All users will need to merge their Diablofans account with a new or existing Twitch account starting Nov 20th. You can merge your accounts by clicking here. Have questions? Learn more here.

I've been farming keys for the Infernal Machine and don't know how can Blizzard say that MP5 gives you 50% chance of the keys to drop.

Always using 5 stacks of NV, Paragon 29, MP5, enchantress gives me +30% MF, I've killed Keywarden Act 2 about 30 times and I've got 1 key, yes 1 key, that 3,3% chance to drop not 50%.

I always span in Vault of the Assassin, clear the level that gives me 5 NV, teleport to town, spawn in Dalgur Oasis and search for the Keywarden, that means another 2 or 3 elites killed in the Oasis before the Keywarden.

50% is 50%. You can go 1000 runs, theoretically, though very improbably, on MP9 and get 0 keys. Your previous games have no impact at all on the chance of it to drop. Math is fine.

Yeah, the game should drop 0.5 keys at MP5, 0.05 keys in MP0, then there should be a plan to combine it Problem solved, and takes RNG out of the game...

my current record is 3 mp8 run back to back with no key drop. the only thing i can say is FML. And i dun think mf affect the drop rate

for act 2 u can try black canyon mine + stinging winds+ khasim output with one dungeon. it will give u more than 5 nv. This is much better than vault since u dun need to find the place in desolate sand each time u change the map.

I've been farming keys for the Infernal Machine and don't know how can Blizzard say that MP5 gives you 50% chance of the keys to drop.

Always using 5 stacks of NV, Paragon 29, MP5, enchantress gives me +30% MF, I've killed Keywarden Act 2 about 30 times and I've got 1 key, yes 1 key, that 3,3% chance to drop not 50%.

You don't know MF doesn't play a role

You didn't kill him 30 times and get just 1 key - you're either exaggerating or just flat out lying

You think the odds of getting 1 key out of 30 at MP5 is 3.3%

When dealing with experiments that are random and well-defined in a purely theoretical setting (like tossing a fair coin), probabilities describe the statistical number of outcomes considered divided by the number of all outcomes (tossing a fair coin twice will yield HH with probability 1/4, because the four outcomes HH, HT, TH and TT are possible). When it comes to practical application, however, the word probability does not have a singular direct definition. In fact, there are two major categories of probability interpretations, whose adherents possess conflicting views about the fundamental nature of probability:

Objectivists assign numbers to describe some objective or physical state of affairs. The most popular version of objective probability is frequentist probability, which claims that the probability of a random event denotes the relative frequency of occurrence of an experiment's outcome, when repeating the experiment. This interpretation considers probability to be the relative frequency "in the long run" of outcomes.[2] A modification of this is propensity probability, which interprets probability as the tendency of some experiment to yield a certain outcome, even if it is performed only once.

Since the key drop is based in an algorithm, that is affected by a scaling factor (MP), first is not truly random secondly it should take in consideration the account stats.

If the probability of winning is not related to the odd of that event happening then why all the casino games, lottery and other games are based in many millions to one probability against the players???

I think many of the people who wrote above that need some math lessons.

The only way to know if the MP level really scales correctly was to know the key drop rate vs the MP level of each player, that would be interesting.

And yes 30 F***** runs, just 1 key. And yes I might be just unlucky.

You fail here, as nothing you just quoted, is relevant to what occurs. The drop chance is 50%. It is 50% every time, no matter how lucky or unlucky you were the run before. You can aproximate a drop chance from sample data, however, the sample data needs to be huge. 30 is not even slightly close to a huge number. Therefore, you were just unlucky and there is no need to make a post about, as everyone is getting a bit tired of people whining about how unlucky they are.

Blizzard's math skills are actually better or equal as yours.

Always using 5 stacks of NV, Paragon 29, MP5, enchantress gives me +30% MF, I've killed Keywarden Act 2 about 30 times and I've got 1 key, yes 1 key, that 3,3% chance to drop not 50%.

This is a perfect example of just how much of a statistical error you can get, if you have small sample of data. Run another million MP5 runs and I guarantee you, it will be much, much closer to 50% (not precisely, something like 49,95% chance of dropping).

To explain it in simplified form, strings of bad luck have a small chance to occur - for example flipping tails on a coin 5 times in a row has only 3,125% chance to occur -, therefore they tend to create a smaller portion of your estimation the larger your sample of data gets. The closer you get to infinity, the more precise is your estimation. 30 is pretty far away from infinity, isn't it?

However, true randomness is kinda mindblowing. Because if each event is independent, it's entirely possible for someone to experience killing key warden on inferno MP<10, with 5NV, INFINITE number of times and NEVER see a key-drop. We can even calculate the probability of this to happen. The chance is... ε%... looks weird, it means the chance is infinitesimaly small. Or aproaching zero, but not quite zero.

Since the key drop is based in an algorithm, that is affected by a scaling factor (MP), first is not truly random secondly it should take in consideration the account stats.

Most PRNG algorithms used now are good enough that you never have to worry about them being a problem. See Mersenne Twister for example.
What account stats should it take in consideration? Key drop is linked to one thing only - MP level.

If the probability of winning is not related to the odd of that event happening then why all the casino games, lottery and other games are based in many millions to one probability against the players???

I love this thread. There are two sides attacking the OP, but they are also unwittingly attacking each other.
One side says that it's normal to kill the warden 30 times at MP5 and have only 1 key drop; the other side says that the OP is lying, because they think that killing the warden 30 times at MP5 and only getting 1 key is almost impossible.

Always using 5 stacks of NV, Paragon 29, MP5, enchantress gives me +30% MF, I've killed Keywarden Act 2 about 30 times and I've got 1 key, yes 1 key, that 3,3% chance to drop not 50%.

I always span in Vault of the Assassin, clear the level that gives me 5 NV, teleport to town, spawn in Dalgur Oasis and search for the Keywarden, that means another 2 or 3 elites killed in the Oasis before the Keywarden.

Keys in Act 3 and Act 1 seem more likely to drop.

MF - also irrelevant.

Ha. Bagstone.

So I'm sorry to say that you have bad luck and worse math skills (or is it the other way around?) :)!

Flip a coin again, 50/50 chance of it landing heads up

Put on your full MF set, level to paragon 29, Load MP5, get 5 stacks of NV and flip a coin.

Still has 50/50 chance of landing heads up.

You can do 1000 runs and each run could end up dropping no keys

read it, think about it, think about it again, come to the conclusion you were wrong, leave the thread with some dignity left

That's what i thought too. it is rather improbable that he is telling the truth.

for act 2 u can try black canyon mine + stinging winds+ khasim output with one dungeon. it will give u more than 5 nv. This is much better than vault since u dun need to find the place in desolate sand each time u change the map.

When dealing with experiments that are random and well-defined in a purely theoretical setting (like tossing a fair coin), probabilities describe the statistical number of outcomes considered divided by the number of all outcomes (tossing a fair coin twice will yield HH with probability 1/4, because the four outcomes HH, HT, TH and TT are possible). When it comes to practical application, however, the word probability does not have a singular direct definition. In fact, there are two major categories of probability interpretations, whose adherents possess conflicting views about the fundamental nature of probability:

Objectivists assign numbers to describe some objective or physical state of affairs. The most popular version of objective probability is frequentist probability, which claims that the probability of a random event denotes the relative frequency of occurrence of an experiment's outcome, when repeating the experiment. This interpretation considers probability to be the relative frequency "in the long run" of outcomes.[2] A modification of this is propensity probability, which interprets probability as the tendency of some experiment to yield a certain outcome, even if it is performed only once.

Since the key drop is based in an algorithm, that is affected by a scaling factor (MP), first is not truly random secondly it should take in consideration the account stats.

If the probability of winning is not related to the odd of that event happening then why all the casino games, lottery and other games are based in many millions to one probability against the players???

I think many of the people who wrote above that need some math lessons.

The only way to know if the MP level really scales correctly was to know the key drop rate vs the MP level of each player, that would be interesting.

And yes 30 F***** runs, just 1 key. And yes I might be just unlucky.

P.S. I killed each keywarden on MP3 and I got one key from each keywarden. Obviously the drop rate is 100% on MP3 lulz. IT'S BROKEN! BLIZZ SUCKS!

Blizzard's math skills are actually better or equal as yours.

This is a perfect example of just how much of a statistical error you can get, if you have small sample of data. Run another million MP5 runs and I guarantee you, it will be much, much closer to 50% (not precisely, something like 49,95% chance of dropping).

To explain it in simplified form, strings of bad luck have a small chance to occur - for example flipping tails on a coin 5 times in a row has only 3,125% chance to occur -, therefore they tend to create a smaller portion of your estimation the larger your sample of data gets. The closer you get to infinity, the more precise is your estimation. 30 is pretty far away from infinity, isn't it?

However, true randomness is kinda mindblowing. Because if each event is independent, it's entirely possible for someone to experience killing key warden on inferno MP<10, with 5NV, INFINITE number of times and NEVER see a key-drop. We can even calculate the probability of this to happen. The chance is... ε%... looks weird, it means the chance is infinitesimaly small. Or aproaching zero, but not quite zero.

What account stats should it take in consideration? Key drop is linked to one thing only - MP level.

What are you even trying to imply?

Yes, you are unlucky - if you believe in luck.

One side says that it's normal to kill the warden 30 times at MP5 and have only 1 key drop; the other side says that the OP is lying, because they think that killing the warden 30 times at MP5 and only getting 1 key is almost impossible.

Maybe side #1 should start attacking side #2.

My US profile