Working out what causes dislikes is too difficult

[grilme99 (brooke)]

As a Roblox developer, it is currently too hard to determine what causes dislikes and, more specifically, who dislikes. There is no way to know which demographics actively dislike an experience, which makes it very difficult to target UX optimizations and reduce those dislikes.

It would be great if analytics in the Creator Hub had graphs showing the influx of likes and dislikes that can be broken down by demographic (age, gender, language, device type, etc.)

The graphs available on Rolimon’s (shown below) are a great example of what these graphs would ideally look like. The crucial part they are missing (that only Roblox can provide) is the accurate demographic breakdown. They aren’t very useful without knowing which groups specifically are leaving a dislike.

This feature request is motivated by our experience being part of The Classic event. Judging by social media, we were one of the most well-received games, but our like/dislike ratio has been tanked during the event (more than the other experiences). We have no way of knowing which groups are leaving those dislikes, so we can’t optimize the experience so that they can enjoy it more.

[signal_zzz]

Thanks for your feedback.

Would you prefer we prioritize showing how like/dislike is trending with ability to filter or giving users a way to leave qualitative feedback after they like/dislike?

[Korblox_Guardian (VOEW)]

I agree that this would be greatly beneficial in finding out which specific groups and audiences might not be enjoying the game. A work around I see a lot of people doing is leaving a “suggestions” tab in game. Though I would imagine this is somewhat bias in that its mostly the older audiences giving more valuable and constructive info compared to younger ones that might simply utilize it as a platform to complain without really stating the in-game cause of their frustrating.

Regarding your case in particular, I played your game during the event, and found it quite fun. Ranking it with my peers in difficulty though, it was one of the harder ones. The issue being that the game is skill based. Being that this is a platform wide event; your attracting players of all ages and player types, many of which don’t play faster and more competitive titles.

If they want to get the event prizes, they see themselves as being forced to play your game, regardless if they enjoy it or not. Because you would be thrown into a match with others random players, majority of which were new and simply there for the event, your chances of being successful were very low.

Your now getting’s waves of people who play games like Adopt Me, tycoons, and simulators, things that are slower pace and not reflex or skill based. Again, this is prominently because your intaking a audience which your game would never naturally attract, they are there just for the event.

To put it simply, your getting more dislikes because a lot of people coming for the event, are only there for the event, and are not the player types who would be interested in your game if they had not been incentivized by a outside source.

[grilme99 (brooke)]

@Korblox_Guardian

The issue being that the game is skill based. Being that this is a platform wide event; your attracting players of all ages and player types, many of which don’t play faster and more competitive titles.

Sorry, I think I should have specified in the original post that those ratings are from the Clip It game.

[grilme99 (brooke)]

I imagine qualitative feedback would be more helpful generally. If I had to pick between one or the other, then I would pick the option for rich feedback.

[NyrionDev (Nyrion)]

This one is kind of obvious. The dislikes occurred during the classic event. When joining the game for the first time, unless someone saw a tutorial, it was incredible hard for them to figure out how to change maps(since there were no in-game tutorials for it). This caused many players to not know how to progress in the classic event and leave a dislike to the game.

However I do agree with the point of the post, we need detailed demographics for the game likes and dislikes. Although I think the individual users that like/dislike the game should remain private for privacy reasons. Most specifically so code doesn’t force users to like by looking if they did or punish them for disliking.

[grilme99 (brooke)]

To be clear, this isn’t about the event. It prompted me to finally make this feature request, but this has been a need of ours for long before the event. I should probably change the wording of the OP.

[xChris_vC (Chris)]

Stumbled on this post when I was about to post a feature request. Our game has a rating of 62% in our first version, and we plan to release a major revamp.

It’s an important KPI to us (since it affects discovery/CTR), so for me a first priority would be to have a like/dislike graph with filters. This would help us understand who dislikes us (age, platform) and improve accordingly.

Dislike surveys with quantitative feedback (or a set of options) would be nice, but I think that’s more complex. It would also need a higher volume of responses to be accurate, since I feel many will just press an option and leave.

[vicy98 (Samarium_Helium)]

People are more likely to like/dislike than fill out a longer feedback form, so I personally would prefer it if filtering and viewing trends was prioritised.

[Reditect]

Adding on to that, I assume people are more likely to dislike due to a bad experience than to like from a good experience.

---

To the thread itself…

The issue of understanding why players dislike a game seems to have been a problem since 2015, probably even before that. I personally would love a huge reviews system comparable to Steam, but there are many questions to answer there. Restrictions (time limits before posting, word limit, age limit) and sorting of review capability come to mind. Bots too.

I don’t think Guilded or things off Roblox are the answer. We need something on Roblox, on the game page itself users can navigate to. And, that can’t be a dislike/like counter. As the OP notes, we (developers) need a way to get feedback from users about their dislikes. Or, I guess just understand them better. And, also to add, players need a way to find games they’d like to play. The like button isn’t exactly helping developers or the players. So why is there not a bigger system? I assume time and resources, maybe also demand.

On a larger note, players finding a game they want to play (or one they think they want to) can be better achieved with Filters & Tags in addition to a Reviews System. I feel as though solutions to these issues (feedback, search, and discovery) have largely been solved in the video game space. And, while analytics to developers can help to an extent… that is not the long-term solution (that would help the entire community -not just devs) that I hope for.

---

There are a myriad of threads on this forum about dislikes and a reviews system. Example: link to post w/ list of other posts… This thread is the only one I’ve seen with a staff response to it. Respectfully, I hope that goes further than a single reply.

[focasds]

I also want to contribute to this by saying that…

As a Roblox developer, I find it challenging to identify the issues with my game and its assets. Pinpointing the problems that have persisted for years feels impossible. I am not seeking help to diagnose these issues here, but rather proposing a feature that could really benefit everyone if implemented correctly.

The analytics provided by Roblox are excellent for understanding game activity, but they don’t really say anything to the average developer. They are just plain numbers and lines going up and down, as well as being compared to games that do not fit within your genre or “similar experience” which to my understanding are games that my users play the most.

The stats I am referring to are Retention/Engagement/Onboarding/Funnels/CustomEvents and you know the rest.

The group wall… won’t even talk about it, with votes happening more often than people talking in the group wall. I want to hear “everyone” not just two users who complained about dying in the game.

---

Anyways, enough talking, the feature I want is for Roblox to allow users to leave a Review under the game once they decide to up-vote or down-vote the game. They can say no if they wish, but I have a feeling a lot will indeed leave a review. I would also opt to private the comments, but everyone I guess has the right to see it, so that’s left for you all to decide.

With the beautiful AI stuff that has been introduced lately; I would love for Roblox to also auto-summarize them and let the developer/user know what they can improve. Which is better than some random insight telling me that my retention is at 6-month height and then telling me the next day to improve the retention, excuse me... If the comment section is public, this AI must filter out spams because imo it can have the ability to do so once it sees someone advertising Robux or a bunch of comments unrelated to the game that would not go well in a summary.

Just like this beautiful Amazon example for a UPS:

---

If Roblox is able to address this issue, it would improve my development experience because I would be able to target those problems that have persisted for years and know what’s actually wrong, as well as pushing updates more often rather than focusing on this retention issue.

At this moment, Like/Dislike doesn’t really say anything about why the game is good or bad, and having the comment section would truly reflect why the ratio is like it is.

Thank you!

[Maxwell_Edison]

I think a “qualitative feedback” system, while useful sounding, is something developers likely could implement in-game as easily, if not easier, than roblox could do itself (Since most games are so different, you’d potentially need to direct that feedback in some way anyways)

But in general, knowing analytics about likes/dislikes, such as age group and platform, itself would be so useful and hopefully not the most complex thing to implement on roblox’s end. Ideally it’d be a page on the analytics pages, and function similarly to how current analytics do.

This is important because it would help identify if certain devices are having trouble on a game as well - You might per say, not have many mobile phone players, but seeing that mobile phone players are a high ratio of dislikes let’s you definitively know that you’ve got something wrong with your game’s mobile accessibility.

Personally unlike what was said by some others here, I do not desire a full review system. Roblox is mostly children, and I don’t believe you’d be able to strike a good balance of ‘Real reviews’ and “Literal spam” - since let’s be honest, roblox fails at keeping spam off of comments like group walls already.
Developers would literally need a way to hide spam reviews… which… would be bad.
But if not, people would just use it as marketting/spam. Reviews are a no-go for me.

---

On an tangential note, does anyone else get weirded out when very bad games have very high upvote ratio? I feel like “Real” and “Good” games usually have lower ratios than pure spam.

Does Roblox potentially have a spam/botting problem for likes, or is there something else at play here?