Something very wrong with chest drops

[Fcknuckle67]

In the past 2 weeks i have opened 100 chests and recieved 8 golden border cards. Nice yes? 6 were Legionairs armour. Thats a bit too suspect. There is something very screwy about chest drops and this will be the last i purchase. Which sort of works out since they will be cutting the free ones (paid for with in game currency gold) . If we are expected to pay real money for chests, i would expect a far more realistic spread of items, and maybe some transparency.

Edited September 23, 2016 by Fcknuckle67

[Hannibal_PJV]

Sound very normal, for player who has put a lot of money to ccg cards... you can get multible same rares and not some at all. So Sounds like it work as it should... not ideal, but belong among normal random distribution.

[Ethics Gradient]

8 Legendary Cards in 100 chests is right in the middle of the probability curve.  That much seems legit.

How the treasure picker actually selects the individual cards, well, that's some RNG magic that I don't have any data on.

 

Also, Obsidian isn't expecting you to pay cash for your treasure chests, they just proposed that in the future, IAPs purchased with gold won't get the 25% bonus.  Who knows if/when that change will even happen.  Regardless, you will still be able to buy 40 chests with 20k gold.

---

EDIT: Forgot to check my cross-tabs.

 

As the only Deck 2 Legendary, Legion Armor probably does have a higher-than-average chance of selection.  Yeah, 6 is a lot, but you're pretty likely to get at least 2 or 3 if you open a hundred chests.

---

Edited September 23, 2016 by Ethics Gradient

[firestormkirby]

Legionnaires Armour and Flask of Magic seem to have the highest drop rates. Both of them easily tripling the next most dropped legendary. I do think they need to work on their weighting system for what cards to drop.

[Ethics Gradient]

Legionnaires Armour and Flask of Magic seem to have the highest drop rates. Both of them easily tripling the next most dropped legendary. I do think they need to work on their weighting system for what cards to drop.

 

Yup.  Legion Armor happens to be the most likely legendary card right now.  Tied for second should be Flask of Magic and Amulet of Inescapable Location.

[Fcknuckle67]

Thanks Ethics Gradient, i hadn't read that correctly, about 500 gold/chest.

As to the probability of getting 6 Legion Armour in a purely random distribution out of 100 draws, and them being 6 out of 8 legendary cards, if you guys think that is "normal" or "fair", I'm obviously on the wrong planet. What is apparently more obvious is the pointlessness of raising the issue.

[_Harwin_]

Thanks Ethics Gradient, i hadn't read that correctly, about 500 gold/chest.

As to the probability of getting 6 Legion Armour in a purely random distribution out of 100 draws, and them being 6 out of 8 legendary cards, if you guys think that is "normal" or "fair", I'm obviously on the wrong planet. What is apparently more obvious is the pointlessness of raising the issue.

 

I think there are two issues here.

If Legion Armor is the only Deck 2 Legendary, and they try to distribute cards per-deck evenly (or even bias to lower decks so you're more likely to get a card that you can use) then it will have a higher priority.  It does feel wrong for it to be more likely than other legendaries, but it isn't necessarily wrong to distribute cards by deck (either evenly between decks or weighting some decks higher).  If they instead *first* decide to give you a legendary and then pick the card, then yeah it probably shouldn't be more common, but that would then start skewing the distribution of other cards between decks (for instance, if a deck had fewer than normal uncommons, then those cards would be less likely to come up, which might be bad for your treasure distribution.

 

The second issue has to do with the law of large numbers.  Flip 10 coins in a row and get heads all 10 times and you'll wonder if the coin is weighted.  But if you've got a game with a 10 thousand users(making up a number here), 10 of them will get those 10 coin flips.   I know this is why most attempts to gather "loot opening statistics" from games like this strongly encourage/require forum posters to decide *before* opening that they will post the data, so you don't only post the data that seems "wrong".  In the coinflip example, if you only decide afterwards, then nobody who gets between 3 and 7 heads is going to post, causing it to look like it's mostly going one way.

 

I'm not saying that the distribution algorithm *is* the best it could be, just trying to provide what I understand the context to be and why one poster might claim it is "normal" even though it's not something any of us would want to draw. 

[Longshot11]

Thanks Ethics Gradient, i hadn't read that correctly, about 500 gold/chest.

As to the probability of getting 6 Legion Armour in a purely random distribution out of 100 draws, and them being 6 out of 8 legendary cards, if you guys think that is "normal" or "fair", I'm obviously on the wrong planet. What is apparently more obvious is the pointlessness of raising the issue.

Somewhere around on the forum I read that the devs did indeed weight the drops so you can get Legendaries that "you can use" (i.e. up to the current AD# you're playing). I find this idea not a little absurd, and this could be the reason I got 4 Legendary Brooches of Shieldings in a row, sheesh. So, my advice would be , you wait until we get AD6 before you open your chests, so you don't get screwed by the weighted chance.

 

And yeah, some transparancy on these drops would probably be more appreciated by people ready to shell out real cash, but of course - if they knew how the current system works, they would probably all wait until AD6, or they're otherwise p(l)aying against the odds; such late payments obviously working against the devs. Then again, I don't presume to tell Obsidian how to monetize their app...

[Ethics Gradient]

Thanks Ethics Gradient, i hadn't read that correctly, about 500 gold/chest.

As to the probability of getting 6 Legion Armour in a purely random distribution out of 100 draws, and them being 6 out of 8 legendary cards, if you guys think that is "normal" or "fair", I'm obviously on the wrong planet. What is apparently more obvious is the pointlessness of raising the issue.

Oh, the distribution isn't even at all.  But the devs have hinted before that treasure cards are slightly weighted towards currently-playable Adventure Decks, and it seems to be reflected in my data.

As the only Deck 2 legendary (and there are no AD3 Legendaries at present), Legion Armor just has an increased likelihood of coming up.  Its the way it is.

 

Somewhere around on the forum I read that the devs did indeed weight the drops so you can get Legendaries that "you can use" (i.e. up to the current AD# you're playing).

I'd say that this very true. Legendary AD(B) through AD4 drop at about 4x the rate of AD5 and AD6.

[Ethics Gradient]

I'm not saying that the distribution algorithm *is* the best it could be, just trying to provide what I understand the context to be and why one poster might claim it is "normal" even though it's not something any of us would want to draw. 

I've got pretty decent sets of data at the moment.  Pulling together up a "treasure chest simulator" is next up on my to-do list.  For mostly academic reasons, I'm interested in determining how many treasure chests it would take to get one (or two) of each card. 

 

I'm not saying OP's results are good or fair, but a rough eyeballing of the numbers should put it somewhere in the top end of a standard deviation.