Darkmoon trinket creation randomness.

This post is part 2 of a series.  In my last post, I talked about how trinket making had 90% profits, with some large caveats.


This post is looking at the randomness of how many cards it takes to make a trinket.  The preious post assumed cards would come in order and you would be left with no wasted cards.  Umm, that was optomistic.  I would allow for 4%-8% wasted cards.


The basic process again is :

• Gather herbs and volatile life
• Mill herbs -> Inks (Blackfallow & Inferno)
• Combine Inferno inks and Volatile life to create an individual card (Ace to Eight of one of the 4 standard card suites)
• Combine 8 different cards of the same suite to create a deck. 
• Turn the deck into a trinket during the Darkmoon Faire.

One idea bandied about is that you need to craft 32 cards on average to make your first trinket; assuming you don't buy individual cards from the AH.  I think the number is closer to 50, based on this google docs simulator.  Even after making 10,000 cards, my simluator suggests that I will have about 1150 trinkets and about 800 (8%) unused cards.  Some readers may make 1150 trinkets.  I doubt I will.

The individual cards can be bought and sold on the AH.  Conventional wisdom says that if you have a surplus of one card (eg 8 of the winds), but need a different card (say 3 of the winds), you will sell your surplus card on the AH, and try to purchase the card you need.

• In my 50 card scenario, trading 6 cards allows me to create another 3 trinkets (using 24 cards); up from 1 trinket using 8 cards.
• In my 10,000 card scenario, trading 30 cards allows me to make 30 more trinkets, using up 240 of the 800 surplus cards.
• However, you have also potentially allowed your competitors to do the same, meaning increased competition for completed trinket sales.

If I only have one (or possibly two depending on my mood) extra card, I will sit on it.

I will sell surplus cards where :

• I have a few extra cards, and they are listed above 1/8 of the completed trinket sale price; or
• I have a lot of extra cards.

I will purchase cards where :

• I need them to complete a deck, and they are at or below market prices
• I need a lot of one card to complete multiple decks.
• The are significantly below completed trinket price and I only have a few of that card

To the peanut gallery, for this post, I am specifically looking for debate on:

• your observations on the randomness of cards, and/or
• selling/purchasing surplus/shortfall cards.

In answer to yesterday's comment : Cold has been trying to get out of darkmoon trinkets, and has some cards he can not shift.  According to my simulation, there will always be cards that are surplus to a servers requirements.  If you want out, liquidate, but try both faction's AH.


Postscript in response to Zamboni 
(It's a bother to put tables in comments, so I have added a postscript.).
Random chance is enough to explain the shortage in cards.  Statistics are funny like that.

You say you have made hundreds of trinkets, lets call it 400, and see what my simlator suggests.
I ran it about 20 times making 3660 cards ~ 1/2 gave more than 400, ~ 1/2 than 400 and one gave exactly 400.

NumDecks	110								Total
NumCards	130	131	110	112	110	147	114	121	975
Suite 1	A	2	3	4	5	6	7	8
NumDecks	95
NumCards	102	95	129	125	99	123	104	116	893
Suite 2	A	2	3	4	5	6	7	8
NumDecks	96
NumCards	117	120	96	122	128	105	99	100	887
Suite 3	A	2	3	4	5	6	7	8
NumDecks	99
NumCards	107	99	140	116	119	102	106	116	905
Suite 4	A	2	3	4	5	6	7	8
									3660
Total decks	400
Consumed cards	3200
Surplus cards	460

As you can see, I am short (by a lot) fo the following cards
Suite 1 : 3, 4 & 5
Suite 2 : 2 & 5
Suite 3 : 3 & 7
Suite 4 : 2 & 6.

The most of any given card was 147 (Suite 1 card 6).  The fewest of any given card was 95 (Suite 2 card 2). 

The most of any deck was Suite 1 with 110 decks, and the fewest was suite 2 with 95 decks.

To show a bias in the RNG over 400 decks you need to show a difference significantly more than 52 cards.  Until then, it's just random noise.