Hearthpwn

Well, that is some really detailed info with a decent sample size there.  Didn't expect the real values to be so closely tied to the disenchant values.

To add more data (while not as detailed as yours) I watched a replay of Lokazi's stream where he opened 112 packs (and another 8 off stream)

397 commons, 132 rares, 24 epics, 7 legendaries and a total of 24 golden cards of different rarities.

* 132 rares in 112 packs = about 1 rare per 0,848 packs
* 24 epics in 112 packs = about 1 epic per 4,667 packs
* 7 legendaries in 112 packs = about 1 legendary per 16 packs (but he did get 2 in a row though)
* 24 golden cards in 112 packs = about 1 golden per 4,667 packs

Comparing our values they seems quite similar with both being slightly higher than my original calculation. The legendary and golden card numbers are the one that stands out. Though, there is still a quite small sample size for those so errors are always possible.

I didn't write up the possible card rarity combos, but I noticed his biggest "total" rarity in a single pack were 1 epic, 3 rares and 1 golden common.
Slight surprice that it was possible to go without a rare (even if it was a golden common). Makes me wonder how the algorithm handles it.

As a bonus, after Lokazi opened all those 120 packs he got a total of 25 extra duplicate rares and 254 duplicate commons so he got another 1770 arcane dust from the mass disenchant. That would be enough to craft another legendary card.

Been looking around on the streams, and while I have not noticed any legendaries yet (other than Geblin) The "one epic per 5 packs" seems quite true at least.

Unless my memory fails me, at the Fireside chat Flux opened 5 packs and got one epic.
Xdega opened another 9-10, but only got one epic.
Ezm0d opened 15 boosters and got an epic on 5th, 7th and 15th pack.

Currently it holds true for most cases. So it's probably likely to be somewhere around that mark.
Anyone have seen/had some other experiences?

Quote from Flamedruid »

Erskow, that looks to be fairly solid reasoning and numbers, and would make sense. Remember however that you can get extra resources to buy packs from daily quests I think, so it wont be as many battles for booster packs. Unless you play 24/7, you will likely do quite a few daily quests in the estimated average of 21 packs. Say if you play 1 hour a day, twice, for 2 hours total, that is only 1 booster pack a day. So less time played, but more packs from daily quests if that makes sense.

The dailies were disregarded in the calculation I did as there is no info how big the reward is, would be a bit troublesome considering the time variable. Somewhere around 1 hour worth of gold would make sense though. But yeah, depending on how little you play the time could possibly be cut in half.

Quote from Rabidwhale »

They said that when you open a pack, the cards have a certain chance to upgrade to the next highest type (Rare, Epic, Legendary) and then they roll for each one after that if they get the first one.

So a common rolls to see if it is rare, then either the common doesn't become rare, or it does and it rolls again for epic and so on.

Quote from fr0d0b0ls0n»

The arena costed 3 packs, not one. Also we get at least a rare per pack, so the average of rares per pack will be higher than 1 for sure.

In the end, this was just some facts based on a fun calculation on the disenchant values and some old/new info that we knew of. The real rarity algorithm would probably not be as linear as this one. However, personally I feel that the values and facts makes somewhat scene and the real value will probably be somewhat close.

That arena cards used to cost 3 packs were news to me though, always thought it was just one pack. Makes sense though as you used to keep the cards during that time.

Quote from Ferrarista »

Also, one of the reasons they changed to gold over just packs, was to reward even the loser of the match, so they can progress (though not for the same amount as the winner).

It does not necessary mean that the average values have changed (it neither means that the weren't changed). The gold values for each game might have been balanced to give enough for a pack after 8 games with a 50% win ratio (if Team 5 now found it to be a suitable value). The difference being that if you are on a losing streak you feel that you are still moving slowly towards the goal (more cards) compared to staying in one spot and getting mad on not winning.

Quote from danilodn »

So, if we are in the right way with this, to craft a Legendary that we want, we will need to spend around 84 dollars (4 * 21).

To craft an Epic, 20 dollars (4 * 5).

Well, if the values I came up to somehow would be legit, you would get one RANDOM legendary AND 104 other cards of different rarities.
If you want the values for crafting (that we disenchant all other cards than the specified rarity or above) we would get...

* An average of 1 random rare + 1 crafted rare per pack
* An average of 1 random epic + 2 crafted epics per 5 packs
* An average of 1 random legendary + 3 crafted legendaries per 21 packs

Not sure what values you used for the real money. Is it the cost of a regular TCG pack or did I miss some gold conversion blue post?

Quote from danilodn »

No one wants to try to connect the chances to the costs of creating the cards by arcane dust?

Not likely to be so, but for those that want to believe.

 Considering the disenchant values

* 1 rare card is worth similar to 4 commons
* 1 epic card is worth similar to 5 rares, or 20 commons
* 1 legendary card is worth similar to 4 epics, or 20 rares, or 80 commons

Now considering there are 5 cards per pack..

* 1 rare per average 5 cards = An average of 1 rare per pack
* 1 epic per average 25 cards = An average of 1 epic per 5 packs
* 1 legendary per average 105 cards = An average of 1 legendary per 21 packs

EDIT: Figure ill keep ranting/speculating.

We still don't know anything special about gold generation (other than 100 gold per pack and a "possible" 15 gold for a forge 1 win play, which all could possibly change later on.) However, before gold they had a system that whenever you won 4 times you were rewarded with a booster pack (shown in the first stream). Considering all those things and an average win ratio of 50% this would mean a number of things.

• You would need to play an average of 8 games to receive a booster pack
• Converted to gold this would mean that you get an average of 100/8=12,5 gold per duel (close to the 15 gold from forge arena)
• Considering Arena used to cost a booster pack to play, the cost for arena should be somewhere around 100 gold.

• As for time, according to a blue post playing a duel takes an average of 10-15 minutes. With 8 games this would mean 80-120 minutes per booster pack.
  
  • If we combine that to the rarity calculated earlier, this would mean that.
    * An average of 1 rare per pack = 1 rare per 80-120 min = 1,5-2 hours
    * An average of 1 epic per 5 packs = 1 epic per 400-600 min = 7-10 hours
    * An average of 1 legendary per 21 packs = 1 legendary per 1680-2520 min = 28-42 hours

   

So yeah this is all some mashup with old and new numbers. Even if true or not, would requiring to play the game for 7-10 hours to get an epic or 28-42 hours to get a legendary, not to mention 1,5-2 hours for a booster, be a valid number for blizzard to use or would it be too high/low? Whats your opinion?