You shouldn't take in small results and add them up because they are not representative unless these people share every single amulet they turn in, as opposed to the occasional 1 in 15.

If you still don't see what I mean, going by the data gathered on this thread alone, add them all up and check the % (4 in 85 remember) and then only add together the samples that had at least 100 amulets. These are much more realistic numbers from what I see.

Thank you~!

Gotcha~

If I ping those who have already posted here, to continue to report their amulet to shard/tome numbers, then it would fix the variability/skewness between scores. I understand what you're saying now. :)

I can separate the large quantities reported for now. I could probably make a second variable (large batch vs. small batch of amulets), thus turning it into a two-way chi-square. <--- double checking with you on this.

I know I sound like a snob, but Chi square isn't really a statistic. LOL It's a fun table you can draw by hand and look at. (Okay, it fits the technical definition, but why use it?) Lots of scientific folks who suck at math and are afraid of computers claim it's a stat because they can easily calculate it by hand and it's easy to grasp. Unfortunately, lots of them teach and they pass this horrible curse on to their students. Crossing the street would show up as significant variance w/Chi2. Anything that fits in a 2x2 Chi table is better off as a T-test. Use one-tailed probability if you're trying to be a bit laxer. Which software are you going to use? SPSS? It's been years since I did my graduate work, so I'll admit to being rusty, but… Chi2… REALLY?
(I tease, but I do adore you.)

PS -- In answer to your question, you can define your variables however you want. Group batches into whatever categories you want. You can do 3 or 4 sizes if you feel like it. Although if you're looking a 2x3 Chi table -- I'd suggest running a one way analysis.

Of the last three amulets I have submitted, all three gave me tomes. aka. 3 amulets = 3 tomes (2 water and 1 Life)

^nothing against you but i think this is why i would rather higher sample sizes be recorded.

doesnt t-tests have a n>30 requirement for the normal assumption to be valid. i honestly barely remember the chi squared analyses so i wont comment on that.

edit:

ran a small sample size, i normally turn in 100 or more, but decided to try something.

I just turned in 30 death mage amulets. 3 tomes, 27 amulets. :/

EDIT of an edit:

just turned in another 800, 100 of each element.

123 tomes. 1 in 6.5 chance. ~15.375% chance.

lol im done now. i have semi-shitty luck apparently.

I just went and bought a bunch of amulets to see what happens. These are turned in order, but I doubt it would matter.

Dark 3 Tomes, 21 Shards, 24 Amulets 12.5 percent tome, Approx 1 in 8

Death 2 Tomes, 11 Shards, 13 Amulets 15.3 percent tome, Approx 1 in 7

Earth 8 Tomes, 65 Shards, 73 Amulets (Had runs of 15 shards or more at least twice, ugh!) 10.9 percent tome, Approx 1 in 9

Fire 7 Tomes, 28 Shards, 35 Amulets 20 percent tome, Approx 1 in 5

Life 4 Tomes, 19 Shards, 23 Amulets 17.4 percent tome, Approx 1 in 6

Light 3 Tomes, 12 Shards, 15 Amulets 20 percent tome, Approx 1 in 5

Magma 6 Tomes, 16 Shards, 22 Amulets 27.2 percent tome, Approx 1 in 4

Water 4 Tomes, 14 Shards, 18 Amulets 22.2 percent tome, Approx 1 in 4 or 5

TOTAL 37 Tomes, 186 Shards, 223 Amulets 16.6 percent tome, Approx 1 in 6

Remarks:

1. From my results, looks like it's still close to 1 in 6

2. You need a sizable amount of amulets to provide a better picture. If I only used my first 50 or 100 amulets, the results will look worse.

3. They do not change the chance during weekend Quentin quests. A friend got poor results and sent a ticket to ask about it, and they said it doesn't change on weekend quests.

4. The cost is approx 327k per amulet = 72.921m 72.921m divided 37 tomes = 1.97m per tome = Approx 2m per tome

5. Your luck may vary, so do it at your own risk. I am not responsible for impulsive gambling :P

Quentin gave me another amulet today, turned out as a shard.

Another bunch of amulets for STATISTICS! thunderclap

Dark 5 Tomes, 35 Shards, 40 Amulets 12.5 percent tome, Approx 1 in 8

Death 9 Tomes, 23 Shards, 32 Amulets (For some reason, death amulets never tend to disappoint me) 28.1 percent tome, Approx 1 in 4

Earth 6 Tomes, 46 Shards, 52 Amulets 11.5 percent tome, Approx 1 in 9

Fire 6 Tomes, 42 Shards, 48 Amulets 12.5 percent tome, Approx 1 in 8

Life 8 Tomes, 33 Shards, 41 Amulets 19.5 percent tome, Approx 1 in 5

Light 6 Tomes, 30 Shards, 36 Amulets 16.7 percent tome, Approx 1 in 6

Magma 12 Tomes, 38 Shards, 50 Amulets (Started out slow but picked up nicely) 24 percent tome, Approx 1 in 4

Water 2 Tomes, 22 Shards, 24 Amulets (Had a run of nearly 15 shards with this one, WTF?) 8.3 percent tome, Approx 1 in 12

TOTAL 54 Tomes, 269 Shards, 323 Amulets 16.7 percent tome, Approx 1 in 6

I got to stop getting hooking on these, it's really costing me lol. And I still can't use them. kicks stat cap

Mostly from 2014... Shards: 8609 / 10271 Tomes: 1662 / 10271

I'll say it again : people doing small samples at a time have had the feeling they get terrible results such as 1 in 10 or worse for YEARS. It's nothing new to the past 2 months. These remarks already existed 5 years ago. Anyone who has provided you with larger samples eventually found it was close to 1 in 6 (the real % could be different, for all I know we're all just trying to round up to something)

I'm not really seeing what chi-square will get you that percentages don't in this case. We really just want to know how many tomes we'll get on average on say, 100 amulets, to figure out what price should be paid for amulets.

Those are valid points Alex. Most people don't turn in hundreds or thousands at a time, instead just turning in a small handful or stopping before the pseudo random number generator balances out (yes, it has some determinism in its distribution of numbers, it's pseudo after all). Usually the terrible results gets posted as complaints about the mage, although he could be an ass frequently.

I quote an ideal rate, and a more pessimistic rate because people tend to have high expectations for a tome, then become disappointed when they get a bunch of shards. I usually don't turn in hundreds at a time either, hence I see lots of little crappy results that I don't record.

12 amulets = 1 tome and 11 shards