Just a little tidbit, you should replace "Expected" with "Average"; there's a rather large distinction to be made here when dealing with probabilities. Expectation is usually within 90-95% odds of receiving it. Average, however, is much, much lower, usually between 60-80%. IE: If you flip a coin 10 times, you will on average find 5 heads and 5 tails(insert argument for weighted sides, not truly random, etc. whatever about that) but the actual odds of this happening to you aren't even close to 100%. 10C5 = 252 / 1024 = .246 or roughly 25%. Even though over infinity you will, on average, get 5 heads 5 tails(in a truly random test), in individual tests you won't generally see 5 heads and 5 tails.

That is if blizzard have programmed everything right.

No one knows how it works.... They may have wanted it to be 16% and they may have said it is unrelated. But I have noticed lots of times that me and my party are getting legs from same mobs / chests. This contradicts whatever they say.

Just a little tidbit, you should replace "Expected" with "Average"; there's a rather large distinction to be made here when dealing with probabilities. Expectation is usually within 90-95% odds of receiving it. Average, however, is much, much lower, usually between 60-80%. IE: If you flip a coin 10 times, you will on average find 5 heads and 5 tails(insert argument for weighted sides, not truly random, etc. whatever about that) but the actual odds of this happening to you aren't even close to 100%. 10C5 = 252 / 1024 = .246 or roughly 25%. Even though over infinity you will, on average, get 5 heads 5 tails(in a truly random test), in individual tests you won't generally see 5 heads and 5 tails.

Average is accurate, expected is not.

Expected is exactly the right term, specifically when talking about probabilities. This was also discussed on the previous page. And this might be helpful as well.

You maybe could include other weapons like TF and the cold damage sword?

Compared to Shard of Hate, Thunderfury is twice as likely and Azurewrath is 4 times as likely, so you need 50% or 25% respectively of those listed shards

From a first glance a few items that might be of interest as well: Burning Axe of Sankis (fire), Serpent Sparker (hydra wand), Gesture of Orpheus (arcane wand) and Sloraks Madness (disintegrate wand).

Just a little tidbit, you should replace "Expected" with "Average"; there's a rather large distinction to be made here when dealing with probabilities. Expectation is usually within 90-95% odds of receiving it. Average, however, is much, much lower, usually between 60-80%. IE: If you flip a coin 10 times, you will on average find 5 heads and 5 tails(insert argument for weighted sides, not truly random, etc. whatever about that) but the actual odds of this happening to you aren't even close to 100%. 10C5 = 252 / 1024 = .246 or roughly 25%. Even though over infinity you will, on average, get 5 heads 5 tails(in a truly random test), in individual tests you won't generally see 5 heads and 5 tails.

Average is accurate, expected is not.

Expected is exactly the right term, specifically when talking about probabilities. This was also discussed on the previous page. And this might be helpful as well.

Interesting, It's a term that(at least in this instance) is entirely synonymous with Average, but leads credence to confusion among people who don't immediately know what the word means. I stand by saying that in this instance Average would be less confused by some, and I don't understand why you'd use a fairly technical phrase in a report to laymen, but I suppose I was mistaken.

Screw that damn Wand of Woh... Inigo farmed 4k Legs to get 1 single one, thats not including shards spent.... Could take well over a year for someone to find one if they play like 4-6hrs a day... Ridiculous..

Updated for 2.05 assuming that Kadala uses the same weightings that were datamined recently

Thank you very much for keeping this up to date. I wonder if it's possible to update this to 2.0.6? We have no new datamined information, but for all we know the class set items' value in the tables have simply been doubled (to make it on par with, e.g., Blackthorne's). That would not *exactly* mean that you can half all those numbers - but it probably reduced the expected shards by about 40% or so per set item. My guess. Haven't done the math. Thoughts?

Also, I'm wondering if anyone is gambling weapons currently - and could keep an eye on the wands/other weapons ratio to find out if it's really 99k (2:1 ratio) or higher (my guess). I'm totally sick of RiF right now, so not many shards incoming anytime soon that could help me to find this out, but maybe if I go back to rifting next week once football's over...

Oh, and completely forgot to add this to the guide. Will add it right now, way too interesting to not mention in the RiF section.

Updated for 2.05 assuming that Kadala uses the same weightings that were datamined recently

Thank you very much for keeping this up to date. I wonder if it's possible to update this to 2.0.6? We have no new datamined information, but for all we know the class set items' value in the tables have simply been doubled (to make it on par with, e.g., Blackthorne's). That would not *exactly* mean that you can half all those numbers - but it probably reduced the expected shards by about 40% or so per set item. My guess. Haven't done the math. Thoughts?

Also, I'm wondering if anyone is gambling weapons currently - and could keep an eye on the wands/other weapons ratio to find out if it's really 99k (2:1 ratio) or higher (my guess). I'm totally sick of RiF right now, so not many shards incoming anytime soon that could help me to find this out, but maybe if I go back to rifting next week once football's over...

Oh, and completely forgot to add this to the guide. Will add it right now, way too interesting to not mention in the RiF section.

I am going to start gambling on weapons tonight and see how many it takes to get a Wand of Woh, right now I eat, sleep, breath, and dream about it, I can't get it out of my head.

Updated for 2.05 assuming that Kadala uses the same weightings that were datamined recently

Thank you very much for keeping this up to date. I wonder if it's possible to update this to 2.0.6? We have no new datamined information, but for all we know the class set items' value in the tables have simply been doubled (to make it on par with, e.g., Blackthorne's). That would not *exactly* mean that you can half all those numbers - but it probably reduced the expected shards by about 40% or so per set item. My guess. Haven't done the math. Thoughts?

Also, I'm wondering if anyone is gambling weapons currently - and could keep an eye on the wands/other weapons ratio to find out if it's really 99k (2:1 ratio) or higher (my guess). I'm totally sick of RiF right now, so not many shards incoming anytime soon that could help me to find this out, but maybe if I go back to rifting next week once football's over...

Oh, and completely forgot to add this to the guide. Will add it right now, way too interesting to not mention in the RiF section.

Ok, have done so, thanks for the appreciation and the link in your guide.
My information was the same that all of those class specific set items now have a 100 weighting instead of 50 before. That means they take less shards to get. However most legs have become more expensive that way, since the weighting-sums have increased.
Also, Blizz lied about all class set items dropping more frequently The Firebird Orb was at 100 before the patch, which results in a lowered drop rate.

And thanks for the data, SuchFear. From my personal WD-gambling experience I'd say the ratio is even worse than 1 in 3. But I'm too lazy to record my gambling results.

At first, I thought this was a great idea to explain how many in average 1 item is in Blood Shards, but I've come to the conclusion its a very bad idea actually and very misleading. Like, with Cindercoat your writing: "Expected shards to get one? 5000"

Pay notice to: "EXPECTED TO GET ONE".

Now, if you know anything at all about statistical math, its very very wrong. Stastics isn't intuitive.

I'l make an example:

So, we all remember when we had to farm MATERIALS before we could craft items such as Aughilds, Crimsons, Borns sets before they changed it. And we've all tried the randomness of this. I'll make an example from this, because it has a LOT higher drop chance than all of the stated items here.

Lets use Razormouth as an example, lets say we kill it on Torment 3 and it has 31 % chance to drop the material. Then at first, you'd think that your chance to get 1 material after 3 kills is 31 % + 31 % + 31 % = 93 % right? Well no, its far from that.

In fact, it goes like this. Lets say we have 25 groups of 4 people (a total of 100 people) who go and kill Razormouth.
- On 1st kill; about 31 (should really say between 25 and 36, but lets estimate here) gets the materials and 69 dont.
- On 2nd kill: now only 31 % of the last 69, who havent found one yet, gets one. So thats 21 more who get its than before. 69-21 = 48.
- On 3rd kill: Now 31 % of the last 48, who hasnt gotten one yet, should get one. So thats roguhly 15 more. Now thats 48-15 = 33.

So, in reality, only 100-33 = 67 % of the people got at least one material in 3 kills, when its 31 % drop chance! NOW ... Imagine this with drop chances such as 2,5 % on Tasker and Theo .... Wand of Woh 1,52 % .. Even cindercoat with 7 % would be tremendous to "at least get one". As you can see, there will always be a high ammount of people who "should" get unlucky. ESPECAILLY when its soooo low drop chances. I don't even wanna make the calculation, about how many wands you'd have to find to get at LEAST one. Because people think, that it must be 67, because there is 1/67 of wands being woh. But no, as I showed you with the materials, its actually waaaay more. This is because, there is people who get lucky and make up for it, like, there is some guys who find 2 or 3 materials in thoose 3 kills ! But of course, over 10000000 kills, it would average out.

I hope this helped someone to understand how fucked up RNG really is if they didnt already know

Average is accurate, expected is not.

No one knows how it works.... They may have wanted it to be 16% and they may have said it is unrelated. But I have noticed lots of times that me and my party are getting legs from same mobs / chests. This contradicts whatever they say.

Expected is exactly the right term, specifically when talking about probabilities. This was also discussed on the previous page. And this might be helpful as well.

Compared to Shard of Hate, Thunderfury is twice as likely and Azurewrath is 4 times as likely, so you need 50% or 25% respectively of those listed shards

10 times the amount of Thundergods

http://eu.battle.net/d3/en/profile/Sol77-2972/hero/66110450

From a first glance a few items that might be of interest as well: Burning Axe of Sankis (fire), Serpent Sparker (hydra wand), Gesture of Orpheus (arcane wand) and Sloraks Madness (disintegrate wand).

All in all though, great thread, and appreciate the work!

Also, I'm wondering if anyone is gambling weapons currently - and could keep an eye on the wands/other weapons ratio to find out if it's really 99k (2:1 ratio) or higher (my guess). I'm totally sick of RiF right now, so not many shards incoming anytime soon that could help me to find this out, but maybe if I go back to rifting next week once football's over...

Oh, and completely forgot to add this to the guide. Will add it right now, way too interesting to not mention in the RiF section.

Shards Spent- 1,500

Rares-91- 30 Wands

Blues-4-2 Wands

Legendary-1-Chantodo's Will

Ok, have done so, thanks for the appreciation and the link in your guide.

My information was the same that all of those class specific set items now have a 100 weighting instead of 50 before. That means they take less shards to get. However most legs have become more expensive that way, since the weighting-sums have increased.

Also, Blizz lied about all class set items dropping more frequently The Firebird Orb was at 100 before the patch, which results in a lowered drop rate.

And thanks for the data, SuchFear. From my personal WD-gambling experience I'd say the ratio is even worse than 1 in 3. But I'm too lazy to record my gambling results.

http://eu.battle.net/d3/en/profile/Sol77-2972/hero/66110450

Pay notice to: "EXPECTED TO GET ONE".

Now, if you know anything at all about statistical math, its very very wrong. Stastics isn't intuitive.

I'l make an example:

So, we all remember when we had to farm MATERIALS before we could craft items such as Aughilds, Crimsons, Borns sets before they changed it. And we've all tried the randomness of this. I'll make an example from this, because it has a LOT higher drop chance than all of the stated items here.

Lets use Razormouth as an example, lets say we kill it on Torment 3 and it has 31 % chance to drop the material. Then at first, you'd think that your chance to get 1 material after 3 kills is 31 % + 31 % + 31 % = 93 % right? Well no, its far from that.

In fact, it goes like this. Lets say we have 25 groups of 4 people (a total of 100 people) who go and kill Razormouth.

- On 1st kill; about 31 (should really say between 25 and 36, but lets estimate here) gets the materials and 69 dont.

- On 2nd kill: now only 31 % of the last 69, who havent found one yet, gets one. So thats 21 more who get its than before. 69-21 = 48.

- On 3rd kill: Now 31 % of the last 48, who hasnt gotten one yet, should get one. So thats roguhly 15 more. Now thats 48-15 = 33.

So, in reality, only 100-33 = 67 % of the people got at least one material in 3 kills, when its 31 % drop chance! NOW ... Imagine this with drop chances such as 2,5 % on Tasker and Theo .... Wand of Woh 1,52 % .. Even cindercoat with 7 % would be tremendous to "at least get one". As you can see, there will always be a high ammount of people who "should" get unlucky. ESPECAILLY when its soooo low drop chances. I don't even wanna make the calculation, about how many wands you'd have to find to get at LEAST one. Because people think, that it must be 67, because there is 1/67 of wands being woh. But no, as I showed you with the materials, its actually waaaay more. This is because, there is people who get lucky and make up for it, like, there is some guys who find 2 or 3 materials in thoose 3 kills ! But of course, over 10000000 kills, it would average out.

I hope this helped someone to understand how fucked up RNG really is if they didnt already know

EU Rank 1 Barbarian (Season 4):

>>> www.twitch.tv/Drahque <<<

>>> YouTube Guides <<<

>>> Patch 2.4: Guide for Season 5 <<<

>>> Patch 2.4.1: Guide for Season 6 <<<

This was discussed in this thread earlier. "Expected" is exactly the right term - especially in context of RNG.

http://en.wikipedia.org/wiki/Expected_value

But it's all in detail in the previous posts here.

Drop rates: https://docs.google.com/spreadsheets/d/16HCfzoijS_Xpzdrtg-C03b7oY97lb15XP5yztWNzH6Y/htmlview?usp=sharing&sle=true# (from http://www.reddit.com/r/Diablo/comments/2fxz4h/datamined_drop_rates_of_all_legendaries/ via diablofans news)