In January 2014 Developer Stats were released and they were absolutely game changing. It has allowed developers to analyse players behaviour and track their sales identifying their popular products.
Since it’s release the stats page has gained many new features with the most recent allowing developers to filter statistics based on player’s ages.
What is my idea?
My idea is short and sweet: I propose an additional feature which displays the genre average for each section of the developer stats.
Why would this be beneficial?
Currently, there is no way to benchmark and compare the statistics of the performance of your game against a baseline average. Developers currently have no way of identifying whether their game is performing well compared to their “competition”.
e.g Having a 10 minute average play time for your game which is in the “Horror” category might seem very positive to a developer but if the genre average is actually 15 minutes then it allows the developer to see that their game isn’t performing as well as the rest in it’s genre.
I am well aware that this could possibly require large amounts of data processing by Roblox so providing the genre average in the "Hourly" or even "Daily" section might not be feasible. Perhaps if this statistic was updated once a week this would still prove to be very beneficial and provide a very good insight to developers.
Genre sort has been removed (a decsion I deeply, angrily disagree with), but genres still exist.
While this would be interesting to see, I don’t think it’s a useful stat.
If anything, in the likely case for most young developers just getting started, it would be depressing to see this on the developer stats page, since it suggests their work is “below average”.
You also have the problem of figuring out where the cutoff is for collecting valuable data versus outlying data. There are millions of games with zero players, and thousands with a very small playerbase, a few huge games, and some very specialized games that serve no other purpose than to show you something quickly, or redirect you to another game - therefore having very low playtime.
In addition to the UX issues, this sounds expensive and difficult, and not worth the trouble.
It would perhaps be better if Roblox released “data dumps” - monthly or yearly on the blog for example, that we could compare our stats against if we wanted to.
The genre sort isn’t really being supported since the removal of it on the actual web page. It’s important to shift towards measuring statistics on a grander scale (I.E the top 150-250 games) which will give you more accurate results as to how your game is performing overall.
I think it was wise of Roblox to realize their extremely limited and non scale-able genre model wasn’t going to last. However I think it is very unwise to not replace it with anything. User generated tags based on a few words the developer hand picks (to then be voted on by the community) could be a good way of doing things. Then players in the game page can search though the top tags, and just scroll through tags and tags (based on popularity) in a fluid way like the games page is fluid.
Therefore, I see no reason to support a benchmark stat such as the suggested idea.
Having a replacement for genres is sort of a side issue.
My focus is on being able to compare my own statistics I receive for my games against an average to be able to see if my games are performing well or not.
It sustains an audience for an extended amount of time without needing constant advertising
It makes more profited Robux from advertising than if you didn’t
The Robux:Visit ratio combined with the visits it gets is enough to justify the work you put into it, or the thoughts of future earnings after an update keeps you working on it
The avg visit time works together nicely with the visits per hour and increases your concurrent players to a place to gather enough new players daily for the next day returning users to continue the cycle
I think this idea is looking into the wrong place for how to improve one’s own game
