RNG log

So i want to know what would be an acceptable way to log RNG for crystal tower? Pictures and time/date and a recording of loot dropped? I know and I have FC the members who have also mentioned the insane amount of bard and monk and dragoon gear vs caster and tank so I decided to see what would be a accepted way to show the devs the RNG algorithm needs to be redone.

Also a sample size, 100-200-300 runs?

1 million runs will be ok.

Pretty sure the devs already track this information, and have a heck of a lot more data then you could ever hope to collect.

I will agree that I personally see a disproportionate drop rate between PGL, ARC, and all the other classes (especially casters). This week I ran CT 5 times before finally getting a drop and quitting. 4 of those runs I saw PGL gear drop almost exclusively. Last week, I could not get a drop for my healer or caster DPS at all the entire week (did about 15-16 runs). I saw the GLA/MRD body once, the LNC gloves drop once, and the LNC body drop once...other than those 3 every other drop was PGL or ARC (you can do the math, that's 57 chests of nothing but archer and puglist shit).

Now I understand that with RNG you should expect anything for any single given run, but sampling and trends should ultimately show equal distribution in the end. From personal experience and data collected from others, this isn't the case, and more and more I'm believing the loot tables are skewed to favor certain classes. It is quite disturbing and making it incredibly hard to find gear for my DoM classes.

How something that is "random" would show equal distribution? Heck, in a 10x run, it could even drop the same loot over and over. As it is RANDOM and not a percentage.

Well loot tables aren't 100% random. Most people think it is like 'Mnk Gloves' are assigned 1, 'Pld Gloves' are assigned 2, RNG 1-2. It is more like Mnk is assigned 1 through 6, Pld assigned 7 through 9, RNG 1-9. (way over simplified to express the point). This is chosen due to a number of factors and is likely not shady at all... in intent.

Yes, SE has more data than any 1 person could ever collect. However there is good reason to do this- A lot of people over a long amount of time and in different circumstances see the same thing. There is a lot of bias in this, but it is consistent enough to be worth looking in to. If there is something off about the loot tables then we need to make it public for SE to bump it up the priority list. Squeeky wheel and all. There are even times where the community points out problems that the Devs don't know about yet. Even if it is only the extent of the problem being new to them that is quite helpful for us. Having the data doesn't mean knowing there is a problem, understanding it or committing resources to it.

Where's that guy with the RNG cartoon strip, he always improves the RNG threads that pop up every day of the week.

Quote:

---

Originally Posted by Yashnaheen

snip

---

I don't see any reason why would they use that kind of method, the one you're referring like "mnk glove = 1-6" "pld 7-9". The loot table per boss is fixed. And what number of factors you're actually referring to?

Quote:

---

Originally Posted by Thorauku

I don't see any reason why would they use that kind of method, the one you're referring like "mnk glove = 1-6" "pld 7-9". The loot table per boss is fixed. And what number of factors you're actually referring to?

---

I think what Yash is getting at is that while its random, each item has a certain chance of popping up. Rng can and often does use percentages to determine the odds of an outcome. item1 = 25% chance, etc. Of course, random outcomes being what they are, we could get the same item 10x in a row without it meaning the system is broken. The only way to know for certain is to track the outcome over thousands of attempts. Even at a couple hundred, the results are still likely to be inaccurate.

Of course, the system could be broken, but there is no easy way to tell. Even a large amount of people complaining about the system doesnt make it broken. At launch, I did ifrit once and got my first weapon. third time i got my second weapon. A friend killed ifrit close to 40 times before seeing his tome for scholar.


For what its worth, in my personal experience of late, ive seen enough WoL and Crimson gear to drive me crazy.

The problem with people posting in these RNG threads:

"Hey guys, check out my random dice number generator. Well, yeah, I know it SEEMS odd that it always rolls either a 1 or a 6, but that's just the way RNG is!"

Quote:

---

Originally Posted by Dreamer

The problem with people posting in these RNG threads:

"Hey guys, check out my random dice number generator. Well, yeah, I know it SEEMS odd that it always rolls either a 1 or a 6, but that's just the way RNG is!"

---

Nope, not a problem, that's just how it really works. Unless you can explain how RNG should work, then there's no problem with the people posting in these RNG threads.

the problem i see isn't so much the RNG as much as the sheer amount of rotting that goes on.

6 weeks to get an 80 set when it's like 8 weeks to get a 90 set seems silly.
having everything but chests be freely lootable and chests be a weekly quest reward would make more sense imo, still looking at 6 weeks minimum to be done with it and its not even endgame content

I just wanted to know what would be acceptable for recording, I'm not going to be running it over a thousand times in a week this is a longer term thing I usually run CT 20-30 times a week so given a month or so I think I could get enoughdata just to look at, maybe iI'll prove the algorithm is working as intended or maybe I won't, I'm just asking for a widely accepted means of recording I was thinking screen shots with time date stamps and just a log in a spread sheet?