Modernization Hub

Modernization and Improvement
The Story of Celeste’s Development

The Story of Celeste’s Development

(“Resurrections”) – The larger game, Celeste, has a lot of story and a lot of atmosphere and it didn’t really exist
in the PICO-8 version. It started off as a much more simple like, we know it’s about someone
climbing a mountain, we know that it’s really difficult, but we were kinda just
making a hard platformer and the limitations of
PICO-8 are quite strict. We’re limited to about 30 screens, limited to like 16 colors. So it was basically just us being like, let’s just make something fun. So we made that little game and we put it out on the internet and a lot of people really enjoyed it, especially speed runners, and they actually played
it at Games Done Quick. It was few months later,
I can’t remember when, but we just decided like, oh let’s like maybe just try
making a slightly larger game but still a small game in
like two or three months, just like take this and we
started getting new ideas for what we wanted to do with that world and we liked the character and we liked the little
mountain and stuff. And we said, okay, let’s
spend like three months and make a larger game and something we’re more familiar with. (“Resurrections”) Yeah, so Matt and I started together and we were working on it mostly
ourselves for a few months. At the time when I was
working with Pedro and Amora, they were the same artists who
worked on TowerFall with Matt so we knew them through that fairly well. And I was actually working
with hem for a few years on my game Skytorn that
never got finished. Around the same time I think we started talking
to Lena about like, we’d heard her music
just through SoundCloud and seeing her at events
like GDC previously and we’re like, oh man, we
really like her SoundCloud stuff and like the work she’s doing so we just hit her up and asked her about it and she was keen to jump on too. And then Power Up Audio,
the sound designers, they’re the same sound designers who were working with Matt on TowerFall and I also had them working on Skytorn. So it is kinda like, a lot
of people were from Vancouver or people that we’d known for awhile or people really liked their work. (“Heart of the Mountain”) I would say Chapter 1 is
kind of what we envisioned like when we were talking about making a few month long project. Chapter 1 is really like, what that was. We started doing that and we
decided we wanted more chapters and we wanted to explore this game further because I felt like there was a lot there that we could do with it. That’s when it really branched out and I’d say that the different chapters really started taking shape. We started like getting
pretty far into even Chapter 3 before we really knew what it all was and even before the story
had really come together. We knew that there was
gonna be a bit more story, we had the granny character at the start of the game who laughs at you but we didn’t really
know where that was going until I’d say around Chapter 3 and then things started falling into place and we started going actually back and kind of redoing some of the dialogue. Oh, this is what this is now, Matt really led the way on that, he had kind of a strong
vision and focus on it. Yeah, even the character
that’s like part of you, the darker side of you, she originally was a gameplay
mechanic, she wasn’t, you know, she started off as something that mechanically was interesting and then when the story
started really taking shape and we understood better
what the game actually was and what it was about
we actually went back and started adding her
as proper character. What would often happen too is Matt would skeleton out levels but everyone
else was like, you know, focusing on catching up with art and finishing various things
and we were like planning kinda where things are going and he would just be like in
the meantime kind of just like making the skeletons of levels. So sometimes he’d get pretty far ahead and then we’d come back and be like, okay, so what exactly
emotionally happens here and fill it in. I’d say that changed by
around I think Chapter 6. (“First Steps”) I think, yeah, it really
comes down to like just making sure that
when someone picks it up it’s fun immediately, especially
if it’s an action game. You need to be able to just like hit a button and have it feel good, like you’re like, I get this. It’s definitely, it took a
lot of iteration to get there and it’s not something
that happens immediately but just kinda like we just
watched people play a lot and whenever something felt kinda weird or kinda wrong we would
go back and tune it over and over and over again until it really just felt
like people were flowing as best they could. And you don’t wanna get rid of everything, you don’t wanna polish off all the edges and make it so there’s
nothing interesting anymore. But yeah, I think it just has to be, is just has to feel really
good to pick up immediately. Because the whole game is just
about the moment to moment. I mean, that’s definitely where
play testing came in a lot is because it was not
always smooth for people. I mean, it’s still not for everybody but we tried to make it
as smooth as possible. Like the learning curve
is as nice as possible. And still teach them and still have it so you feel like you’re
learning and getting better. But the nice thing
about a platforming game is you can basically hand it to someone a week into making it and be
like, how does it feel so far? And they can play it, like
you only need a few levels for it to like, okay, I can feel this. And we’d have friends come
by for hours at a time and just kinda like, we
would watch them play it and make changes and then
we would live update it. So it’d be like, okay, that’s not working. Just stop for one second while tweak it and then we’ll commit it and pull it down and you can auto reload the game and have you try that again. If we never implemented like analytics, maybe you had something where you can see where people are dying a lot or something, you can kinda get a good idea. But I think it’s a lot better to kinda see like people emotionally and like what they’re getting handled, you can see it visually and
that’s a lot more useful. I think often times feedback
from players, like verbally, it can be good but in an action
game it’s all about feel. It’s much, much better to
like watch what they’re doing and what’s really working
for them and what’s not than it is for them to say, like, I didn’t like this or
like this or like this and you don’t really know why. (“First Steps”) You know, we hoped that
it would do okay and stuff but we weren’t expecting the
reception that we got at all. Everyone has a lot of input and influence and it was a very, very
collaborative experience. But Matt I would definitely
put as the role as director. I mean, I was his roommate
for the whole of Celeste so we were just hanging
out in the living room talking about it every day. My goal personally with it,
besides making a good game was to be like, I wanna
keep living making games and hopefully this can, and that, it definitely did that which was awesome. Yeah, it was really cool, I hope we can keep making
stuff that people like. (“First Steps”)

98 comments on “The Story of Celeste’s Development

  1. I wish this was a full dock (edit) but thanks for making it. With you find the time please revisit Celeste and the studio formally know as "Matt makes games"

  2. Celeste is one of my biggest discovery of the joy from indie games. After Hollow Knight, i couldn't think of a game that could blend great design, amazing soundtrack and a steep difficulty curve from any indie games that i've ever played.

  3. I'd really like to hear about the whole development process of hollow knight even if it is 12 hrs long. The developers are awesome and I'm really looking forward to silksong.

  4. I love when Developers (Indie or Big Companies) put their hearts in "small" projects and turns up to be the biggest step in their careers. The game carry the old school nostalgia and as a gamer, I love that.

  5. Bummed this is so tiny cause I adore Celeste. I'm interested to see where Noel and Matt go from here. Towerfall was a neat little game but being strictly multiplayer I never really cared but Celeste was truly something special. I remember jumping out of my chair and fist pumping the first time I finished climbing the Summit. I loved the game so much I went back and found the hearts so I could tackle The Core and all the B-Sides.

  6. Been following Noel Berry and Matt Thorston since 2009 or so, I used to make little flash games based on the ActionScript framework Flashpunk…Back then Noel was this 16 yo kid that popped up in the community alongside Thorston and co…It's crazy to see him become so successful, making a great game like Celeste. Looking forward to seeing what he and his team do next!

  7. ^‿^ Adored this game. I love that a teensy-tiny indie developer managed to make a title that — at least to me — lives up to the big leagues.

  8. It's so weird to think that the same Noel that I saw at a cafe in Winnipeg ages ago as a teenager learning XNA and moving away from FlashPunk work on a small passion project that just spiraled out of hand and paid off in the end. I mostly wish the documentary was full length, but it's understandable why it's not.

  9. That game was really great. It was nice (though a little too short) to hear where the game came from. My son and I really enjoyed the game a lot. The music was great and the gameplay really was challenging.

  10. Great video. More folk should appreciate the art that is Celeste.

    Do you guys take suggestions?
    I was hoping you guys could do a doc about the Batman Arkham series.

  11. I am constantly fascinated by how developers discover their own games. It seems like the best projects rarely start as what they end up as. I remember Klei saying that Invisible Inc (my favourite game in recent years) started as a card game… which is so bizarre to me that such tight, and purposeful game was a result of an entirely different project. I think in crowdfunding we see a lot of games, which are put together competently but suffer from having to be what they were promised to be. It feels like games to some extend reveal what they can be during the development and sticking to vision one had before the game was forming just isn’t good for the final product.

  12. 4:56–5:28 is my personal highlight. you cannot be clever, you cannot automate the playtesting process with statistics or self-reported surveys, you've got to empathize with the individual to make any sense of what your game experience is like.

  13. The best platformer of 2018, without question. I attempted speedruns of it pretty close to launch, there are some ridiculously brutal techs involved that I just couldn't get, but the sheer process of playing it multiple times in a single sitting rapid-fire was a class of reward all its own.

  14. To anyone wanting more behind the scenes Celeste stuff, there’s a really interesting GDC talk Matt Thorson gave on the level design of Celeste and I’d highly recommend it

  15. This was an incredible game, I was going through a transition period in my life and anytime I doubted myself, one of the thoughts that got me through was "I beat Celeste" … might be silly but anyone that's played the game and fought through the levels we thought were impossible knows what I'm talking about

  16. God this was great. I also felt teased and realized now I need to hear more about not just Celeste, but Towerfall too. ugugugh, great work Danny!

  17. I thoroughly enjoyed this game, great level design (especially the B stages!) I applaud these guys for the quality they showed, a true gem!

  18. … Oh! I thought this was a trailer! It's the whole thing!
    I would love to see like an hour long video on Celeste! But this is good too.

  19. Man this game is satisfying. When you're able to pass through a screen that seems impossible at first, it feels great.

  20. I think Celeste might be perfect. I'm really picky when it comes to platformers (and bad) and this one feels amazing and the difficulty curve is just right. And I never expected the story and characters to connect with me so deeply. And the art and music are just outstanding. I really can't say enough about how much this game means to me

  21. It's great to see how developers started with small projects that eventually turned into pieces of art that are getting nominated for world-wide awards. I've had a blast with Celeste and even though there isn't going to be anything new after Chapter 9, I'm looking forward to what new projects they will start.

  22. After watching this, I realise what I really want to hear is the emotional side.

    It sounds like an experienced team working a smooth process, so I'd like to feel more of the heart they put in. What is the personal inspiration for the team members, what are stories from their lives that influenced this.

    I want the cornier doc ?

    Maybe that's because I also have been following the indie game scene since 2007, flashpunk, local multilayer scene, pico 8, etc. I knew this broad line story already.

  23. not everything is pink and dandy about celeste… if you look up the "I'm sure it's just my tinfoil hat, but all the attention on Celeste has me Paranoid" reddit post, I think its pretty interesting

  24. The fact that they analyzed feedback from play testers from an emotional lens rather than a data-driven formulaic lens is a really underrated in a fascinating insight into how games that feel "right' are made. There's no committee judging what direction this game should go based on causation without correlation from statistics of testers, to understand how the game is going to work you need to let the art breathe and exist with the player. Companies like EA pretend video games are stocks. Indie devs remember that games are meant for two things: artistic expression and fun.

  25. You took meatboy and said "this dude needs an airdash and double jump i am a genius, not even kojima would do this shit."

  26. As a game creator myself this is very inspiring. I've had the game in my steam library for a while, but this has made me want to play it. Keep up the great work.

  27. One of the best games I've ever played, this should be made physical not just a limited run game.

    Everyone should have the chance of owning and playing this game.

  28. I really love this game and its developers.

    I live in a kinda poor country (Argentina), and buying games it's expensive, if it's single player you just pirate and you're done, i did it with Celeste.
    But after replaying i was decided, i went to pay the steam bill with effective to a supermarket like 15 km from my home (rural zone), a really sunny hot summer day, i was walking under the sun and really felt like the effort Madeline does to climb the mountain, it was kinda an empathic experience.

    I also play piano and bought the physical book Piano Collections to learn the music, Lena is an incredible composer, i really feel what she does and totally merge it with the level design (ej: from Quiet and Falling to In The Mirror she swaps the notes and works natural and perfectly)
    It was another headshot to my wallet but it's totally worth.

  29. This game is my favourite! Idk of I want to be a games developer but computer science is damn cool and listening to how projects form is so interesting.

  30. This game is so good. it is easily up in my top 5 favorite games ever with Mario 64 DS Mario Odyssey and Hollow Knight. I absolutely love the work the team has done with this game. Thank you to all of the developers as a part of Matt Makes Games Inc.

  31. This guy is like, amazingly honest. I don't think he has any moment where he holds back or reconsiders anything, not even movements and expressions. Man, that is good to see; that people still are that way.

  32. 5:50 Randy Pausch once said; "If they're standing close to eachother, the world is good".Just look at that dream team! Still hoping it'll one day arrive on

  33. 5:39
    Jesus. What´s up with his face expression while receiving that award? Almost looks like he's pissed. That was really weird.

Leave a Reply

Your email address will not be published. Required fields are marked *