Hello,
To introduce myself I am an older game developer that has come from doing mobile phone games in the unity engine to explore roblox this year. The platform has a lot of nice features and is attractive to me as a developer.
Unfortunately, I am having a very difficult time understanding discoverability and advertising on the platform and I would like to highlight the performance of two of my projects as an example of this:
Project #1: Bus Driver Vietnam
This was just a quick idea I had whilst I was learning how to develop on roblox. I made it over 5 days.
ADS: I have not bought ads
UPDATES: I have never updated the game.
LIKES: It has 800k plays with 34 likes and 33 dislikes, a 50% approval rating and 0.00004% of people like the game.
FAVOURITES: It has 340 favourites, 0.0004% of players favourite the game.
Project #2: Civilisation Explorer
This is a remake of a popular mobile game I made, spent a month developing it and it is the culmination of all my time spent learning roblox.,
ADS: $10/day consistent ad spend for 6 weeks
UPDATES: I have updated the game with large content updates every week for the last 6 weeks
LIKES: It has 540k plays with 2400 likes and 600 dislikes, a 80% approval rating and 0.004% of people like the game. (100x better than game #1)
FAVOURITES: It has 6800 favourites, 0.01% of players favourite the game. (200x better than game #1)
A detailed comparison of the stats:
So my question here is why does game #1 consistently have more players than game 2? Why does roblox give 10x more home recommendations to game #1? How can a game with worse stats, no ads and no updates, be getting more players than a game with better stats, weekly updates and a fair ad spend?
I decided to try out a $50/day ad spend over this weekend to see what happened. As I write this, game #1 has 95 players (NO AD SPEND) while game #2 has 55 players ($50 / DAY AD SPEND).
I’m not writing this post to advertise the games but for what its worth, here are the links to the games:
Game #1: [BDVN] Bus Driver Vietnam 3 [NEW CARS] - Roblox
Game #2: Civilisation Explorer 🛠️[Combat Update] - Roblox
2 Likes
I will also add that the recommendations for game #2 have just generally behaved weirdly.
You can see on the graphs I posted above that after release, roblox picked up on game #2 and recommendations were rocketing. It made sense - the game had good stats and roblox was therefore recommending it. We kept getting more users and we reached a peak of 700 CCU.
Then, suddenly, with no apparent change in anything else, on 1st June 2024 roblox just stopped recommending the game as fast as they had started. Now we’re left at the current point where it underperforms even game #1. What happened on June 1st? I have no idea
Roblox’s recommendation system is very weird but to explain it simply the more weird and oriented the game is to kids and the more monetization, the more it gets recommended it’s been a pretty big problem for a while and your likes and favourites dont matter much
1 Like
Thanks for the reply, but this doesn’t really answer what is going on here. If its just about more monetization, we can compare the two games and see that game #2 has >1000x better monetization performance than game #1, yet it is game #1 which roblox promotes!
Is designing a game to meet the metrics that roblox highlights just a waste of time?
I have reported this as a bug also but really hoping somebody from roblox is going to have a comment for me on this. Please, give me some suggestion of how or why game 1 can be promoted over game 2 so I can understand??!
Very interesting to see the difference in the two and the factor of “luck” that’s still involved with ROBLOX’s recomendation system.
Honestly < 100 ccu doesn’t feel like any feeds or “the algorithm” has really picked up either game, and the spikes you’re seeing might just be the system checking your game out and see what players think of it.
Like @MrChickenRocket says it looks like the algorithm isn’t picking up either game much at all, just checking them out to see if they should get boosted or not - and to me personally it looks like it’s because of monetization stats. If you’re not at least within the benchmark for monetization you will (sadly) not get any real traffic from home rec, only from other places such as organic (friends etc) discovery.
I also believe running sponsored ads may be netting you much lower quality players - ones that play less, pay less, or both. We’ve seen this consistently while running sponsored ads for our games and have a running theory that it may be adversely affecting the statistics each time we do. After no longer running sponsored ads our game actually seems to have a higher player count and better discovery.
If you want to spend some time reading posts from top developers on the subject, and their frustrations, this thread may interest you: Roblox’s Algorithm Changes Are Killing Our Game
2 Likes
Thanks for those replies and link to the other thread. I think the spike we saw was the algorithm pickup as I’ve noticed on other projects, roblox will start recommending them about 2 weeks after publish.
My main takeaway so far is that roblox does not actually care about or weigh in many of the metrics that they encourage developers to watch. Monetization is king and it appears they want us to churn out multiple low-quality games with heavy monetization over doing larger projects, so that is what I will try next!
1 Like
