Yes, I know that moderators would not be the people looking at the source code. I just used the word moderators, as it can have multiple meanings.
I know that ROBLOX is allowed to inspect the contents of the game, but I am just saying that for a reason like not putting down the correct percentage of a crate, I would not be happy.
However, it is just my opinion.
I donāt think this rule applies to a store that has randomized stock. The announcement says devs must include the odds when they buy an item. The store isnāt an item like a lootbox that, when bought, will give a random item, so your store should be fine.
This is really good for users knowing what they are really buying, but what if the developer, or owner doesnāt even know what the chances are? Like what if the chances are random, and that is random, just if they put a lot of randoms to the point where it would be almost unknown, would they have to redo it, or what? Not that Iād do it, I just was wondering if anybody was curious.
Previews are super helpful to me, as a player. Iām not going to buy something if I donāt like any of the potential rewards, so itās nice when developers choose to show all of them.
Then you say that itās 100% random (see Murder Mystery 2 for example)
Yea, the announcement says you must show the odds for all lootboxes, including ones that can be bought with an in-game currency that can be bought for Robux.
Iām glad ROBLOX supplied guidelines towards random items. Itās a constant issue where users could be mislead on the possibility of items that they could receive.
Is there anyway this could actually be monitored tho? Creators could theoretically just write random numbers and users wouldnāt be able to tell if it is accurate or not.
To be honest, you probably could lie but at some point someone is gonna find out and it will be a huge disappointment for you and your game. The community wouldnāt trust your game anymore. This is worse for games with larger audiences.
Whether itās a Roblox moderator who finds out or some Robloxain testing the odds, I wouldnāt lie about the odds one bit.
Not saying I would lie. Anytime Iāve done lootboxes itās basically 1/#items for every item of getting. But just pointing out that a player would not be able to detect this, nor do I think it would ever be questioned.
If someone said 50% chance for this item, Iād take their word for it, and if I never got it, then the odds just arenāt in my favour.
This seems pretty good, but some developers might provide inaccurate or misleading information to their players.
Yeah, itās pretty hard to guess whether the odds are true based off of a few spins. However if someone does spin the thing 10 times and never won the item they wanted (50% odds for getting the item). I know I would definitely be very suspicious about the odds. Which could lead me to report the game.
What if your system that gives rare items doesnāt run off of %s numbers, but rather, a large set of variables the script checks when the user uses their currency?
(Variables like where they clicked, what color they are wearing, what time of day it is, etc)
Becoming more of a game that players can actually persuade their outcome, study results and actually learn how to get the rarest item, even if itās incredibly difficult? Basically being not a game of chance but an incredibly intricate game of observation and studying outcomes, but on the surface it might seem as if itās ārandomā for a first time player and with no external insight from friends etc
You can simulate 10,000 random draws (throw random variables through your model) and then display the simulated percentages.
If 10,000 random draws yield item X 150 times, then item X has a chance of āabout ~1.5%ā.
Thanks, was just kind of wondering if itās required to display this data if itās not actually random but based off of circumstance that a player can dictate.
Either way I think Iām going to just remove the part where you use currency, since itās not really necessary and not a big part of it.
Thanks for the response:)
A regular RNG is not actually random either. If you have one based on percentages, it will never actually yield those percentages over a small sample size either. You could feed into it 5%/25%/70% as item rarities but actually observe 4.9%/25.2%/69.9% after some draws. By simulating the draws and then computing the percentages over those, you are pretty much doing the best you can and itās a good enough estimation of the actual percentages for a large sample size (i.e. > 10,000).
I would definitely display the data as courtesy to players, but you could also include an explanation about how it actually works for those that are interested.
The hypothetical system Iām talking about has nothing to do with rng or chance - Iāll give you an example of how it would work
So testing to get a % chance is not too reliable, as I can do the test 1,000 times and if Iām doing the same thing over and over again (wearing blue, itās day time etc) it will always yield a 100% chance of the same thing happening
it may seem like chance if you donāt know any variables but you can observe them and find new variables as you go. Wasnāt sure if this needed to be fully disclosed Iām guessing the same rules donāt apply since it isnāt a chance based game?
Thank you xx
(There will also be hints around the game about variables giving you specific outcomes, the idea is to get them knowing that their actions matter, but I wasnāt planning on disclosing every single variable and itās result, but itās something they can observe and learn themselves through the system)
Itās also basically skill based because some variables will be a little bit tricky to hit perfectly, think like super smash bros melee where you run into the finish line at exactly xx:04 seconds and it yields a result
If you just clearly tell the player somewhere near the fountain in the game that it is based on variables like that and that there are hints in the game that let you figure out how to get what items, you probably donāt have to disclose the odds then, since itās not actually random and they could figure out how to get the item they want by game play.
Why make a system like this in the first place? Do the players know the loot boxes arenāt truly random? I would be up-in-arms if I found out a loot box I bought was using preset variables I could have changed instead of being 100% random and beyond my control.
I havenāt implemented this system yet but players WILL know its not random - I wouldnāt really call it a lootbox, just another little mini-game to make some side currency. it isnāt expensive at all āto playā, it can be earned in a matter of minutes, or seconds, depending on how much you want to put in. (i say that because its more like a skill-based, knowledge game and less of a ālets open up a random boxā, also thereās a 3 hour cooldown on it)
this system is inspired by a current system in the game that iām going to rework to make it work like the above, pretty much based off of players requests. i have seen players speculate about āif you click here it gives a different resultā, but the truth is itās just random right now, and i think itās a lot more fun for it to be skill-based so they can tell their friends their secrets they know about the system. 
think like neopets where you can visit the sleeping dragon, if you attempt to steal treasure from him at a certain time of the day, it consistently works. neopets doesnāt let you know that theres certain variables, but players figure it out, and they spread it around, and its really cool and exciting to figure out an algorithm or pattern.
players love creating game-guides and stuff for āsecretsā like this when they find out it consistently works. and i love seeing them find out secrets <]:^)