Your estimates are wildly incorrect and overestimating the percentage of bots.
5 Likes
It was just a guess. But look at Pets Go and Pet Sim 99’s server. It’s a bot infestation!
4 Likes
You can already find investigations into these bots. Several programs have been created to analyze the percentage of bots by identifying them through their avatars—specifically, avatars that share the same URL within the server list. (Of course this is still not fully accurate from comparing bots with avatars that has not been changed)
2 Likes
The replyee modified his post to reflect mine.
If you’re a bot, you’d want to avoid changing your look. That takes up some effort.
New Rōblox players might also not change their avatars for a while until they become more comfortable with the controls.
You’ll probably get inflated results that way.
2 Likes
I don’t see the value of anything in pets go decreasing, which is the game that should be most affected
1 Like
The reason people are overestimating the number of bots in a given game is due to how the game server page is designed. It works as follows:
-
The default list of servers is by number of players, and then by age of the server. This makes full + old servers rise to the top by default.
-
The 5 avatar images shown are displayed based on the amount of time that account has been in that server. This makes it so accounts that have been in the server for the longest period of time will be the 5 shown.
So when you look at a game which does have a lot of bots (e…g Pet Sim 99) you see images like this:
100% bots, agreed. But that’s because:
- Bots join servers and stay in-game for long periods of time. This ensures the bots rise to the top of the 5 bacon headshot list.
- When real players join those servers, they leave, because too many bots. You want to know who stays in a server with too many bots already? More bots.
- Those server instances never go away because they’re at full capacity running 24x7 bots.
The game server list is a prime example of survivorship bias. Are there lots of bots? Absolutely. But due to simply the order in which the servers + avatars are displayed, it drastically overstates the problem and people think “omg 100% bots”.
15 Likes
The reply was modified before you posted yours thanks for the note you had to add on yours
2 Likes
( Im not sure how completly accuutre this is )
On websites that track the values of cosmic items in Pet Simulator 99, you can see the value rates dropping after the botting has started to grown in the graphs they have posted.
4 Likes
Don’t they already have market saturation by this point? I can’t think of any reasonable reason (har har) for them to do this.
2 Likes
It doesn’t matter if some ppl expressed distaste when the ID verification thing was first introduced to Roblox. This isn’t a real issue anymore, it’s been a crucial part of the platform for years, as it is being used to separate games with 17+ content and to determine voice chat eligibility I haven’t seen anyone suggest this was a major issue with getting voice chat. Also, even if it was true that this is a real and large issue, community distaste towards ID verification doesn’t impact the performance of the metric so why would that matter? The metric would still function effectively in its purpose.
Second, the community already has no choice of what content is promoted because right now it’s being swayed by bots and account farming. If you aren’t suggesting any other solution are you saying we should continue letting bots/farming decide what games get promoted? (non-constructive) Also, this does not have any impact on the choice of content each user gets recommended as you suggested. Roblox has some a way of matching specific types of games to players (which is failing too imo based on my recommendations but that’s a separate issue) this metric would be part of Roblox’s main ranking algorithm which determines which games are promoted within each specific genre/subgenre; the same area where the issue mentioned in the original post exists in most and is impacting most. So yeah it would have no impact on people’s choice of content (what types of games get shown to who), and that’s a separate issue.
Third, it’s the other way around for hidden gem games. It would only make sense that the games currently supported by small but strong communities get a large advantage because their ratios (ratio per player) would be substantially higher. They’d outperform substantially in this benchmark, and the games that currently have super high user counts but aren’t actually even good games would suffer most in this metric. Having it so that games are not able to detect if a user hits the buzzer or have text everywhere telling them to hit the buzzer would make it harder for bad games to succeed in this metric. In the end the bad games are the ones that need to ask for it but the good games don’t need to ask for it. Therefore, having this restriction benefits hidden gems and it would actually harm smaller games to have it the other way around. If it were the other way around dopamine harvesting games that put text all over your screen would end up getting more buzzer hits than they deserve from a younger audience that doesn’t use it sparingly as @Vamizy suggested. Also it’s not at all hard for that to be implemented correctly, it’s just a percentage (you divide by the number of players) how would that be hard for roblox?
Roblox’s current ranking algorithm focuses on a game’s metrics for session time, retention, monetization, and thumbnail click rate. These can all be impacted by things like bots/farming accounts, pay2win, gamepass-spam, and clickbait. The metric I suggested is an effective way of recognizing quality content while being immune to those issues. If it were added as a 7th metric alongside the other 6 it should carry more weight than the other 6.
6 Likes
Hindenburg’s quotes from former employees show that the company has a growth-above-all approach, where anything that increases their CCU number is a good thing.
This is why every RDC opens with them displaying their ‘record high’ CCU growth all within a minute of starting the presentation
RDC 2023 @ 51 Seconds
RDC 2024 @ 53 Seconds
This is what they prioritize above all else, anything that makes this number go up whether it’s botted or not.
The consequence of this is, of course, that normal players are now forced to play with bots and cheaters. Roblox has a competent anti-cheat that detects every single one of them yet they aren’t allowed to clean things up. Which is why I question what these new AI models will even change. To me it just seems like more PR when they’ve had all of 2024 to correct the increase in bots/cheats yet refused to.
7 Likes
I’ll believe it when I see it. We’ve had this type of conversation on the DevForum for three years now. The Ban API wasn’t worth its weight at all and ROBLOX seems more keen on MAU growth than fostering competitive, fair, mature games on the platform.
The funny part about this is tripling down on allowing developers to remove bad actors machines completely, having a zero tolerance approach to people banned for scripting (the three strikes you’re out policy currently in place isn’t working), and allowing developers to enact a verified phone policy on games would actually lead to more revenue - fun competitive experiences bring the dough.
6 Likes
I’m going to be honest, Hyperion isn’t very effective. The Ban API isn’t very effective either. I still see many exploiters in-game frequently.
6 Likes
Hyperion doesn’t do HWID bans.
1 Like
In the current situation, one of the most effective ways for developers to limit the influence of bots in the game is to verify the age of user accounts using AccountAge properties.
Another solution would be a system that identifies a user’s accessories when they log into the game. If a player only has standard accessories or their accessories are below a certain value, they could be considered for kicking the user.
While these methods are not guaranteed to be accurate or effective under certain conditions, they are viable in the absence of significant anti-cheat tools.
3 Likes
even not bots have accessories in bacon style. That peoples who dont wanna spend money on avatar, even roblox employe have bacon avatar.
Account age is main solution for this problem
4 Likes
preston after roblox bans 99% of his playerbase: 
3 Likes
This is wonderful news! However, I’m curious when you plan to address the ongoing issues on your platform, such as:
- The “Allow Strong Language” feature not working as intended.
- The frustration among 18+ Australian users who can’t access gambling features, despite the legal gambling age in Australia being 18. Meanwhile, in many parts of the United States, the legal gambling age is often 21 (though it varies by state), yet U.S. users seem to face fewer restrictions.
Additionally, different countries have varying legal gambling ages, for example:
- United Kingdom: 18
- Germany: 18
- France: 18
- Canada: 18 or 19 (depending on the province)
Or better yet: make it a parental control option that is automatically enabled by default giving users parent’s more control on what their children can see / do within the games they play.
Could you share any updates or a timeline for resolving these concerns?
2 Likes
so glad this is happening, the amount of cheaters still exploiting is massive - gotta respect roblox and their entire team - thought the ban api was fading.
seriously i do love this update and PRAY for more news like this!
4 Likes
3 Likes