Yes. I admitted I was wrong about 3/65536, and said that it was 24/65536 based on my assumption. You chose to ignore that and just say I fail at math, which you continue to do despite other people agreeing on the result. I tried to make my assumption more clear in the first post.

There are three different "correct" answers, depending on how the loot table is rolled for. 8 independent rolls for one of four pieces of loot 4 independent rolls for one of ten combinations of two pieces of loot 1 independent roll for one of 165 combinations of eight pieces of loot Analagously: You have eight sets of four cards (one for each class). Every time a piece of loot drops, you pick one of the four cards to determine which piece it is. You have four sets of eight cards (two for each class). For every boss, you pick two of those eight cards simultaneously to determine which classes get loot from that boss. You have 32 cards (eight for each class). You pick eight of these cards simultaneously which determines how many loot pieces drop per class for the whole instance. Practically, the last one is unlikely because each boss drops a different slot; order matters in the sense that you'd have to assign each piece of loot to a boss to know what specific item drops. The second one is more likely, but that'd require storing the 10 combinations (and more given that you have other drops). From a programming perspective, I'd say its most likely that it's the first since you can also use the roll to determine if a piece drops at all. How the loot is rolled for internally absolutely does matter. Just because the user sees the same end result (8 pieces of loot) doesn't mean that the probability is the same for each case.

Thank you. I'm gonna develop a web app to allow people/guilds to submit the drops they see in HM's in the hopes of getting enough data to make some real conclusions.

I like how you edited your post afterwards to make it even more condescending. Go troll somewhere else.

You're still wrong. Sorry, bud. But the condescension is very mature of you and contributes a lot to the thread.

Still tiny! I think its time to start actually collecting some loot table data. Let's find out what the drop rates really are.

Chill out dude. It's not a big deal to be wrong. No need to be an *** about it. You're making a big assumption in stating its 165 combinations. You don't know how the loot is generated. No one does because Bioware won't tell us. It could just as easily be 3/165 as 24/65536. Nice ninja edit. And it does matter how the loot is generated internally. You still have the same number of outcomes in terms of pieces of class loot, but the probability of each result will be different.

While there is no way to determine which is first or second, the correct result depends on how they implemented it. If they choose loot one piece at a time, then its still 4^8. If they choose loot one boss at a time as a choice of 1 of 10 combinations, then I agree it would be 10^4. It'd be interesting to see how they do it. I suspect it would be easier to implement it as a series of random rolls for whether a piece drops, and then another roll for what it is, as opposed to storing all the combination set and picking from that. You're definitely right about this. 3/65536 is the chance of getting 7 smuggler pieces for the first 7 and one other class piece on the last boss, which is an exaggerated number to illustrate this point. So then assuming there are 4^8 possibilities, shouldn't it be 24 different ways of getting 7 smuggler and 1 other? You have AAAAAAAB and its other 7 similar results, as well as C then D. 8x3=24.

I'd like to address a few points that people are bringing up: 1. Random Number Generation Random number generators are not truly random. True randomness in a computer RNG is impossible, but the algorithms used can get close enough. That being said, if not implemented properly, you can get results that are consistent; exactly the opposite of what the RNG should be doing. There are plenty of examples of failed RNG implementations in gaming that have been abused. In this context, a poor implementation could mean certain pieces of loot favored over others. 0.0046% means 1 in 21845 Karagga 8man HM runs will result in this loot set. That's not nothing, but its staggeringly low. In practice, you should almost never see that. If you compare this to the probability of getting 2 pieces of Rakata for each class, which is undoubtedly significantly higher (someone better in probability might chime in with this result), you'll realize how ridiculous this is. 2. Selection Bias Selection bias, roughly speaking, means that we only remember the times when things are unusual, as opposed to the times when things are normal, even if the normal occurrences happen far more often. While it is true that in general, people won't post on the forums except to complain about something like this, I find it strange that so many people are complaining about Smuggler/IA drops and nothing else. I realize that it's one of the least played classes, but the complaints are very consistent. And if I ran Karagga's and 7/8 drops were for Sages, you can bet that I'd be complaining just the same. But it seems that you just don't really see as much complaining about large amounts of drops for other classes... And to the guy who was correcting my probability: I made the assumption that each piece of loot is determined one at a time, completely independently, which would give you 8 IID's. In this case, "order" matters, and I think this is a more likely scenario than the entire loot table being generated as a set, especially since a result has to be assigned to each piece. If it was generated all simultaneously, however, you would be right. Unfortunately there's no way of knowing.

Bioware, this needs to be fixed. Assuming each Rakata piece always drops and each piece is rolled for independently, there are 4^8=65536 possibilities. 24 of those possibilities are 7/8 smuggler, and 1 of another class. 24/65536=0.0366%. This means either: 1. Certain classes are favored over others OR 2. Something in the loot tables is broken. Whatever the case, this needs to be fixed. We had one smuggler who got 7 pieces of loot (not counting all of the exotech that he got too). He couldn't even use 3 of them because they were dupes. Edit: Fixed the math using the assumption that there are 8 independent rolls for loot.