Speaker A
Roblox is actually easy and by the end of the video you will see why. You made a game and nobody came and that sucks.
A Roblox dev explains why small games get no players: ads, likes, and clickbait don't drive the algorithm; first-minute retention and playthrough rate do. Covers the explore/expand loop.
Key Takeaways
- Ads, likes, and clickbait do not improve Roblox recommendation ranking; engagement signals do.
- The first minutes of a game determine whether new players stay, which drives the algorithm.
- Small games can compete because metrics are averaged per user, so quality of experience matters more than player count.
What the video covers
- Myth 1: Running ads will boost ranking. Roblox states ads don't affect recommended-for-you ranking, and players from ads, friends, search, or curation aren't counted.
- Myth 2: Begging for likes helps. Likes are not among Roblox's top recommendation signals; playthrough rate, first-play bounce, play days, and playtime are.
- Myth 3: Clickbait brings players. Qualified playthrough rate is designed to penalize clickbait, so immediate leavers hurt the game.
- Myth 4: Small games can't compete. Signals are calculated as per-user averages, so a small game with engaged players can rank well.
- Home page drives over 90% of Roblox traffic, so getting recommended is essential.
- The recommendation loop works as explore (small test group) then expand (wider audience) based on how players behave.
- Roblox measures first-play bounce in two windows: under 60 seconds and 61-180 seconds. New players decide within minutes, so the first minutes must be strong.
- Play time is capped at 60 minutes per user per game per day, so AFK time doesn't help.
- Since June 2026, Roblox evaluates player behavior over 28 days instead of 7, covering day 1, days 2-7, and days 8-28.
- Practical advice: make the game simple, deliver fun within the first seconds, and use the sibling/mom test to check enjoyment before wider release.
Chapters
- 00:00Introduction: Why Roblox Games Struggle
- 00:56Myths About Ads, Likes, and Clickbait
- 01:54Why Small Games Can Compete
- 02:51How the Homepage Recommendation Loop Works
- 04:57Roblox's Top Ranking Signals
- 05:57Explore and Expand Phases and Play Time Caps
- 07:57Winning the First Minutes
- 09:48Thumbnails, Testing, and Action Plan
Full Transcript — Download SRT & Markdown
Speaker A
You spent months on this game and it's not hard. You're just fixing the wrong thing. Most devs think that getting players is luck, money, or a rigged algorithm. And that belief is the reason why they're not making it. It's the
Speaker A
reason why you're not making it. And why listen to me? I built a site that tells you why your game has no players. And I've had over 1,300 devs in the past two months check their game with my website.
Speaker A
And I've seen trends way before they get picked up in the Roblox algorithm. So here's the plan. First, the myths that make it feel hard. Then how home really picks games. Then the one thing to fix first, then how to check it worked. And
Speaker A
if you have a small game and almost no players, then this video is for you. So chapter one, we're going to go over why it feels hard. So, the first myth is to run ads and the algorithm will boost
Speaker A
you. You see, a lot of devs spend their Roblox on ads and wait for home to kick in. But Roblox said it themselves that ads don't change your recommended for you ranking. And it goes further.
Speaker A
Players from ads, friends, search, or curation aren't counted towards ranking at all. So, the second myth is to beg for likes. Likes aren't on Roblox's official list of top signals.
Speaker A
playthrough, bounce, play days, and play time are the signals to actually look out for, not likes. And myth three, clickbait gets you more players. Roblox built qualified playthrough rate to get rid of clickbait on the platform, so people who click on your game and leave
Speaker A
immediately actually hurts you more than it helps. And this is the fourth myth that makes a lot of people quit. And that's that small games cannot compete.
Speaker A
Roblox measures signals as averages per players, not totals. So, even a small number of players on your game can send a signal to Roblox and you won't have any major issues competing against the bigger games as long as your game is
Speaker A
good. And actually, here's proof from small games. You can see these games like Roblox is Unbreakable, Defend Your Plushy, and Draw a Flag going over 200% increase this week because they're able to keep their players. So, if it's not
Speaker A
ads, if it's not clickbait, if it's not begging for likes, how is your Roblox game going to get picked up by the algorithm? All right, so here's why this matters so much. According to Roblox, the homepage drives over 90% of traffic
Speaker A
to the platform. So, if the homepage does not show your Roblox game, almost nobody will find it. So, let's start with how the homepage actually decides what to show. Roblox shows your game to some players, then it watches what those
Speaker A
players do. And if those players have a good time, Roblox shows your game to more people. And that's the whole loop.
Speaker A
And everything else in this video is about winning it. You see, home works like a chain. Someone sees your game, they click, and then they stay past the first minute, and they come back the other days. It's really simple. And a
Speaker A
chain is only as strong as his weakest link. Almost every small game, the weak link is always the same thing. And it's the first minutes of your game. The first minutes of your game has to be so good. And guys, I just want to say as I
Speaker A
continue this video, the only way you're going to win is if your game is good.
Speaker A
And it's a lot easier said than done. So, I'm definitely going to explain to you how a game can be good. in this video and future videos as well just so I can educate everyone. I can't necessarily personally look at your
Speaker A
game. Um, you can check out Game Doctor on my website and you can check that for sure. But one thing I want to talk about before that is what Roblox watches.
Speaker A
Roblox lists the top signals it uses for recommendation and there are four big ones. There's playthrough rate, which is when people actually see your game, click and play it. First play bounce rate is when new players leave really
Speaker A
quickly. Play days per user is whether they come back on other days. and playtime per user is how long they stay.
Speaker A
Here's the thing that should make every small dev happy. The data is calculated as averages per user, not total value.
Speaker A
Which means that a game that's bigger won't just destroy your game simply because they have more players. If your 50 players in your game love your game more than someone else's 50,000, that counts for the Roblox algorithm, and
Speaker A
you're definitely going to shoot to the top. So, Roblox says its recommendation system uses explore and expand phases.
Speaker A
So, in the explore phase, your game gets tested on a like really small group of players. And if those players really like your game, it moves into expand and more people will see your game. So, the first players that Roblox sends your
Speaker A
game, like the first players they send your game to, they're just like a little test to see if your game is worthy or of just getting pushed to the algorithm.
Speaker A
And you have to make sure like 100% that those people pass that test. And this is one thing I want you guys to remember, like even a small number of people playing your game can send a signal to
Speaker A
Roblox. So, you have to prioritize the enjoyment and the happiness of those people who are playing your game no matter what. Like, you don't need a huge player count. You just need to have the people playing your game have a good
Speaker A
time. In order to do that, one way I like to do it is just imagine you give your little brother or your little like just your little sibling your game to see if they'll enjoy it and have fun and
Speaker A
then try to give it to your mom. If they don't have fun playing your game, other people won't have fun playing your game for the most part. Like you have to make your game so simple, as simple as
Speaker A
possible and as enjoyable as possible for everyone, right? Like how Tetris is or Mario Kart or Minecraft. Now, one thing that people usually miss is that play time is capped. You see, Roblox caps play time to 60 minutes per user
Speaker A
per game. So, per day. So, if someone's sitting in your game for more than 60 minutes AFK, it's not going to help you.
Speaker A
So, don't even bother doing that. And the whole point of the algorithm is so people playing your game have a good time, not just standing around doing nothing. And one thing that you guys need to know about, earlier this year in
Speaker A
June 2026, Roblox changed its recommended for you algorithm from looking at seven days to looking at 28 days. So now it tracks players on day 1, days 2 to 7, days 8 to 28, pretty much for the whole month, and that covers
Speaker A
things like play time, how many days people play, how often they come back. So the next thing I want to talk about, and arguably one of the most important topics in the Roblox algorithm space, is winning the first minutes. And this is
Speaker A
pretty much why I made this video in the first place. So, Roblox measures first play bounce rate in two windows. So, the first window are players who leave in under 60 seconds. And this is going to hurt you if they leave real fast. And
Speaker A
the second window are players who leave between 61 and 180 seconds. So, one minutes two and three. Roblox said this themselves that new players typically decide within minutes whether or not they like your game. Not hours, not days. So, if your content or your game
Speaker A
has like a lot of content, like 10, 20 minutes within like the actual game, it's over for you, sir. Everything amazing has to be right at the beginning. I do not care what you think.
Speaker A
That's just how it is. And you can't decide what you want the algorithm to be. You have to play the game or just quit. And this connects to a number you'll see in like your analytics page called playthrough rate. So, Roblox
Speaker A
defines this pretty much as the rate which people play your game after seeing it in the recommended for you. So, in simple terms, did they see your game?
Speaker A
And then did they go in and actually play that game? And then Roblox will go ahead and check to see that if those players who joined your game leave within the first few minutes. And the reason why Roblox made it this way is to
Speaker A
kill clickbait. If your game promises one thing in the icon in the title and people join and get a completely different experience, that is just going to kill your game. It's going to hurt you and you just wasted your time vibe
Speaker A
coding an entire game for a month. You have to deliver on the promise of the icon in the title. As a matter of fact, not even deliver, but exceed expectations. So, not only does your icon have to be flashy, but when people
Speaker A
join your game, the game has to be just as flashy as the icon. So, think of it as two jobs. Your thumbnail and title have to get the click, but then your game itself in the first minute have to
Speaker A
keep the player. So, this is how you fix your first minute. So, step one is very important, and I've stressed this a lot, but you have to get to the fun quickly.
Speaker A
a lot of games, and I've done this test myself. I've joined a lot of the top Roblox games, and you get to having fun within the first five seconds. Not the first minute, not the first 30 seconds, but within the first 5 seconds, you're
Speaker A
already doing something. How do you actually do that? So, join your own game as like a new brand new player, brand new account, and time how long it takes to do something fun. So, cut any long intros. And I've seen new Roblox games,
Speaker A
not the top ones, but the ones that want to blow up, they've been putting long animated intros at the beginning. Cut that out. I'mma be for real. Nobody wants to see that. Walls of text. Nobody wants to read their kids playing your
Speaker A
game. Let's be for real. They don't even read at school. Why would they read playing your game? And everything that's fun, move it to the front. Now, the second step is to make sure players know what to do and why they're doing it.
Speaker A
Players have to understand both. If someone's confused as to what to do and what that thing that they're doing gets them, it's over for you. You have to make it as simple as possible that they spawn in. They know what they're doing
Speaker A
and why they're doing it. For example, if they're walking to fill up a meter to walk even faster. That's so simple. So, when someone joins your game, a a new player should be able to answer these two questions, which is first of all,
Speaker A
what do I do? And the second thing, why should I? And step three is one that a lot of people skip. Actually, Roblox says a custom thumbnail is required.
Speaker A
It's absolutely needed for your game to get impressions from the homepage. If you're still using the default thumbnail, your your game isn't going to be shown on the the homepage. episode, it's over. And you can use my website. I
Speaker A
have an AI icon generator if you want to check it out. It works pretty nice. Um there I mean there are other ones as well out there if you want to try them.
Speaker A
All right. You can also test thumbnails. Don't stop on just one and think it's okay. Uh Roblox will allow you to upload up to five thumbnails and then Roblox shows them to different people and finds out which one works. This is not a new
Speaker A
system. YouTube does this as well where you can do AB testing on uh thumbnails and titles and Roblox will just test your thumbnails to see which ones work for the players. But the thing is you have to learn patience. Now I'm not
Speaker A
really a patient personally. I need everything now. But unfortunately we are going to have to learn patience. And every time you change your thumbnail or have a variation, give it a second where you have to wait seven days because if
Speaker A
you keep swapping every day thinking your game is going to blow up the next day, that's not how it works. And this has worked before. In 2024, Roblox reported that games using thumbnail personalization saw playthrough rates go up by 8 and a half% on average. Some
Speaker A
people even saw it go up 50%. And that's just letting Roblox pick the right thumbnail for each player. This is pretty huge. Take it as you will. Now, step five I want to talk about is kind of newer and a lot of devs, they
Speaker A
probably know about it by now, but people don't talk about it. It's uploading gameplay videos on the homepage. gameplay video saw up to a 39% more play time from the recommended for you. Now, the video itself doesn't change your ranking, but it helps you
Speaker A
get the right players to click and so the people who click will stay. I love this feature cuz I get to see what a game looks like before I join. So, if I join and leave, I don't hurt their game
Speaker A
and their stats and then the people who really want to join will join the game and play for a long time. It works amazing. So, let's do a real fast recap for the first five steps, which is have
Speaker A
put all the fun at the beginning of the game. Show players what to do and why.
Speaker A
Have a custom thumbnail and then test up to five thumbnails and let each one be live for 7 days. And also upload a gameplay video. Now, chapter five are small games that blew up. So, let's look at proof that small games can break out.
Speaker A
You've probably heard of Grow Garden. Even my grandma has. When the game first came out, it had about 500 players online. And then the developers started pushing updates every week. According to Games Beat, it went from around 500
Speaker A
players at once to a 100,000 players in about 2 weeks. And that and within a month, it was at a million players, which is crazy. And the thing that makes it even crazier is that the game was built by a 16-year-old in just a few
Speaker A
days. It just it wasn't made by a giant studio with a huge budget. It was just a simple game. But what we know is that this game started small from a single person and they consistently updated it.
Speaker A
And that's the lesson I want to teach you guys. You don't have to start big with a giant ambitious project. As a matter of fact, that's worse for you.
Speaker A
You're better off making a simple game and making it more complex as it goes on. Here's another game I want to talk about. Fish. It was one solo developer with four months of work. And the growth was mostly organic, meaning that they
Speaker A
didn't spend a lot of money on a lot of ads cuz it probably wasn't a lot of money. Having updates and just good quality gameplay helped it get pushed to the front page. So, if you have a small
Speaker A
game, this is where all your time should go. Have fun first at the beginning of the game within the first five seconds to the first five minutes because that's what's really important. And make sure your thumbnail is so good because
Speaker A
without it, people aren't going to click and you're not really going to get pushed. And then as your game grows, you start to test more and more thumbnails and that actually makes a big difference to how your game performs. And I also
Speaker A
want to be clear that there is no magic number and there is no shortcut to this.
Speaker A
Anyone who gives you an exact number is mostly guessing, mostly just throwing something up in the air or making an educated guess. But at the end of the day, as long as you have a good game, and it doesn't even matter if it's AI,
Speaker A
because AI is meant to help you, not make your entire game. So, if you guys are letting AI just do the whole thing, your game is going to be trash. Your game is going to be so bad. There's so
Speaker A
many bad AI claw generated games out there. People don't understand the fundamentals of game design, gameplay, attention. These are things that giant studios or experienced Roblox devs understand the best. So, this is your homework. If you don't know what is
Speaker A
making your game dead, you want to revive your game. Run your game through my game doctor if you'd like and you can see what's wrong and try to work on it.
Speaker A
This is not easy. Roblox is getting harder and harder, especially with more devs joining the scene and it's understandable. So, if you guys enjoy this video, I wish you luck. Keep going.
Speaker A
There is no magic number. There's no magic answer. All you have to do is just do research. You have AI. Do research on how to make your game as good as possible. See what's working and make sure you make a quality game. even if
Speaker A
you're using AI because you can still make a quality game using Claude.
Topics:RobloxRoblox algorithmgame developmentplaythrough rateretentionRoblox recommendationsindie game marketinggame designfirst play bounce
