Hearthpwn

In response to http://www.hearthpwn.com/members/Irealys

Who stated:  "if you understood statistics you would understand it is a result of randomness. What you seem to be hoping for is even distribution which would be the DIRECT OPPOSITE OF A RANDOM SYSTEM."

I am not commenting or complaining about the rate that legendary cards appear, I am complaining about the rate of duplicate cards. Guess we DO need statistics to understand this. 

Nope. That is not how a random system works, or how one determines a random system.  The higher the sample size (in this case a number of packs opened), then the closer we get to an understanding of a distribution system, and it becomes more evenly distributed... that is exactly what a random distribution system with a large sample size is. This would not be an even distribution, but a 'bell curve',  https://upload.wikimedia.org/wikipedia/commons/thumb/2/25/The_Normal_Distribution.svg/800px-The_Normal_Distribution.svg.png  . Given that the legendary cards lie to the far right of this curve (as you see it), the number of legendary received would be quite low. Receiving 4 of the same identical legendary in 120 packs (600 cards) is astronomical.  Perhaps I am the guy who could have won the lottery... and that would be random, but the statistical odds are very very very low that I would receive the same legendary card 4 times. 

Let's use this data: http://hearthstone.gamepedia.com/Card_pack_statistics

Let's find a real Statistical P-value and do this right.   Let the null-hypothesis be "I don't understand statistics, and this i is a true random system". The tested hypothesis is "This shit is broken, and I might have a limited understanding of statistics".

This states that the rate of a 1.10% Legendary per pack.  We have 120 packs, so 600 cards X 1.01%= 6.06 legendary drops. Number of legendary cards in the expansion is: 23.    So we can expect 6 cards with a 1/23 chance for each card for 120 packs.  A 26% chance for each legendary to show up.   Expected value= 26%. Actual value of the number of Clucthmothers x 4 = 66%. (4/6) (I got the promised 6 legendary cards, but 4 were duplicates).

Lets do a Chi-squared test of this ((66-26)^2)/26=  61.54  (our p-value).

Sample size: 600 cards

Chances of this happening in a completely random system: less than 0.005%.  My T-table couldn't go lower than that.

This is the "DIRECT OPPOSITE OF A RANDOM SYSTEM".  But thanks for getting me to remember my stats homework bro!

Edited to take out a '9' that was meant to be a parenthesis :)9.

Maybe I am the outlier in this case, anecdotally there seem to be many others with the same situation, but perhaps that is because those of us who are unlucky are more vocal.  I would like to see some more transparency in the Blizzard system of pack opening.

This is bs.... 57 packs 122 common cards dusted... Do the math, let that sink in...

I think the complaints about the amount of legendaries gotten is actually a side effect of the expansion being really good. People want ALL the cards!

You are wrong. The complaint regarding the duplicate of legendaries is due to last expansion issue with tri-class cards (which being part of 3 classes appeared 3 times as much as normal cards... I wonder how that happened) and since Blizzard offered free packs as a result people jump to conclusion.

The people complaining (at least on the official forums, not sure if it is the same here) about how this game "forces them to buy packs" is because for once there isn't a clear cut 3 to 5 legendaries you should have and instead made class legendaries unique and/or interesting. So basically those people complain that the game will be more diverse and that they have options...

I opened about 120 packs... 50 from pre-order and 60 from buying them. I am now the proud owner of 4 Clutchmother Zavas. Got 2 other legends. You don't need statistics to understand that this is not a random system.  Coincidentally (or not), I have been playing Discolock heavily for the last season. This might be a case of misplaced benevolence on the part of Blizzard, in which they identified the decks and classes that a player uses most frequently and then apply a filter to provide them with the cards they prefer. The odds of pulling an identical legendary 4 times in 120 card packs (600 cards) is as close to impossible as Blizzard acknowledging that there is a problem with this. 

Quote from Shazaam1 >>

I opened about 120 packs... 50 from pre-order and 60 from buying them. I am now the proud owner of 4 Clutchmother Zavas. Got 2 other legends. You don't need statistics to understand that this is not a random system.  Coincidentally (or not), I have been playing Discolock heavily for the last season. This might be a case of misplaced benevolence on the part of Blizzard, in which they identified the decks and classes that a player uses most frequently and then apply a filter to provide them with the cards they prefer. The odds of pulling an identical legendary 4 times in 120 card packs (600 cards) is as close to impossible as Blizzard acknowledging that there is a problem with this. 

"You don't need statistics to understand that this is not a random system." No, if you understood statistics you would understand it is a result of randomness. What you seem to be hoping for is even distribution which would be the DIRECT OPPOSITE OF A RANDOM SYSTEM.

The way their pack works is exactly like any other pack of cards from any card collecting (game or not) works. You open a pack and you get random shit. The system doesn't care about what other cards you have. Hell Blizzard has a pity timer which is more than normal pack opening would do.

Maybe I am the outlier in this case, anecdotally there seem to be many others with the same situation, but perhaps that is because those of us who are unlucky are more vocal.  I would like to see some more transparency in the Blizzard system of pack opening.

I am not commenting or complaining about the rate that legendary cards appear, I am complaining about the rate of duplicate cards. Guess we DO need statistics to understand this. 

Nope. That is not how a random system works, or how one determines a random system.  The higher the sample size (in this case a number of packs opened), then the closer we get to an understanding of a distribution system, and it becomes more evenly distributed... that is exactly what a random distribution system with a large sample size is. This would not be an even distribution, but a 'bell curve',  https://upload.wikimedia.org/wikipedia/commons/thumb/2/25/The_Normal_Distribution.svg/800px-The_Normal_Distribution.svg.png  . Given that the legendary cards lie to the far right of this curve (as you see it), the number of legendary received would be quite low. Receiving 4 of the same identical legendary in 120 packs (600 cards) is astronomical.  Perhaps I am the guy who could have won the lottery... and that would be random, but the statistical odds are very very very low that I would receive the same legendary card 4 times. 

Let's use this data: http://hearthstone.gamepedia.com/Card_pack_statistics

Let's find a real Statistical P-value and do this right.   Let the null-hypothesis be "I don't understand statistics, and this i is a true random system". The tested hypothesis is "This shit is broken, and I might have a limited understanding of statistics".

This states that the rate of a 1.10% Legendary per pack.  We have 120 packs, so 600 cards X 1.01%= 6.06 legendary drops. Number of legendary cards in the expansion is: 23.    So we can expect 6 cards with a 1/23 chance for each card for 120 packs.  A 26% chance for each legendary to show up.   Expected value= 26%. Actual value of the number of Clucthmothers x 4 = 66%. (4/6) (I got the promised 6 legendary cards, but 4 were duplicates).

Lets do a Chi-squared test of this ((66-26)^2)/26=  61.54  (our p-value).

Sample size: 600 cards

Chances of this happening in a completely random system: less than 0.005%.  My T-table couldn't go lower than that.

This is the "DIRECT OPPOSITE OF A RANDOM SYSTEM".  But thanks for getting me to remember my stats homework bro!

 Edited to replace a 9 and 0 with parenthesis as intended. 

http://www.polygon.com/2017/4/10/15247906/hearthstone-journey-to-un-goro-free-packs-pack-problems-too-few-legendary-rarity

And on the other end, i never played Druid, my one and only class that's not level 60, and yet, i open only Tyrantus from 50 packs. Stop talking bullshit  guys, it's random, just some people are more lucky than the others. Is it that hard to understand ?

87 Packs opened, 3 Legendary (1 Quest). $50 inverted plus 3700 in Gold.

And I need open like 544.5 Packs for craft me the missing cards, because the % get it in Packs is ridiculous.

yes no problem with the pack sure just 1 pack 3 wasps and 10. same rare card, yeah this is normal.

pls don't waste your money on this game. if you want a nice deck need to spend 100 dollar... ASTOUNDING !

What if it turns out we all just have really bad luck 

When you end up with 7 of the same Rare card and the closest competitor is 4 copies of several different Rares after opening 58 packs, something is wrong with the distribution. Look harder Blizzard.

I feel your pain, 87 packs and two legendaries-the same one no less-at least I like warrior.

Quote from Khazeroth >>

I feel your pain, 87 packs and two legendaries-the same one no less-at least I like warrior.

I don't normally complain about these things and I'm not even going to complain about the frequency of Epic and Legendary cards.  My big issue at the moment is the frequency of duplicate cards.  During my pack opening I managed to open 2 Legendary cards yet they were the same card.  Out of 23 possible cards they could have been, it was the same card both times.  A few days later I decided to craft a Legendary.  I just knew deep down inside that crafting this Legendary is going to guarantee that it's the next one I'll open in a pack.  Sure enough, the next Legendary is the same one I just crafted that same day.  Makes it hard to justify actually spending money on packs instead of buying with gold since it would take well over $800 to open every card in the set and with 3 sets releasing each year it's just not sensible unless you make a living streaming Hearthstone.

Opened 131 packs, got 3 Clutchmothers aswell!

120 packs... 4 Clutchmothers... Question: Do you all play Warlock (specifically Discolock) often? I do. Perhaps that influenced the packs?