tired…

But she did give me some very good advice. 
She said: you want feedback on this game,

but you want people to tell you about the 
game as a whole and about the puzzles. You

don't want people getting bogged down in all 
the little bugs and glitches and annoyances

and frustrations. So maybe just spend a couple 
days fixing those up before you release the game.

So I begrudgingly agreed. So day 24, fix up all 
of the annoyances that my wife discovered. And day

25, okay, you know I am, like, a better programmer 
at this point so I should try and make a better

magnetism system. Yeah, that actually looks 
a lot better, let's put that in the game.

Now I am done, though, right? Well 
I gave the game to my dad to play

and he discovered a whole bunch of bugs 
that I had never seen before. Like,

you shouldn't be able to throw that switch off 
the screen, that's going to break things. And, oh

no, shouldn't be able to do that. That’s, that’s, 
that's wrong, that's a disaster. So day 26, fix up

all of those bugs. And day 27, if I'm doing this 
I might as well do it properly. So I added in a

save and load system so you didn't lose all your 
progress if you quit the game. And a better pause

screen that is less ugly and stops you from 
quitting the game with a single button press.

But now I really am done, right? Well I gave the 
game to my brother to play and he discovered a

bunch of ways to essentially cheat the puzzles 
and skip the solutions. So day 28 fix all of

those problems. Oh and I left the game with my 
brother and he gave the game to my eight-year-old

nephew who managed to break the game in ways 
that only a child can break it. So day 29

fix all of those bugs. And day 30 
polish, finish, compile, upload.

So in the end I did get to 30 days, and 
that's because in my original plan I had never

considered play testing. I mean 
I'm not a complete idiot: like the

point of this was that after I finished the MVP 
I would throw it out to the world for feedback.

But I now see the value of giving the game to 
a very small, select group of trusted people

to see how they play the game, and see what bugs 
they run into, and what assumptions they make that

I never did, and to then fix and tweak 
the game based on those play tests.

Better to get those problems out of the way now 
as part of the development process, than after

giving it out to thousands of people on the 
internet. But it is finally that time! The

minimum viable product is now done and I have a 
playable demo for the world to play. Or, at least,

a very small, tiny fraction of the world because 
I'm giving this demo exclusively to my Patrons on

any tier for the first month of its 
release. And I'll also give it to

some designers and developers, including 
Oliver who set me this MVP challenge.

Because as it turns out, the MVP 
was a brilliant idea. I just kind of

screwed it up. So if you want to do one of these 
yourself, here is my three pieces of advice.

Number one, it's got to be pretty 
much your singular focus for the

duration of the project. 
You can't go off for a month

and make a bunch of YouTube videos and expect 
to keep up that same level of motivation.

Two, you have to factor in play testing. 
Give yourself at least five days

to give the game out to various 
people see what their feedback is

and tweak and adjust and fix the 
game based on those play tests.

And three, you don't actually need loads of 
content for this. It's not about generating stuff

and getting the game built, it's really about 
seeing what the game feels like, and plays like,

as a complete whole. The way I see it, a prototype 
is a way of testing the viability of a mechanic,

but an MVP is a way of testing the viability of 
that mechanic within a larger game structure. So

you want to see: is it working? Is this the right 
direction? Is it fun? And, ultimately, that's up

to you now. Or at least people who give me money 
every month. So, please, if you are a GMTK Patron

go download the game, see what you think, leave me 
some feedback, and then next month on Developing

we will look at that feedback, see what you 
think, and decide where to take the game next.

Thank you so much for watching 
and I'll see you soon.