Untitled

Ok, this is just some quick proof of my dorkiness. WoW patch 2.1 included a change to the prospecting math where attempts are always guaranteed to yield a gem now. Having read the patch notes ahead of time, I stockpiled a bit of ore to take advantage of what I hoped would be an upgrade in the gem drop rate. I prospected over 15 stacks of ore just now, and here are the results:

math

The quick math shows that fel iron prospecting produced 1.13 (35/31) uncommon gems per attempt.

Adamant gave me a 1.14 (40/35) uncommon gems per attempt and 0.26 (9/35) rare gems (total of 1.4 gems per attempt). Getting two uncommon gems at once is not unusual, and getting an uncommon and a rare gem together is also not unusual (looks like what’s guaranteed in the new math is at least one uncommon gem, anything else is extra).

I also had one adamant prospecting attempt out of the 35×5 adamant that produced three gems at once (2 uncommon, one rare). This could be a case of the two different chances (14% chance for an extra uncommon gem and 26% chance for a rare gem) both rolling at the same time. The small numbers math says that there’s a ~3.5% chance of both coming up at once, which jives neatly with my experience (1/35 ~= 2.8%).

One point of interest is that the uncommon gem production seems to be roughly even now between the two ore types (roughly 4.5 uncommon gems per stack). The only difference giving adamant a higher drop rate is its production of approximately 1 rare gem per stack of ore.

The sample isn’t big enough to make any definitive statements about individual gems having higher drop rates than others but my sample at least does seem to hint that orange gems are slightly more likely to drop than average and that blue and red are less likely. The difference here is well within statistically meaningless fuzzy grey territory, however, so don’t complain if you prospect 100 stacks of ore and get all azure moonstones

old math

By popular request, I have hunted down some of the old prospecting math. Kaliope has a wonderful blog on WoW crafting issues and recorded some numbers on the subject. She also has posts on mithril, thorium, and lowbie ore prospecting results before patch 2.1.

Her numbers use smaller sets than mine and are therefore even more prone to the vagaries of the random number deities. What she did notice was an 83% drop rate on gems from thorium (15/18), 1.35 gems per prospecting with mithril (27/20), and 1.29 gems per attempt with adamant (18/14). These numbers jive with my memories of the experience. She reported rare cases of getting 3 gems per attempt and several attempts w/o getting anything.

These are all over the charts but show two things:

1. the results from prospecting were much more random before this recent patch
2. rare gems really were pretty rare

Her one talasite from 3.5 stacks is a 7.1% drop rate… compared to the 26% drop rate that I noticed under the new math. The higher randomness prior to 2.1 meant it was possible to burn an entire stack of ore and get only dust. I remember very acutely a stretch of 6 adamant prospectings where I got no gems.

If her rate of ~1.3 gems per prospecting (assuming the thorium exercise was a fluke) was accurate, then the numbers are similar enough to mine to assume that no real change in the overall rate of gem production resulted from the patch… they just normalized things a bit, and I like it

thottbot

It just occurred to me (4pm, May 23rd) that Thott might have some data on the subject. Turns out they do. I’ve saved a snapshot of the data before it becomes too corrupted by the changes in this patch, but the trends they show are pretty easy to spot.

The adamant prospectings show a rate of 2.2-2.4% for each rare gem and 18.8-19.7% for each uncommon gem. Add things up and they give a 13.8% chance of getting any rare gem and 115.6% chance of getting an uncommon gem for a grand average of 1.29 gems per prospecting – confirming Kaliope’s numbers but showing that rare gems aren’t quite as rare as in her experience.

The fel iron attempts have about half the sample size, but are still quite statistically valid. Rare gems had a 0.7-1.0% chance of dropping (5.3% total) where uncommon gems had a 15.5-16.8% chance (97% total). So with fel iron, people were getting ~1.02 gems per attempt. Fel iron clearly had a lower drop rate.

So in conclusionary fashion, I will finally give a rest to this post by stating that my current guesses as to the changes to the new math:

• Both fel iron and adamant have been given the same base chance to prospect green quality gems. This chance is probably similar to the old adamant rate of 1.15 gems per attempt.
• Adamant rare gems look like they might have had their drop rate doubled. Clearly, making this statement from only 35 samples isn’t as accurate as it could be… but something like 0.25 rare gems per attempt sounds like a perfectly reasonable number to me – especially if I were the one pulling these numbers out of the proverbial hat when tweaking the algorithm