#11
|
||||||
|
Quote:
Quote:
Quote:
| |||||
|
#12
|
|||
|
TLDR: doesn’t really understand statistics
| ||
|
#13
|
||||
|
Quote:
Statistics isn't really that complicated, but I understand you struggle with pretty much everything in life. | |||
|
#14
|
||||
|
Quote:
| |||
|
#15
|
||||
|
Quote:
So about four in every hundred would-be lords of the ring will see such a stretch of bad luck. Given how many people camp that thing, it’s not at all surprising that you know someone who is experiencing such a run. I guess if you know exactly two people who have ever camped the ring, you’d be in some reasonable territory. But you likely know many people who have and you’ve only got data on a pair of recent bad runs. So it’s a biased data set. Again, if you want to call the wiki bull shit or argue that the drop rate has changed, you need a run of bad luck so exceptional as to fall outside expectations. 160 is just not that surprising. Edit: I found the 3.5% source! It is based on 224 kills. So 3.5% is the point-estimate, with the margin of error that means the drop rate was most likely between 1.5% - 7%. So, as above, a string o 160 failures is consistent with that dataset. That is, if you think the drop rate has changed, you'd need to find evidence that the drop rate is decidedly outside that interval than your two friends' recent string of (not outside expectations) run of bad luck. Would be interesting to collect a fair sample of people’s camping experiences to refine that estimate, though! | |||
Last edited by jadier; 02-28-2021 at 08:13 PM..
|
|
#16
|
||||
|
Quote:
(1) this isn't a "fair sample" so a strict application of standard deviations isn't really appropriate. My guess is that DMN knows way more than two people who've camped the ring, and is *only* reporting the two most egregious bad luck runs he's seen (because most people only talk about really bad or really good runs of luck; you brag when you get it in <10 couriers, and complain when it takes >50, but no one spontaneously posts in guildchat or whatever about that time they got it in 31 pops). (2) there's a margin of error associated with the 3.5% estimate. If we think the drop rate has changed, we'd need either a more representative sample of camp results, or a staggeringly large outlier dataset (which this is not) given the uncertainty in the 3.5% estimate & the bias of the sample. Edit: To put it another way, if you take the 224 kills with 8 rings the 3.5% is based on, and just add another 160 pops with 0 rings from the "DMN dataset" here (which, again, you shouldn't do as the DMN is biased in ways the original isn't) you get an estimate for the drop rate of...between 1 and 4%. Even with the bias reporting, it doesn't exclude the 3.5% estimate. It's just not that bad a run. Edit edit: To be even more precise: DMN's friends would need to observe 221 total pops without a ring to *just barely* exclude 3.5% if you lumped them together (which, again again, you **cannot do** since the friends-complaining-about-bad-luck dataset is heavily biased). | |||
Last edited by jadier; 02-28-2021 at 08:24 PM..
|
|
#18
|
||||
|
Quote:
| |||
|
#19
|
||||
|
Quote:
That is, if you know that the normal chance for something is 10% with a sd of 2% and I tell you, "hey, I got a drop rate of 5% with this dataset where I just ignored all the people who got the drop more often than 7%" you can't just go "oh well 5% is more than 2 sigma away from 10, so I guess the droprate changed". [ I'm not saying DMN did this. This is an extreme example of why using Z-scores to interpret data relies on fair sampling. If sampling isn't, fair, it's meaningless to calculate Z scores ] In other words, you can only meaningfully interpret a Z score if the sample's fair. If it's a biased sample, it doesn't readily translate to the probability expected from the standard deviation...because it's biased. Regarding the assumption: the OP's question was whether the drop rate changed. 8 / 224 = 1.5 - 7% drop rate, 1 / 160 (even with the bias) translates to a <0.1 - 3.5% drop rate, and 9 / 384 = 1 - 4%. That is, they're all consistent with one another. So you're correct that assuming a 3.5% drop rate, 1/160 is an unlikely observation...but my point is that although the point-estimate may vary by 2 sigma, when you account for sample size, even this biased dataset doesn't preclude 3.5%. Eg, nothing posted here implies the droprate changed at all, even if you assume the true drop rate was exactly 3.5%. OP's friends just had bad luck, and the true drop-rate is uncertain but we'd need way, way, way more evidence to suspect a change. (edit: tldr; if you assume a 3.5% drop rate, the point estimate of 1/160 is, as Isomorphic says, >2 standard deviations away from 3.5%. However, (1) the data are biased, so there's no way to actually connect a probability to the observation, and (2) even if it weren't biased, 1/160 is still not inconsistent at 95% confidence with a 3.5% drop rate because while an assumed distribution doesn't have error, 160 is a small sample size when dealing with a 3.5% chance event so the confidence interval still overlaps the previously-estimated drop-rate) | |||
Last edited by jadier; 02-28-2021 at 09:52 PM..
|
|
#20
|
||||
|
Quote:
Got 3 CoS before getting the damn idol. | |||
|
|
|