harsh, Narco!!!!

(this probably shows, but probability / combinatorics has never been strong suit. i'm a psychotherapist by profession. LOL)

Great, Math!

But +1 for a Nameless thread regardless, and +1 for a Nameless return sometime shortly, and +2 for several Piffle shoutouts!

As for your overall premise ... I dunno! I have tried to type out my thoughts on the subject three times now and have deleted it every time. I'm sleepy and nothing sounds right. Basically - I agree that the cost can be high, but I think the potential reward is worth it.

03/29/2013 08:49 PMPosted by PiffleI agree that the cost can be high, but I think the potential reward is worth it.

That's the only thing that really matters imo, at least to me.

And one got hist quad facta craft amulet in his 3rd tried.

RNG RNG RNG....

(chew the numbers)

RNG RNG RNG....

(chew the numbers)

03/29/2013 08:53 PMPosted by gotaplanstanI agree that the cost can be high, but I think the potential reward is worth it.

That's the only thing that really matters imo, at least to me.

I think that my point / idea is that it's only worth it if you already have gear that's close to being BiS. Essentially, I would discourage monks who are showing something like <200K DPS (sheet) from using crafting as a main source of upgrades... I just need to see if the math supports the idea.

for gearing my WD, just fyi...

I think that my point / idea is that it's only worth it if you already have gear that's close to being BiS. Essentially, I would discourage monks who are showing something like <200K DPS (sheet) from using crafting as a main source of upgrades... I just need to see if the math supports the idea.

I don't have the money, nor am making enough, to think about upgrading via AH. So that leaves me with crafting. I've already got 2 upgrades, and am working on replacing them (with Vit rolls) since they're both Dex rolls and I need more EHP to support a few changes I'd like to make.

Okay, so I'm trying to figure out some numbers want to see if this makes sense. Using Vrkhyz's numbers as a backdrop, I want to see if this makes any sense

I'm using http://www.mathsisfun.com/combinatorics/combinations-permutations-calculator.html to calculate the number of combinations you can make out of different... umm... combinations. lol.

So the number of ways you can show 1 affix amongst 5 affixes (in one roll) is 5

The number of ways you can roll 1 affix from 20 affixes available is 20.

So the probability of rolling 1 affix (amongst 20 available) in 5 slots is 5/20 = 25% or 1 chance in 4 [This is what Vrkhyz calculates too]

The number of ways you can combine 2 affixes amongst 5 affixes (in one roll) is 10

The number of combinations of 2 affixes amongst 20 available is is 190.

So the probability of rolling 2 affixes (amongst 20 available) in 5 slots is 10/190 = 5.26% or 1 chance in 19

[This is what Vrkhyz calculates too]

The number of ways you can combine 3 affixes amongst 5 affixes (in one roll) is 10

The number of combinations of 3 affixes amongst 20 available is 1140.

So the probability of rolling 3 affixes (amongst 20 available) in 5 slots is 10/1140 = 0.877% or 1 chance in 114

The number of ways you can combine 4 affixes amongst 5 affixes (in one roll) is 5

The number of combinations of 4 affixes amongst 20 available is 4,845.

So the probability of rolling 4 desired affixes (amongst 20 available) in 5 slots is 5/4845 = 0.00103% or 1 chance in 969

The number of ways you can combine 5 affixes amongst 5 affixes (in one roll) is 1

The number of combinations of 5 affixes amongst 20 available is 15,504.

So the probability of rolling 5 desired affixes (amongst 20 available) in 5 slots is 1 / 15,504 or 1 chance in 15,504

Given these numbers, it would take 114 crafts (assuming 20 affixes available) to craft a trifecta item, and even then, there will be no indication that the rolls would be even all that great. Assuming linear RNG calculations (more generous than a normal distribution), a top 25% roll for each of the rolls would come out (1/4 ^ 3) which works out to 1 chance in 64. Combine that with the 1 chance in 114 crafts, it would seem that 1 craft out of every 7,300 items would yield a decently rolled trifecta item (not perfect, just decently rolled all around).

Is this reasonable? Or is this completely non-sensical?

this sum complicated thread..

hurts my head..

hurts my head..

03/29/2013 08:49 PMPosted by PiffleBut +1 for a Nameless thread regardless, and +1 for a Nameless return sometime shortly, and +2 for several Piffle shoutouts!

Thanks Piffle. Sunday night is my targeted return time. Eagerly counting it down. :)

https://docs.google.com/file/d/0B0gMgiAJ3WsmN3lJWVFfVnJlVTQ/edit?usp=sharing

Pretty much explains how affixes are rolled and how to determine the probabilities of crafting them.

Using the numbers in there Mandlebarb actually make a simple program in Matlab to run a 10,000,000 amulet simulation to figure out the probabilities. I can't seem to find that output spreadsheet but I'm sure it shouldn't be too difficult to program into Matlab.

If you want to do the easy way though you can approximate it using equation 3:

http://imgur.com/XNhqWY1

And thanks for the feedback, guys. I just updated the original post to reflect corrections in the mathematics. Hope this is more accurate.

BTW, I will probably not be around for the rest of the day to address any feedback. But I hope to get to it when I get home later tonight. One more sleep before I can get back into the game! :)

As always, great thread. Thanks for your contribution.

I can't comment pertinently on the math myself (I'm a History major!), but the revised math seems to make intuitive sense when looking at it.

The only comment I would offer is an addendum to your short section as to why new(er) monks should not look at crafting as a primary source of upgrades (and you kind of said this in so many words, just wanted to make it very plain) : It is too difficult to guarantee the proper EHP stats.

When some guys come into the forums posting for gearing upgrade help and they get responses like "Keep crafting X and Y", I always cringe a little bit. That should ALWAYS be followed by a caveat that says, in my opinion, "Make SURE the EHP is maintained!" The truth is, when somebody comes to me to build a set and wants to craft away, I'll almost always say that's not a good idea. I'll just about always propose that he let me buy on ah first, so that I can guarantee the EHP stats that I want.

In short : on ah, you have the luxury of picking EXACTLY what you want, which is a huge advantage when building a set from scratch. Once the set is built, THEN guys can craft to their heart's content - but they had a solid base to start from.

------------

Now, a question. Are we absolutely sure that every affix has an equal chance of rolling? I realize my experience is singular and very subjective, but I would say I've crafted about 60 amulets before getting to the one I'm wearing now. In those 60, however, I've ended up with two more that are quite useable, with crit chc and crit dmg. One of them even had + armor. I've ID'd hundreds upon hundreds of rares and have only come up with ONE useable bifecta. The same story happened with gloves. This took me about 100 tries to get the ones I have, but I did tend to see damage modifiers much more often than I did when rolling simple rares. I realize that the sheer number of guaranteed affixes on crafted items has something to do with that, but still - is it a possibility that Blizz somehow biased some of the more desirable affixes?

All I got from this insanity is that I will add it to the sticky.

What is the formula for which section to put it in?