Possible RNG unbalance explanation

[Horror Stories]

With the event coming to a close, many people are questioning the true randomness of the the RNG in Defiance. I think I have an explanation, but it might not be correct. I used to play an MMO on PC where players could upgrade their gear with a % based chance. Example: If a player wanted to upgrade from a +12 item to +13, the success chance was around 0.3% Players in that particular MMO could only got to try an upgrade their gear maybe twice a week. And yet certain people still had all their gear at + 13 and higher. It was always the same people who would "hit" upgrades like that, while the majority of others could play for a year and never get past +11. A developer for that game explained to me why that happened. He said that every players account is assigned a number when it is created. Depending on where the number is on a line, it would further affect a players chance at "hitting" an upgrade outside of the standard upgrade chance %. For a few certain lucky people, they would get an account number that put them in a sort of "hot zone", that caused everything they do to have hidden but better % with anything they do. So if one of these lucky players goes for a 0.3% chance upgrade, the actual chance for them to successfully upgrade would be much higher than the default 0.3%. While everyone else with normally placed account numbers have no type of hidden bonuses and that's why the majority of players never hit endgame upgrades or pull endgame loot. I feel like that type of system is in place with Defiance, because it is the same people pulling orange items over and over and over and over. I've seen people in my clan call out the orange items they've gotten today, one person has gotten more orange items today than I have in the entire 2 week event. I would really like for the developers to go back and check their coding for the RNG, because it doesn't seem to be all together fair and balanced. Not that the current RNG is a bad thing, just that their might be a few logic flaws in the coding that are giving certain players/account better individual chances of procing that weren't caught at the time of the original coding.

[SUPREEM]

That makes plenty sense. I also notice the same handful of players (ps3/na) scoring the best drops daily. I know a player that got nearly all hh named guns from arkfalls alone. Also, why the hll were t4 HB ar recoil stocks and power bores so rare? Was that on purpose or what? And t4 HB precision scopes for HB pistols, really devs?

[DJ51]

I'd love to believe the RNG is truly random but after two years of play and seeing similar characteristics i often wonder after all that time only two jackpots ? must be random then

[Ryker Vorton]

i know that's the case of like 99% of korean games. I remember in the one i worked for three years one game aster waas fired when we discovered he had found a way to manually change the "lucky numbers" of his friends in the game (among other things) so they not only had the highest upgraded items in game but also the rarest ones .

Though i believe the system used in defiance is much more balanced and quite "honest" i can't make blind eyes to what i see in my own characters, whie one of them almost never find anything good the other one has been on a lucky streak since day one.

[ShatterST4R]

Sounds like you described MU. xD

[Keeper Riff]

It's more complicated than this. The main problem of the computer-generated RNG is that it's usually based on a clock which the game itself runs on. Simply put, you query the RNG at the time of 15:35:51,59 and receive a "random" number of 159. Next player queries at 15:36:06,98 and gets the number of 698. So the only thing that is random in the RNG is the timestamp at which it was queried. It usually works fine until queries start to follow in sheer numbers and are put into a queue. If a queue is resolved every fixed period of time, then all RNG results may come the same or the gradually increasing or decreasing pattern may be seen.

To counter this, other pseudo-random parameters are added into the RNG function, like player's ID, character's coordinates, ID of the equipped item and so on. Usually not the whole ID but only its lower orders are used to prevent the ultimate "lucky" characters appearing. But when it comes to very low probabilities, still some characters may become more lucky than others. For example, those with odd IDs may be luckier than those with even IDs or the other way around. There can be lucky weapons you need to hold in your hand or lucky places you need to stand and so on. The worst thing is there can be pathologically unlucky characters who will never have success in some action just because their ID is divisible by 256 or something like that.

There are also such things like popularity-based probability. The RNG function auto-tunes itself depending on how many people "won" already in the designated period of time, having hard or soft daily/weekly/monthly limits on giving away good stuff. The "good" stuff is determined by players' behavior, it's things that they use frequently and don't vendor sell or destroy. This way a lottery increases people's interest in it when it's not popular but makes them spend more money/time to invest before they get the desired result when more people start playing it.

[brokenshield]

This was what I was wondering. It just seems that every game I've played, that had loot tables and RNG, have been lopsided. The problem is the ones on the receiving end of any issue like this, refuse to see that it is at least possible that something is amiss. I have yet to see someone on the underside of loot drops defend the randomness of numbers generators.

Until I start to see some averaging in my drops, I will firmly believe the RNG has an issue. Throw statistics and charts at me all day long if you must, but the facts speak for themselves. The Earth revolved around the Sun, long before someone finally believed it.

Edit: I have heard the above mentioned in other games. Things like time of day played and even region. Also could explain why the gear I do get tends to be the ones I never use.

[Tekrunner]

I bet one fun part of being a developer on a loot-based game is watching people come up with all sorts of conspiracy theories and all the anecdotal data they present as "evidence" that your RNG is not actually random.

[brokenshield]

Or versa vice.

But again, there isn't an issue here because you weren't affected. So anytime you do have an issue it is, by your definition, a conspiracy. Interesting.

I remember now why I find forums less than useful. No room for discussion, just the "facts" as stated by the select individuals.

I leave this place to you. Peace, out.

[Tekrunner]

I don't especially want you to leave this place, even if you're implying I said things I haven't actually said. But anyway, this type of posts appears regularly in every single loot-based game, and the response is always the same. Using anecdotal data from yourself and your group of friends isn't enough, because of things like self-selection and confirmation biases. I'm sure there are games out there that skew their RNG, but if you want to prove it, you're going to have to come up with actual evidence.

If your evidence is "I keep getting crap drops at this event while these people over there are swimming in legendaries", well guess what, that's how RNG works. Low chances happen to have a large variance over a limited sample, which means that during a single event, over a large player population there will be a significant number of people who get completely screwed, and a similarly significant number of people who "get lucky". In this case what you should ask for is not more or better randomness. What you should ask for is less randomness, with systems like token rewards and crafting that reduce RNG and increase the likelihood that you will get exactly what you want.