![]() |
|
![]() |
|
Thread Tools | Display Modes |
#11
|
||||
|
![]() Quote:
My stats books are in a box that is inaccessable right now and I cannot remember how to set up the question. Thanks! I'm genuinely interested in re-learning. | |||
|
#12
|
||||
|
![]() Quote:
Secondly, where is this 9% number coming from? The odds of killing Grachnist 128 times and not seeing the earring are indeed roughly 2%, assuming a 3% drop rate. However, assuming a 9% drop rate, that percentage drops into the 10^-4 range. So changing the probability of a non-drop from 0.97 to 0.91 makes a huge difference. I suspect that the drop rate is lower than 9%, and/or your 128 figure is perhaps inflated a bit. The calculation itself is trivial, just take the drop rate percentage to the Xth power. | |||
|
#13
|
|||
|
![]() You were right on my figure being wrong. It's 129 from 1:30 P.M. on 26 sept until 2:53 P.M. on October 3rd. Can post the log line items. Or you can choose not to believe me. I welcome Uthgaard or anyone else looking into the matter. That's the purpose of this thread. On those 129, I saw 2 totems, 129 gloves, and 0 earrings (as well as some skulls, words, etc.).
You were right to question my drop rate. It's a bit north of 7.5 percent. Pre-patch figures tracked: 10200 placeholders killed 332 Grachnists seen 5 totems looted 25 earrings looted % of the time that Grachnist spawns = 3.25 (this is a little equivocal because I killed a few goblins that I knew were not placeholders b/c I was working on faction - so the number is a touch higher - but also is irrelevant to the discussion) % of spawned Grachnists that drop the earring = 7.53 - and that's the number in question. Less interesting is the longest dry patch heretofore was 23. What's in question is whether the droprate is broken or has changed. It's arguable that I don't have a large enough pre-patch sample and that I was extraordinarily lucky - repeatedly. That's a harder position to defend than the one I'm taking, though. The evidence that I have suggests that the droprate's been nerfed into oblivion. The number I get in excel when I use your formula is so small that I cannot understand it - as a percentage it's 0.004635982529% - as a number that's easier to figure with it's .00004635982529. What that means to you and me is that (given the previous dataset and the 7.5% number), I'd have a 1 in 22,000 chance of having such terrible luck. (I think I did the math right.) That is to say that if you killed grachnist 128 times as a set, you'd have to kill 22,000 sets of 128 to not get an earring in one (on average). That's enough to give one pause. Which supports my contention that what I've observed is statistically impossible (hell, it's possible that 2 and 2 are 5 and we've all added wrong from the beginning of time - but that's not a useful supposition) - and I'd hope that when the devs have time they can look into it. | ||
Last edited by Pan; 10-04-2012 at 06:52 PM..
|
|
#14
|
|||
|
![]() 139 now. I really need to leave this alone...
| ||
|
#15
|
|||
|
![]() Given your pre-patch statistics, if accurate, yeah it's extremely likely it got changed in the patch.
| ||
|
#17
|
|||
|
![]() Good to hear that it can still drop, but I'm going to echo Pan's request - can some one please verify that the drop rate hasn't changed (decimal shifted or something like that, perhaps)?
I know it isn't necessarily off - but it would suck if everyone just assumed it wasn't when it was.
__________________
Don't be a Kleat...
| ||
|
![]() |
Thread Tools | |
Display Modes | |
|
|