Gospel:
This issue in this case is that people asked here, and we ended up testing it and documenting it in the Craftsman forums.
The reason it was originally asked (8+ years ago) was that the Devs said that humans had a flat 40% chance of getting colored ore, everywhere, and that Elves would have a 20% bonus, but didn't say if the bonus was additive (adding +20%) or multiplicative (multiplying the human chance by 1.2 (aka 120%)).
Now, this didn't match pre-ML mining, where the thousands of us who put together Mining books for our or others' use KNEW through long term use and testing that -
- some spots rarely produced colored ore (I had one Verite spot that was more likely to produce a Valorite elemental with a GPA, than even a single Verite or Valorite (with tools) ore),
- and, others that ALWAYS produced 80%+ colored ore (and allowed 20-30 digs each time, in Trammel!)
Nor did it match ML-era mining, where it quickly became apparent that the elven "20% bonus" was
- MAYBE a 20% increase in swings (usually by increasing the number of medium and small ore for an elf)
- and/or multiplying the existing percentage for humans by 120% (so that a 20% chance of colored ore for a human was 24% for an Elf, and 80% human became 96% for an elf). The bonus was most apparent on spots that were good, but not great before ML (1/4 and 1/5 are well within statistical range of each other, especially with humans getting a extra ore 10% of the time, and the expected range of iron on the highest-color-output spots was also not likely to show much (how obvious is it that you are getting only 1-3 iron ore out of 30 digs as an elf, when you used to get 2-4 as a human?)
And, the ore randomization proved that while the ore colors might change, the iron/colored ratio for each spot stayed the same no matter what color it became (and even spots that were currently 100% iron preserved those ratios so that they came into play if hit with a PT or GPA, before or after ML and the randomization).
A lot of things got documented solely because people asked, and not even experts could find answers in official documentation or unofficial answers from devs - and sometimes even those answers don't help. Some things are just TOO BIG to test on one character, in a timely manner (Valorite BOD drop rates, for example).
A few of the things in the last 5-6 years that questions, and multiple people providing data, resulted in...
- Proving that there was a bug in the Smith LBOD system (it was a a matter of a missing parenthesis - spellcheckers don't help with code, and much of UO was written before coding software would tell you if you left a hanging open parenthesis), and that the advertised +10% to exceptional quality BODs was either removed in Publish 19, or was never implemented in the original release (despite being in the design document).
- After that fix, confirming that Smith BODs now worked as advertised (to within 0.1% of expected values after tabulating 3000 BOD pulls from over a dozen people)
- Discovering that no one had ever updated the design document with Tailor BODs, and the normal to exceptional ratio for Tailor BODs was 2 normal per 1 exceptional, not an even distribution of either (as Smith BODs).
- Semmerset and others' work on Reforging.
And, that's just stuff for THREE skills in one aspect of the game (crafting).
The only documentation that's been done officially by EA since AoS has been when new systems were introduced or old ones massively changed, and a lot of that got changed between announcement and release, but the docs never updated to match.