I Made a Successful Roblox Game to Prove It’s Not Luck — Transcript
Full transcript
- 0:00One year ago, I released my first
- 0:01successful game. It got millions of
- 0:03plays and it was great, but it made me
- 0:04wonder, was this just luck or could I do
- 0:06it again? So, about 3 months ago, it's
- 0:08the spring semester of my sophomore year
- 0:11in college and I'm still spending a lot
- 0:13of my time working on Manhunt, my old
- 0:15game, releasing updates and events to
- 0:17keep the game active. But, I was getting
- 0:18more and more eager to start a new
- 0:20project. I wanted to apply what I
- 0:22learned from growing Manhunt to a fresh
- 0:24new game. And I had a game idea that had
- 0:26been stuck in the back of my mind for
- 0:27months. You see, while I was messing
- 0:29around in studio with some different
- 0:31physics, I discovered that you could use
- 0:33an Align Position to make a part follow
- 0:35you around and that dodging it was
- 0:37actually pretty fun. What you're seeing
- 0:38right now is the early testing phase of
- 0:40this, but I thought the idea would be
- 0:42cool and had some potential, so I got to
- 0:44work.
- 0:46To start, we would need a movement
- 0:48system, one that felt smooth and
- 0:49satisfying to use. So, we added running,
- 0:52double jump, and of course, dodging.
- 0:53That would end up being the theme of
- 0:55this game. After that, I created a round
- 0:57system that would put you in a map and
- 0:58spawn balls in that you could dodge.
- 1:00These balls were controlled by an Align
- 1:02Position and they still are today, but I
- 1:04went on to do a bunch of fine-tuning and
- 1:06changed up their movement a bunch to be
- 1:07quite a bit more complicated. Now that
- 1:09that's done, it's time for hit
- 1:10detection. And you might think that
- 1:12that's simple, but no. I actually wrote
- 1:14down three full paragraphs talking about
- 1:16this and I am not going to be reading
- 1:18that uh just because I know 90% of you
- 1:20guys are clicking off I do. So, I'm not
- 1:22going to nerd out. Just know that it was
- 1:24complicated and it took a few days to
- 1:26get just right. Little annoying, but we
- 1:28move on. And with the game's main
- 1:30gameplay loop roughly done, we now need
- 1:32something for the players to work
- 1:33towards long term. So, I started with
- 1:35upgrades. Players can earn cash to
- 1:37upgrade their movement and win more
- 1:38rounds. [music] And then I added tool
- 1:40crates so players can gamble, of course.
- 1:42They can unlock tools, which make
- 1:43gameplay more fun and less repetitive.
- 1:45And this was the point where I just
- 1:47grinded out tools, okay? We got guns. We
- 1:50got swords. We got bananas. We got
- 1:53freaking jetpacks. And all of that made
- 1:55it so that now we have a way for players
- 1:57to progress in the game that's fun and
- 1:58exciting. And after this, I focused on
- 2:00the lobby, adding a spinner wheel for
- 2:02more gambling, a leaderboard for
- 2:03grinding, and an obby, which later I
- 2:05removed and turned into a practice area.
- 2:08And of course, I added some
- 2:09monetization, you know, some game passes
- 2:11and some developer products. But yeah,
- 2:14obviously I'm summarizing and skipping
- 2:16over a lot of development. Honestly,
- 2:17development took about a month, but I
- 2:19did this while I was still in college
- 2:20dealing with classes and also working on
- 2:22Manhunt, so it was for sure a grind. But
- 2:25I honestly had a bunch of confidence in
- 2:27this game idea cuz I personally found it
- 2:29[music] very fun. But first, if you've
- 2:30ever tried making your own 3D assets for
- 2:32a game, you know how much time it can
- 2:33take to model and texture everything
- 2:35from scratch. That's why I've been using
- 2:37Meshie, an AI tool that lets you turn an
- 2:39idea or even just an image into a
- 2:41complete 3D model in minutes. And the
- 2:43process is actually really simple.
- 2:44First, I give Meshie a text prompt or
- 2:46upload an image of the asset that I
- 2:47want, then Meshie generates a 3D model,
- 2:50and once I'm happy with it, I can export
- 2:51it directly into tools like Blender,
- 2:53Unity, or Unreal. For example, I can
- 2:55take this 2D concept art, upload it
- 2:57using Meshie's image to 3D, and Meshie
- 2:59turns it into an actual 3D asset. One of
- 3:01the biggest improvements with Meshie 7
- 3:03is how accurately the geometry and
- 3:05textures match the original image.
- 3:06Details, patterns, and the overall art
- 3:08style are preserved much better across
- 3:10the model, even when you rotate around
- 3:12and look at it from different angles. It
- 3:13also has smart topology, which
- 3:14automatically separates the model into
- 3:16clean individual parts and lets you
- 3:18control the poly count depending on what
- 3:20you need for your game. And if you're
- 3:21trying to create an entire set of
- 3:22assets, Meshie's 3D agent can lock in
- 3:25one art style and match it across
- 3:26different props, characters, and
- 3:27environments. So everything actually
- 3:29looks like it belongs in the same game.
- 3:30So instead of spending hours creating
- 3:32every asset completely from scratch, you
- 3:34can go from an idea or a 2D image to
- 3:36something you can actually start using
- 3:38in your game incredibly fast. If you've
- 3:40got an idea, mesh it now and get 50% off
- 3:42using my link in the description. Huge
- 3:44thank you to Meshie for sponsoring this
- 3:46video. Let's get back to it. And another
- 3:47thing I didn't mention is that
- 3:48throughout development, I was doing a
- 3:50ton of play testing with fans from my
- 3:52Discord servers. This is super
- 3:53important, especially if you're a solo
- 3:55developer. So, make sure that at every
- 3:57step in development, even when the game
- 3:59is at its most basic, unpolished stage,
- 4:01you are getting feedback from your
- 4:02friends and other developers. So, with
- 4:04the spring semester coming to an end,
- 4:06and the game done for the most part, it
- 4:09was time to start ads. I put 280 ad
- 4:12credits in to start for what I think was
- 4:147 days. And as the ads ran, I did not
- 4:16stop improving, okay? Every day, I would
- 4:19get a list of things I needed to
- 4:20improve, and I would try my best to
- 4:22improve [music] it. You cannot release a
- 4:24game and just expect it to take off on
- 4:26its own, right? You need to constantly
- 4:28be improving it. And the game was doing
- 4:30okay. We were averaging about 50
- 4:32concurrent users, and the metrics were
- 4:35not the best, but I was still hopeful,
- 4:37okay? Because everybody knows your
- 4:39players are just going to be worse in
- 4:40that initial stage, because they don't
- 4:42know your game, and they're all just
- 4:43joining from an ad. And on top of that,
- 4:45the dislike ratio was pretty insulting.
- 4:47But, I was making sure to constantly
- 4:49listen to player feedback and just keep
- 4:51pushing out update after update, no
- 4:53matter what. And then, the first week of
- 4:55ads ran out. We were still barely
- 4:57getting any home wrecks, which,
- 4:58honestly, was really scary. So, I
- 5:00decided to run two more campaigns over
- 5:02the next week, in hopes it would finally
- 5:04catch the algorithm. It didn't, not at
- 5:06all.
- 5:07>> [music]
- 5:07>> And it had now been 2 weeks. The home
- 5:09wrecks were low, and I was getting very
- 5:11seriously worried. But, I had been here
- 5:12before, okay? If you watched my Manhunt
- 5:14video, it did the exact same thing,
- 5:16where I thought it was a total flop
- 5:18right before it slowly took off. So, I
- 5:20knew I had to take a risk and turn the
- 5:23ads off. When I turned them off, the CCU
- 5:25dropped. But, day after day, the metrics
- 5:27slowly climbed. And when the metrics
- 5:28grew, the CCU followed. And the game
- 5:30began to blow up quickly. Home wrecks
- 5:32were coming in higher every day. CCU was
- 5:34booming. The weekend came, and the game
- 5:36peaked at 900 concurrent users. And with
- 5:39this increase in players, came some
- 5:41offers. I'm not going to name any
- 5:42studios in specific, but one studio
- 5:45ended up offering me $24,000
- 5:48for 70% of the game, and they would
- 5:51handle everything for me. I wouldn't
- 5:53have to keep putting in hours working on
- 5:55the game. This was a lot of money, and
- 5:57considering the metrics were very mid,
- 5:59it might just be the best option. It
- 6:01would allow me to work on more projects,
- 6:02and I had a very hard decision to make,
- 6:04which I was seriously very indecisive
- 6:06about for days. I was going back and
- 6:08forth, you know, do I take the money and
- 6:09work on something else, or do I double
- 6:11down on the project and go all in?
- 6:14I decided to not take it, okay? I
- 6:16decided to bet on myself and the game
- 6:19that I've been working hard on. I turned
- 6:20down that offer. I told them no. I put
- 6:22together a plan to grow the game
- 6:24further, and over the next few days, the
- 6:26game the game dropped, yeah. Its CCU was
- 6:29just tanking, okay? All of our momentum
- 6:31had stopped. The metrics were looking
- 6:33worse, and not going to lie, it was
- 6:35definitely at this point where I felt
- 6:37horrible for not taking that money and
- 6:39running. For a solid week, I thought
- 6:41that I had lost over $20,000. We were
- 6:44down to 300 CCU, and the metrics were
- 6:46looking horrible. But, I had made some
- 6:48pretty significant changes to the game,
- 6:50and at the end of the week, the game
- 6:52just started taking off, all right? It
- 6:53just started killing it. Almost out of
- 6:55nowhere, it gained that momentum again,
- 6:57and it took off. And me, I was quick to
- 7:00feed that fire, adding a battle pass,
- 7:02clans, more tools, more cosmetics, and
- 7:05the game continued growing. It's been
- 7:07about 2 months now since the game
- 7:09released, and I couldn't be happier that
- 7:11I did not sell. Right now, the game has
- 7:135,000 concurrent users, and it's a
- 7:16freaking Wednesday night, dude. Insane
- 7:18insane stuff, and it's been doing that
- 7:20for weeks now. Last weekend, we peaked
- 7:23at over 7,000 concurrent users, and the
- 7:25game just hit 20 million visits. It has
- 7:28honestly been a crazy journey so far,
- 7:30but honestly, I have loved working on
- 7:32the game, and it's been an absolute
- 7:33dream come true. And on top of that,
- 7:35I've been learning a ton. So, take this
- 7:37as inspiration that it can be done as a
- 7:39solo developer. If you put the work in,
- 7:41it will be worth it. I promise you. And
- 7:44if there's one thing that you take away
- 7:45from this, know that you need to keep
- 7:47learning and improving your game. Find
- 7:49an idea, prove it works by play testing,
- 7:51and do not give up on it. If you want to
- 7:52check out Dodger Dot, link will be in
- 7:54the description of this video. Thank you
- 7:55so much for watching. Goodbye.
About this transcript
This page contains the full transcript of I Made a Successful Roblox Game to Prove It’s Not Luck by Zyonic, generated from the public captions YouTube serves with the video. The transcript has 1,806 words across 251 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.