Treasure chest RNG has serious issues.

[Bullwinkle]

Do all cards of a certain rarity have an equal chance to drop within that rarity or not? Because (and I've noticed this before) the duplicate drops for certain specific cards are off the charts. I've been opening a bunch of chests, and out of the ~16 legendary cards I've got, 6(!) of them were Legion Armor, and 3 were Flasks of Magic. Does this seem reasonable? Seriously, this is infuriating.

[Ethics Gradient]

You're unlikely to get a dev response... but ..

 

Q: Do all cards of a certain rarity have an equal chance to drop within that rarity or not?

A: No.  Cards of equal rarities have nearly similar chances, but there are some probability variances nearing 0.01.

 

Q: I've been opening a bunch of chests, and out of the ~16 legendary cards I've got, 6(!) of them were Legion Armor, and 3 were Flasks of Magic. Does this seem reasonable?

A: Reasonable? Dunno.  Expected? Yes.  You probably have a couple Amulets of Inescapable location as well.  Legion Armor probably has the best odds of any legendary.

[nsr]

Yeah, it took me quite a while to get my first legion armor.  then i got 3 or 4 in a row.  I wonder what the probability distribution of the legendary cards are.

[Ethics Gradient]

I wonder what the probability distribution of the legendary cards are.

Currently mobile and away from my spreadsheets, but so far as I remember, it fell pretty neatly along the following lines before dice were introduced:

75% Uncommon

12% Rare

12% Epic

01% Legendary

 

The "one in four per chest redeemed" dice probabilitiy shuffled things around when it was introduced. Making room for an extra drop class seems to have had the effect of making some treasure cards have slightly increased/decreased probability compared to others in the same rarity tier.

 

I haven't seen enough data since AD5 to know if anything recently changed. Maybe I'll blast through some chests this weekend to find out.

[nsr]

The % change of legendary is different from the % chance of each legendary.  Are all legendary cards found at equal rates, or is the rate equally divided across the card types (armor/item/ally), or is it across the deck level?

[Ethics Gradient]

The % change of legendary is different from the % chance of each legendary.  Are all legendary cards found at equal rates, or is the rate equally divided across the card types (armor/item/ally), or is it across the deck level?

Pretty sure it used to be evenly divided pre-dice.  That, or it was close enough not to be particularly noticeable.  There were also fewer treasure cards then.

 

Post dice, the probabilities are nearly the same, but there appears to be some stratification by deck level.  Legion Armor likely stands out because it is the only Deck 2 Legendary.  If there is extra weighting given to the drop tables by deck number, that card would have better chances because it wouldn't have to "share" that statistical edge with other cards in the AD.

 

Even though it is a Legendary card, when you break everything down into the chances of acquiring a specific card when you open the four chests, Legion Armor is statistically somewhere in the Rare/Epic spectrum.  Granted, the odds of finding Legion Armor are still relatively small, but an extra 0.5-1% boost means you'd see it at 4 or 5 times the rate of some other legendaries.

 

[For the record, I'm not defending the system, or saying it is, or should be, fair.  That's just what the data points towards.  Also, haven't rerun any numbers since AD5.  I'll try to get around to another meta-gaming math project sometime soon. ]

[nsr]

Yes, that's what I was asking for, if the data suggests some type of uneven distribution.  Of course knowing this has no practical benefit, except to have more accurate complaints )  No point in complaining that the RNG for a D6 roll is bad if it rolls 1-6 evenly.  I trust that the developer's have gone and tested that extensively.  Instead, complain that out of the 6 possible outcomes, 4 of them are bad )

[Ethics Gradient]

Yes, that's what I was asking for, if the data suggests some type of uneven distribution.  Of course knowing this has no practical benefit, except to have more accurate complaints ) 

I'm not 100% certain, but the "unevenness" is likely an artifact of the dice introduction.  To compensate players for the fact that one out of four chests redemptions will give you a dice skin, legendary and epic cards got a little boost.  Some cards, like Legion Armor, just got boosted more than others.

[zarrick13]

I seem to recall on another post somewhere that the treasure boxes have some correlation to your progress too.  I don't know if this means your highest story box or just the one you played most recent, but treasure cards have a little bias where if you're supposedly chugging along in AD5, treasure boxes you open should contain a higher percentage of Deck 5 loot.

 

I don't remember if that was actually implemented or just how the devs said they wanted it to work though.

[Ethics Gradient]

I seem to recall on another post somewhere that the treasure boxes have some correlation to your progress too.  I don't know if this means your highest story box or just the one you played most recent, but treasure cards have a little bias where if you're supposedly chugging along in AD5, treasure boxes you open should contain a higher percentage of Deck 5 loot.

 

I don't remember if that was actually implemented or just how the devs said they wanted it to work though.

Everyone should be drawing from the same loot table.  What the devs did do is slightly bias the treasure tables towards the released Adventure Decks.

 

For instance, back in AD4, you were a tiny bit more likely to draw an AD1-AD4 treasure card than any AD5-AD6 card.  The rationale was giving a slight preference to Treasure Cards you could theoretically use over those you cannot.

Edited November 25, 2016 by Ethics Gradient