Orb of Chance Community Log

4 Magic, 2 Rare, Onyx Amulets.

updated, ty

burn 14 orb of chance + 13 scouring for 1 vaal pyramid map

13 magic
1 rare

ty

Speculation time:

Chris said the orbs give about a 1:5.4 ratio. The fact that he gave a ratio like that rather than saying "about 18.5%" I think tells something

The way these sorts of games often run their rolls is through a weighting system.

So basically, imagine that a rare result has a 100 weight and a magic result has a 440 weight. They choose a number between 1 and (rare+magic) and that means 100:540 of the items are going to be rare, and 440:540 are going to be magic.

Now thinking about uniques. What would you do if you had to make a very low unique roll to fit in here? Well, if we imagine that the weights are exactly 100 and 440, a unique roll might have a weight of 1.

That would end up meaning that:
100:541 would be rare. (18.48%)
440:541 would be magic. (81.33%)
1:541 would be unique. (0.18%)

This could potentially explain the data collected on rares so far. It's a very fuzzy match, but I think it's actually a decent guess considering Chris described it as a ratio and not a percentage.

A ratio like 1:5.4 would be easier to think of and write down off the cuff if you remembered in your mind that magic had a weight of 440 and rare had a weight of 100. And if that were the case, 2:542 might be too high to match the data and 1:541 is quite reasonable. These weights are typically integers not floating point numbers so 0.5:540.5 is unlikely, and if it was something like rares are weighted 200 and magics are rated 1080 (which is definitely possible) then he might have given a ratio like 20:108 or something rather than 1:5.4 since that's where my head goes when I try and reduce that.

Anyways the fact that the numbers were off the top of his head to me indicates that they are numbers created by a human instead of a formula. The fact that he used a ratio to describe it suggests to me that they use a weighting system rather than a fixed percentage type system, and the fact that he says that it's from memory and it's "something like" indicates that they use pretty nice round numbers (otherwise he'd probably have to check or would give a percentage type response as a percentage might be easier to remember) and that he knows that the number is a bit off (IE: 100:541 rather than exactly 1:5.4)

I'm totally full of crap, but it's fun to speculate like this in the absence of a large reliable data set.

I saw math and ran away.

In all seriousness, I am pretty slow when it comes to that sort of stuff. However, I'm skeptical about the ratio meaning anything special. IIRC, RNG in programming is generally handled in ratios rather than percentages which would explain why he stated it that way. In any case, I don't know much about either topic (math or programming), so I really have no business discussing it. :S I'd take part in the potential conversation there if I could, though!

3 chance on map, all magic.

Awesome thread; i was actually keeping a log myself, but it seems im a bit unlucky;

4 rares, 34 magic, 0 unique

5 magic

Ive been working on Onyx amulets for quite sometime now. this is the result:
35 magic, 13 rare, o unique.