The Kadala “Double Legendary” Buff is broken
I started to doubt the drop rate was increased as claimed…
So I started logging data about the outcome of all my kadala gambles
In total:
751 Kadala Gambles were performed (searching for boots on a wizard)
70 legendaries were dropped
0 primals (which is the expected result)
So, How unexpected is this result?
Well, the total number of legendary drops is a binomially distributed random variable with N=751 and p=.2 (if you believe the drop rate was double from 10% to 20%)
This means we would expect
150.2 = Expected Number of Kadala boots drops in 751 gamble
-80.2 = How much we under performed = “actual boot count” - “expected boot count” 70 - 150.2 =
10.96 = Standard Deviation of “boot drop count” = sqrt(751* .2 * .8)
-7.317 = zScore = 80.2 / 10.96
The result of this experiment is 7.3 standard deviation below expectation if we believe the buff is active!!
Therefore, I have concluded the Kadala Buff is bugged/broken. The regular drop rate was 10%. The buffed drop rate is not 20%.
Raw data
starting shard count ( number of gambles) → num drops (what was gambled)
1056 (42 trials) → 4 (boots)
1040 (41 trials) → 1 (boots)
1049 (41 trials) → 2 (boots)
1058 (42 trials) → 8 (boots)
1042 (41 trials) → 1 (boots)
1051 (42 trials) → 4 (boots)
1035 (41 trials) → 4 (boots)
1044 (41 trials) → 6 (boots)
1044 (41 trials) → 6 (boots)
1053 (42 trials) → 5 (boots)
1037 (41 trials) → 4 (boots)
1585 (63 trials) → 7 (boots)
1960 (78 trials) → 4 (boots)
1909 (77 trials) → 8 (boots)
1960 (78 trials) → 6 (boots)
10 Likes
I concur. I have been playing every day since season launch and from the large quanity of bloodshards spent at kadala, I am no longer in doubt that the drop rate for boots is significantly lower than for other item slots. Of the many inventory windows that I have filled with purchases from kadala, I have recieved roughly 1-3 legendaries when buying boots, per window, subtracting base gem slots. This rarely happens when purchasing other item slots, I don’t actually recall this low count happening on another slot. Actually quite the opposite, one time, I had 3 empty slots after buying shoulder slot for all my bloodshards and salvaging yellows. Yes, I am normally counting empty slots/yellows after a full buy + salvage. This was tested on a DH, Monk and Wizard. I also have a suspition about rings too, but since they are more expensive and I care about my progress, I have not tested this enough to be sure.
1 Like
Rng is a fluid algorithm so the variables change. 20% today isn’t the same opening weekend. That’s my theory at least.
The law of large numbers - Essentially, with enough data, the randomness evens out and we can make probabilistic statements with a high degree of confidence. Some variables may change, but with enough of them, we can say something to a certain degree of precision about the constants.
I have more news to report
Different Gear Slots Have Very Different Kadala Success Rates
Conclusion: The Kadala Buff has been applied to some, but not all, gear slots.
Method
I tracked the following Kadala gambling. (In addition to my original post)
Boots
1054 “Kadala Trials”
92 Kadala drops
8.73% = Observed Success Rate
Gloves
692 “Kadala Trials”
130 Kadala drops
18.79% = Observed Success Rate
Chest Pieces
515 “Kadala Trials”
50 Kadala drops
9.71% = Observed Success Rate
95% Confidence Intervals
For each gear slot, lets estimate 95% Confidence Interval for its “Kadala Success Rate”.
Boots
8.73% = Observed Success Rate
0.87% = Standard Deviation
95% Confidence Interval for "Boot-specific Kadala Success Rate"
8.73% plus/minus 1.7%
7.03% to 10.43%
Gloves
18.79% = Observed Success Rate
1.48% = Standard Deviation
95% Confidence Interval for "Glove-specific Kadala Success Rate"
18.79% plus/minus 2.91%
15.883% to 21.7%
Chest Pieces
9.71% = Observed Success Rate
1.89% = Standard Deviation
95% Confidence Interval for "Chest-specific Kadala Success Rate"
9.71% plus/minus 3.71%
6.00% to 13.42%
Kadala is bugged … but its predictable and repeatable.
3 Likes
that explains the page of primal gloves in my vault 
I agree with Ivan’s observations. most main body gear is double legendaries. I noticed rings are absolutely not. I did not record the numbers, though I have clearly spent my full 1600 bloods on ring and came up flat or below 10% repeatedely
1 Like
Can confirm. Rings are completely a waste of time unless you just want to throw away blood shards in a single go for trash.
I can’t get primals from any mode of gameplay enough to get me where I need to be to finish the altar. I’m pissed off and annoyed at the number of gr’s I have done trying to do this garbage of a game system. Primals need a buff to drop rate if you are going to require them. Otherwise, put in a bad luck prevention system. RNG is notorious for being extremely punishing to people.
Not getting drops is how you turn a game into a chore and make it feel unrewarding. I don’t give an f about the ladder boards. Everyone at the top just cheats. I’d rather be playing last season with the sanctified power items and the altar. This is a huge disapointment to play as it is. Who the f cares if you get double primals when you go days with out seeing any primals before you finish the stupid altar.
2 Likes
I’m just a casual player, so I don’t have tons of data, but I did go through most slots and tried to get 200+ gambles on each (didn’t make it with weapons or amulets). Below are my results. Odds = odds of getting this many legendaries or fewer assuming a 20% legendary rate (computed using BINOM.DIST(legendaries,gambled,0.2,TRUE) in Excel). Unless indicated otherwise, gambles all occurred on my main, a Necro.
Slot Gambled Legendaries Success_rate Odds
Helm 201 44 21.8% 77.8%
Boots 237 26 11.0% 0.015%
Belt 200 31 15.5% 6.3%
Pants 200 44 22.0% 78.9%
Shield 207 47 22.7% 85.5%
Gloves 202 43 21.3% 71.1%
Chest 209 22 10.5% 0.018%
Shoulders 204 38 18.6% 34.9%
Bracers 215 34 15.8% 7.1%
1-hand Weap 43 11 25.6% 86.5%
Quiver 201 40 19.9% 52.8%
Orb 221 29 13.1% 0.50%
Mojo 210 58 27.6% 99.7%
Phylactery 209 40 19.1% 41.8%
Ring 223 23 10.3% 0.008%
Amulet 41 9 22.0% 70.4%
Ring(Wiz) 122 7 5.7% 0.001%
Helm(Wiz) 64 13 20.3% 59.8%
From this, and from reading what other people have posted, I am pretty convinced that Boots, Chest armor, and Rings are either not doubled, or were doubled but started at a lower legendary drop rate. (Since nobody has ever postulated that these slots drop legendaries at half the rate of other slots, I’m more inclined to believe that the buff just isn’t working on these slots.)
Orbs seem low, and Mojos seem high, but those numbers aren’t so far out of the realm of possibility that I’m convinced anything is amiss there.
Bottom line, I would avoid gambling rings, chest armor, or boots.
13 Likes
Good stuff Tobit.
Thanks for taking the time to collect and share the data
1 Like
I’d posted Kadala 2x legendary only affects char that unlocked node - #5 by Taphos-2526 also related to this. There have been a number of posts for this but no ack from blizzard that this is an issue.
1 Like