Fix ATMA drop [LIKE] Thread

[Lexiii]

So I mentioned this earlier in this thread but it took me 360 fates for all 12 atma. In the last 3 days, I just grinded all 12 atma again. This time it only took 190 fates. Maybe SE did already increase the drop rate? Or maybe I was just luckier. Regardless, grinding atma in a short amount of time is totally doable. Hours-wise, it took me <15 hours (I used 30 Gysahl Greens plus sometimes my choco died). Anyway, I'm sure SE will increase the drop rate eventually. I wouldn't be surprised if they did it with 2.4 but I'd say definitely by the time of the expansion.

RNG is RNG is RNG. I just finished mine and it took over 550 fates. The 3% drop rate hasn't changed.

[Squa11_Leonhart]

RNG is RNG is RNG. I just finished mine and it took over 550 fates. The 3% drop rate hasn't changed.

Statistics involving atma drops have already set the drop rate at 1%.

[Lexiii]

Statistics involving atma drops have already set the drop rate at 1%.

There's absolutely no way this is true. This puts the average time for Atma at 1,200 FATEs. The distribution doesn't match anything I've seen.

[Squa11_Leonhart]

There's absolutely no way this is true. This puts the average time for Atma at 1,200 FATEs. The distribution doesn't match anything I've seen.

And when you average a group of players thats about where the data medians in at.

However for a single person, http://wow.joystiq.com/2010/01/13/dr...ity/#continued

[Lexiii]

And when you average a group of players thats about where the data medians in at.

Source? Give me unbiased data. If you understand probability, then you also understand sampling bias :P.

[Duuude007]

Source? Give me unbiased data. If you understand probability, then you also understand sampling bias :P.

Indeed. You would need multiple sets of twelve (30+ completed Atma weapons if we're being serious about the analysis) using the same exact means to track each: number of fates divided by the number of atma.

Find the mean value, the standard deviation, then you have sufficient data to calculate probability which can be used to help forecast drop rate.

Based on my experience, my results have implied a 3.25% droprate on average. But I would prefer at least 4x more data for it to be considered a sufficient sample size.

[RiceisNice]

...

I've always seen 3% being the magical number for the droprate here (which is still too damn low imo for something aimless like this). Probability on paper always plays out differently than actual practice, especially with very few variables that influence it in the latter (straight up drop rate.)