Alright, at the start of the beta I was looking through replays of other players that were streaming Hearthstone to collect data about the arena algorithm. However, due to a surgery and the start of college the project got put on hold and I don't have that much time to get going with it again. So I figured that I'll post what I have. It might not be that accurate due to a low sample size but it might be to use for something.
The sample data was taken from Darhoth and Ezm0ds first day's of Hearthstone Beta.
Now, on to the conclusions.
In all 7 games, card number 1,10,20 and 30 were always rare or higher. Probably the guaranteed “higher rarity cards” blizzard mentioned.
Of all 210 cards chosen in those 7 games: *A total of 23 cards were upgraded above standard rarity. (assuming rare is standard for 1,10,20,30) *Considering cards that were upgraded multiple times, there were actually a total of 30 upgrades.
Now, if we assume the chance of a rarity upgrade is flat, it would mean that there is a 30/210=14,28% chance of a card getting upgraded. Which means the chance for each card to become a certain rarity is.
Common to Rare = 14,28% chance
Common to Epic = 2,01% chance
Common to Legendary = 0.287% chance
Rare to Epic = 14,28% chance
Rare to Legendary = 2,01% chance
If these numbers are correct, it would mean that there is a.
68,15% chance of getting at least one epic in your deck.
14,43% chance of getting at least one legendary in your deck
0,9493% chance that you wont get any higher rarity cards than the 4 rares that you are guaranteed.
Now, this is a really rough calculation. To get more accurate data you would have to look at each individual case by itself and look how often it happens. This is necessary due to that the card pick algorithm might not be based on a flat chance as I assumed for my calculations. However, attempting to do that would require a much higher sample size than what I have. With just seven games it would be very easy for the calculations to be incorrect.
Just for an example, in the 7 games I logged there were 3 of them that had a legendary in it's deck. Even worse, one of the decks did not contain any other rarity besides the guaranteed rares, which according to my rough calculations has only about 1% chance of happening. The calculations would be totally different if I calculated it that way instead. Simply making it rough like this gives a better hint on the actual numbers.
Anyhow, as I said I don't know if these numbers are close to the real algorithm, but it might give a somewhat hint how it might be.
The packs you get from ending your arena experience are not related in any way to the cards that you choose to your arena deck at the beginning. Thereby the answer is no.
Well this does fit with what they said in one of their fireside duels. Something along the lines that there are certain "tiers" where you are guaranteed to draw something of a rare or higher.
Alright, at the start of the beta I was looking through replays of other players that were streaming Hearthstone to collect data about the arena algorithm. However, due to a surgery and the start of college the project got put on hold and I don't have that much time to get going with it again. So I figured that I'll post what I have. It might not be that accurate due to a low sample size but it might be to use for something.
The sample data was taken from Darhoth and Ezm0ds first day's of Hearthstone Beta.
Now, on to the conclusions.
In all 7 games, card number 1,10,20 and 30 were always rare or higher. Probably the guaranteed “higher rarity cards” blizzard mentioned.
Of all 210 cards chosen in those 7 games:
*A total of 23 cards were upgraded above standard rarity. (assuming rare is standard for 1,10,20,30)
*Considering cards that were upgraded multiple times, there were actually a total of 30 upgrades.
Now, if we assume the chance of a rarity upgrade is flat, it would mean that there is a 30/210=14,28% chance of a card getting upgraded. Which means the chance for each card to become a certain rarity is.
If these numbers are correct, it would mean that there is a.
Now, this is a really rough calculation. To get more accurate data you would have to look at each individual case by itself and look how often it happens. This is necessary due to that the card pick algorithm might not be based on a flat chance as I assumed for my calculations. However, attempting to do that would require a much higher sample size than what I have. With just seven games it would be very easy for the calculations to be incorrect.
Just for an example, in the 7 games I logged there were 3 of them that had a legendary in it's deck. Even worse, one of the decks did not contain any other rarity besides the guaranteed rares, which according to my rough calculations has only about 1% chance of happening. The calculations would be totally different if I calculated it that way instead. Simply making it rough like this gives a better hint on the actual numbers.
Anyhow, as I said I don't know if these numbers are close to the real algorithm, but it might give a somewhat hint how it might be.
I was watching Trump do Arena and in one game he got Two Sylvanas Windrunner. It was the dumbest.
Do the percentages change if you are winning the packs in Arena or buying the packs???
The packs you get from ending your arena experience are not related in any way to the cards that you choose to your arena deck at the beginning. Thereby the answer is no.
As for the likehood of getting a certain rarity card in a booster pack, regardless if you buy it directly or win one from the arena, it was discussed earlier in another topic.
http://www.hearthpwn.com/forums/hearthstone-general/theorycrafting/398-cards-rarity-chances-on-packs
That's really interesting information, though it feels like the sample size is really small.
Well this does fit with what they said in one of their fireside duels. Something along the lines that there are certain "tiers" where you are guaranteed to draw something of a rare or higher.
I must be really lucky then I just got 4 legendaries in my rogue deck