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.

You do realize there are a lot more affixes that are rolled every time an item drops besides "usable" and "trash" right? As a poster above stated, there's maybe a best-case scenario of 5% of the items you get being "usable" based on all the possible affixes each item can potentially have.

If the RNG on diablo 3 is truly random then it should be impossible to get two of the same items with exact stats,name,gold sell price and durability form 2 different mobs, well folks a friend of mine did.

That's like saying it should be impossible to roll 6 in a dice twice... Although that does sound like a bug.

If the RNG on diablo 3 is truly random then it should be impossible to get two of the same items with exact stats,name,gold sell price and durability form 2 different mobs, well folks a friend of mine did.

no not really, even in truly random scenarios you can get 2 equal outcomes, even in a row at times

That only happens if the loot table is from 1 to 10, the loot in this game is billions, it may be possible to get the same item some time down the road but not that soon. Also it seems they are just reusing the random hashes.

This is from my own observation with 200+ hrs of farming in Inferno act 2 and data read from this forum.
It seems that of all the rare drops we get for iLvl61 to iLvl63, only a small number of the items are usable in inferno. I don't exactly have a count but it seems less than 5% of the items are usable. This bring me to conclude that the RNG in D3 is a rigged system where Blizzard only allowed a small number of usable items for the end game. If it were true RNG then the ratio for "usable vs trash" would be close to even.

Note: Usable is an item that someone would use in any of the Act in inferno. Trash is an item someone would just vendor or salvage.

Is this old info? If not, can someone help me backup this claim? Is this how the drop system works?

I'm at a point where I'm frustrated with this game. Hours of doing the same thing over and over again in hopes of finding an upgrade is boring me to death.

This makes zero sense, since you are basing something mathematical (affixes and their ranges) with something conjectural (what YOU would consider useful).

This is from my own observation with 200+ hrs of farming in Inferno act 2 and data read from this forum.
It seems that of all the rare drops we get for iLvl61 to iLvl63, only a small number of the items are usable in inferno. I don't exactly have a count but it seems less than 5% of the items are usable. This bring me to conclude that the RNG in D3 is a rigged system where Blizzard only allowed a small number of usable items for the end game. If it were true RNG then the ratio for "usable vs trash" would be close to even.

Note: Usable is an item that someone would use in any of the Act in inferno. Trash is an item someone would just vendor or salvage.

Is this old info? If not, can someone help me backup this claim? Is this how the drop system works?

I'm at a point where I'm frustrated with this game. Hours of doing the same thing over and over again in hopes of finding an upgrade is boring me to death.

U think usable vs trash should be even?
Look how many friggin stats there r.
So first, u need to have the stats u want 2-6 stats on an item depending what you are looking for. Then it has to have high/good values. So rng stays rng, no conspiracy theories.

Also what you consider to be garbage trash might be considered decent for someone just starting in act 1 inferno. You say you farmed 200 hours act 2, then its no friggin wonder u dont find upgrades cause u have most likely found them earlier.

if an item has a chance to have dps ranging between 250 and 1300 and people consider 1100+ as a great usable item, then of course you arent going to see 50% of your items as "great usable" items you are going to see like 5% good and 95% crap

If the RNG on diablo 3 is truly random then it should be impossible to get two of the same items with exact stats,name,gold sell price and durability form 2 different mobs, well folks a friend of mine did.

no not really, even in truly random scenarios you can get 2 equal outcomes, even in a row at times

That only happens if the loot table is from 1 to 10, the loot in this game is billions, it may be possible to get the same item some time down the road but not that soon. Also it seems they are just reusing the random hashes.

I believe there was(is?) a bug which causes the same item to drop twice in a row from different, subsequent packs.

If it were true RNG then the ratio for "usable vs trash" would be close to even.

Sorry but no, the game isn't rolling "usable vs trash", it is randomly rolling the ilvl, the affixes number, the affixes themselves (from a pool of many useless ones) and, in the end, the actual values.

It is quite clear, statistically, that the vast majority of items will be vendor trash.

If it were true RNG then the ratio for "usable vs trash" would be close to even.

Sorry but no, the game isn't rolling "usable vs trash", it is randomly rolling the ilvl, the affixes number, the affixes themselves (from a pool of many useless ones) and, in the end, the actual values.

It is quite clear, statistically, that the vast majority of items will be vendor trash.

This.

Say you have an item of Dex between 1-10
Any values below 8 is deemed trash.
so you have 70% chance of getting trash.
Now multiply the chance by the number of attributes and affixs, and the change to have a number of attributes getting 8 and above.
so that's your chance of getting the trash

Most items are bad because there are many more items than characters.

Let's look at a simple example. For simplicity, pretend that you never look at the auction house. You've finally hit level 60, reached Inferno, and found your first level 60+ weapon, which is better than whatever you had before and you equip it.

5 minutes later, you find another level 60+ weapon. What are the odds that this is an upgrade for you? 50/50. Half the time it's better, half the time it will be worse, not accounting for anything else.

5 minutes after that, another weapon! But is this one an upgrade? Well, it would have to be better than either of the previous weapons you have found, which it has only a 1 in 3 chance of being.

The next weapon? 1 in 4.

It keeps going like this, with the odds of any drop being an upgrade for you being proportional to 1/T, where T is the time you've been playing. This is totally without regard to the randomization algorithm or how items are generated or anything of the sorts - once you're no longer improving your drop tables, your upgrade frequency is going to drop as you keep playing, until you essentially stop finding upgrades at all.

Now take the auction house into account. What this does is redistribute items between players. Assume, for example, that there are a million players with a level 60 on the auction house. That would mean that, roughly, only the million best weapons ever found are worth much at all, with the price of weapons right on the edge of that million being exceptionally low since they're so similar to items that just miss the cut. Then add in that this is going to suffer from the same 1/T relationship as the solo play example - we're 10 weeks into the game so far, and after another 10 weeks the number of weapons found will have doubled, which means half the weapons that currently exist will have dropped from 'marginal' to 'crap'.

What does that mean for your own drops? It means the usefulness of your own drops, from the perspective of the auction house, does not depend on how long you have played, but how long *everyone* has played. The more everyone plays, access to high quality gear increases, while an increasing number of items on the edge slip from useful to trash.

This has nothing whatsoever to do with the randomization method, or even how item mods are generated; it's just the nature of picking out the best items from a random pool with a constant generator algorithm. There's nothing whatsoever Blizzard can do about this; it's baked into the game.

Was gonna write a long post explaining the chance for a perfectly rolled item, and why we haven't seen any screen shots of any yet, but the poster above me (Ensign ,+1 btw) just make it a lot easier to understand for laymen ^^

1st lvl 63 weapon drop you find will be the best lvl 63 weapon you have found.
2nd lvl 63 weapon drop you find will have a 50% chance to be better or worse (ignoring the infinitesimal chance of it being the same).
3nd lvl 63 weapon drop you find will have a 33.33% chance to be better then the two previous ones.
...
100th lvl 63 weapon drop you find will have a 1% chance to be better then the other 99 previous ones.

This logic goes for all gear slots. The more stuff you find (or buy) the less and less chance there is that anything that drops will be better then what you have. And in SC where items "never" disappear, it will just be harder and harder to find upgrades, so the "average" quality of stuff on AH will go up and up, until eventually every item on the AH is perfectly rolled (in like a few million years ^^).

Fun theoretical example
You want this chest:

Archon armor
+max str
+max vit
+max str/vit
+max all resist
+max armor
+3 sockets

So... How many packs would you have you kill to get this item? Well we don't have all the statistics yet from loot distribution and/or stat allocation but at least:

Chance of getting a rare item (depends on mf ofc, but lets assume you always have 5 NV buff for that extra one) < 200%="" (edit:="" ops="" was="" gonna="" type="" 300,="" ohhh="" well="" still="" generous="" in="" the="" other="">
Chance of it being a chest piece <>
Chance of it being lvl 63 <>
Chance of it having 6 affixes < 20%="" (edit:="" i="" was="" a="" bit="" harsh="" here,="" but="" to="" lazy="" to="" change="" the="" math="">
Chance of it having str affix <>
Chance of it having vit affix <>
Chance of it having str/vit affix <>
Chance of it having all resist affix <>
Chance of it having armor affix <>
Chance of it having sockets affix <>
Chance of having maxed out str stat for the affix <>
Chance of having maxed out vit stat for the affix <>
Chance of having maxed out str/vit stat for the affix <>
Chance of having maxed out all resist stat for the affix <>
Chance of having maxed out armor stat for the affix <>
Chance of having maxed out socket stat for the affix <>

So you will have to kill at least (the percentages above are VERY generous) to have an average change (fifty-fifty) of finding the item: 1 * (1/2) * (1/0.1) * (1/0.5) * (1/0.2) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.05) * (1/0.05) * (1/0.05) * (1/0.1) * (1/0.1) * (1/0.5) = 109,739,368,999 packs.

If the RNG on diablo 3 is truly random then it should be impossible to get two of the same items with exact stats,name,gold sell price and durability form 2 different mobs, well folks a friend of mine did.

no not really, even in truly random scenarios you can get 2 equal outcomes, even in a row at times

That only happens if the loot table is from 1 to 10, the loot in this game is billions, it may be possible to get the same item some time down the road but not that soon. Also it seems they are just reusing the random hashes.

I believe there was(is?) a bug which causes the same item to drop twice in a row from different, subsequent packs.

So when was that said bug fixed? Because the item was looted on the day I also posted it.

So you will have to kill at least (the percentages above are VERY generous) to have an average change (fifty-fifty) of finding the item: 1 * (1/2) * (1/0.1) * (1/0.5) * (1/0.2) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.05) * (1/0.05) * (1/0.05) * (1/0.1) * (1/0.1) * (1/0.5) = 109,739,368,999 packs.

That number is so high that people will most likely not realize the true meaning of it.
I took the freedom to manipulate it a bit so people may realize the size of farming required.

109,739,368,999 packs required
Let's assume the average farmer with decent gear can manage to kill (on average) 35 packs/hr
--> 3,135,410,543 hours required.
------> This equals 130,642,106 days (farming 24/7).
---------> which also equals 357,924 years (farming 24/7) for just 1 person.
(this means that if the average lifetime of a person is ~75 years of which you are capable of farming for ~70, you would need at least 5,000 lifetimes).

Put another way:
It would take ~358,000 people approximately a year of farming 24/7 to farm a perfect item. That population is equal to ½ of the population of Detroit in Michigan!

If the RNG on diablo 3 is truly random then it should be impossible to get two of the same items with exact stats,name,gold sell price and durability form 2 different mobs, well folks a friend of mine did.

Actually this statement is completely false
you can obtain an exact copy of an item from another mob randomly, it is just a very small chance.

I think Kripparian's video, which someone linked on the first page, explained it really well. In a world where 50% of the drops were what we consider "good" right now, your standards for what counts as "good" would go way up, to the point where you wouldn't consider most of those drops "good" any more.

If a great item somehow dropped every 10 minutes, you wouldn't be able to sell a single one of those items on the AH. No one would bother to buy them, because all they have to do is play for 10 minutes to get a great item on their own.

So you will have to kill at least (the percentages above are VERY generous) to have an average change (fifty-fifty) of finding the item: 1 * (1/2) * (1/0.1) * (1/0.5) * (1/0.2) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.05) * (1/0.05) * (1/0.05) * (1/0.1) * (1/0.1) * (1/0.5) = 109,739,368,999 packs.

That number is so high that people will most likely not realize the true meaning of it.
I took the freedom to manipulate it a bit so people may realize the size of farming required.

109,739,368,999 packs required
Let's assume the average farmer with decent gear can manage to kill (on average) 35 packs/hr
--> 3,135,410,543 hours required.
------> This equals 130,642,106 days (farming 24/7).
---------> which also equals 357,924 years (farming 24/7) for just 1 person.
(this means that if the average lifetime of a person is ~75 years of which you are capable of farming for ~70, you would need at least 5,000 lifetimes).

Put another way:
It would take ~358,000 people approximately a year of farming 24/7 to farm a perfect item. That population is equal to ½ of the population of Detroit in Michigan!

To be fair, that is the chance to find a PARTICULAR perfect item. Since there are several perfect items (i.e. other item slots, in some slots different affixes might be desirable depending on your build / character), the actual chance would be a bit higher. It would still be astronomically small, but slightly bigger.

To be fair, that is the chance to find a PARTICULAR perfect item. Since there are several perfect items (i.e. other item slots, in some slots different affixes might be desirable depending on your build / character), the actual chance would be a bit higher. It would still be astronomically small, but slightly bigger.

Yes - but the calculation was made on chances that were VERY generous and hence (even though it depends on build and/or character), the true chance is actually much much smaller

actually if the item-drop were to be perfectly random they should put every single possible item combination of every item whit all the affixes in one big bag of billions and billions of item and then just give you one item randomly from the bag every time you kill a mob.

Haven't read the whole thread, so i'm not sure if anyone else have pointed out allready.
Weather an items is usable or not, is often a matter abut the the right stat and the right combination of these.
There are a bunch of (don't know the exact number) different affix, and the chance to get the right affix-combination, with hight enough stats for inferno, is very low - probably lower than the 5% you mentioned in the OP. You also have to take into account, weather an item is good or bad differ from class to class. Loot that you find trash-material can for some people be quite good....

Rollback Post to RevisionRollBack

To post a comment, please login or register a new account.

That's like saying it should be impossible to roll 6 in a dice twice... Although that does sound like a bug.

That only happens if the loot table is from 1 to 10, the loot in this game is billions, it may be possible to get the same item some time down the road but not that soon. Also it seems they are just reusing the random hashes.

This makes zero sense, since you are basing something mathematical (affixes and their ranges) with something conjectural (what YOU would consider useful).

U think usable vs trash should be even?

Look how many friggin stats there r.

So first, u need to have the stats u want 2-6 stats on an item depending what you are looking for. Then it has to have high/good values. So rng stays rng, no conspiracy theories.

Also what you consider to be garbage trash might be considered decent for someone just starting in act 1 inferno. You say you farmed 200 hours act 2, then its no friggin wonder u dont find upgrades cause u have most likely found them earlier.

I believe there was(is?) a bug which causes the same item to drop twice in a row from different, subsequent packs.

Sorry but no, the game isn't rolling "usable vs trash", it is randomly rolling the ilvl, the affixes number, the affixes themselves (from a pool of many useless ones) and, in the end, the actual values.

It is quite clear, statistically, that the vast majority of items will be vendor trash.

Say you have an item of Dex between 1-10

Any values below 8 is deemed trash.

so you have 70% chance of getting trash.

Now multiply the chance by the number of attributes and affixs, and the change to have a number of attributes getting 8 and above.

so that's your chance of getting the trash

Let's look at a simple example. For simplicity, pretend that you never look at the auction house. You've finally hit level 60, reached Inferno, and found your first level 60+ weapon, which is better than whatever you had before and you equip it.

5 minutes later, you find another level 60+ weapon. What are the odds that this is an upgrade for you? 50/50. Half the time it's better, half the time it will be worse, not accounting for anything else.

5 minutes after that, another weapon! But is this one an upgrade? Well, it would have to be better than either of the previous weapons you have found, which it has only a 1 in 3 chance of being.

The next weapon? 1 in 4.

It keeps going like this, with the odds of any drop being an upgrade for you being proportional to 1/T, where T is the time you've been playing. This is totally without regard to the randomization algorithm or how items are generated or anything of the sorts - once you're no longer improving your drop tables, your upgrade frequency is going to drop as you keep playing, until you essentially stop finding upgrades at all.

Now take the auction house into account. What this does is redistribute items between players. Assume, for example, that there are a million players with a level 60 on the auction house. That would mean that, roughly, only the million best weapons ever found are worth much at all, with the price of weapons right on the edge of that million being exceptionally low since they're so similar to items that just miss the cut. Then add in that this is going to suffer from the same 1/T relationship as the solo play example - we're 10 weeks into the game so far, and after another 10 weeks the number of weapons found will have doubled, which means half the weapons that currently exist will have dropped from 'marginal' to 'crap'.

What does that mean for your own drops? It means the usefulness of your own drops, from the perspective of the auction house, does not depend on how long you have played, but how long *everyone* has played. The more everyone plays, access to high quality gear increases, while an increasing number of items on the edge slip from useful to trash.

This has nothing whatsoever to do with the randomization method, or even how item mods are generated; it's just the nature of picking out the best items from a random pool with a constant generator algorithm. There's nothing whatsoever Blizzard can do about this; it's baked into the game.

Sprechen sie statistics?

1st lvl 63 weapon drop you find will be the best lvl 63 weapon you have found.

2nd lvl 63 weapon drop you find will have a 50% chance to be better or worse (ignoring the infinitesimal chance of it being the same).

3nd lvl 63 weapon drop you find will have a 33.33% chance to be better then the two previous ones.

...

100th lvl 63 weapon drop you find will have a 1% chance to be better then the other 99 previous ones.

This logic goes for all gear slots. The more stuff you find (or buy) the less and less chance there is that anything that drops will be better then what you have. And in SC where items "never" disappear, it will just be harder and harder to find upgrades, so the "average" quality of stuff on AH will go up and up, until eventually every item on the AH is perfectly rolled (in like a few million years ^^).

Fun theoretical exampleYou want this chest:

Archon armor

+max str

+max vit

+max str/vit

+max all resist

+max armor

+3 sockets

So... How many

packswould you have you kill to get this item? Well we don't have all the statistics yet from loot distribution and/or stat allocation but at least:Chance of getting a rare item (depends on mf ofc, but lets assume you always have 5 NV buff for that extra one) < 200%="" (edit:="" ops="" was="" gonna="" type="" 300,="" ohhh="" well="" still="" generous="" in="" the="" other="">

Chance of it being a chest piece <>

Chance of it being lvl 63 <>

Chance of it having 6 affixes < 20%="" (edit:="" i="" was="" a="" bit="" harsh="" here,="" but="" to="" lazy="" to="" change="" the="" math="">

Chance of it having str affix <>

Chance of it having vit affix <>

Chance of it having str/vit affix <>

Chance of it having all resist affix <>

Chance of it having armor affix <>

Chance of it having sockets affix <>

Chance of having maxed out str stat for the affix <>

Chance of having maxed out vit stat for the affix <>

Chance of having maxed out str/vit stat for the affix <>

Chance of having maxed out all resist stat for the affix <>

Chance of having maxed out armor stat for the affix <>

Chance of having maxed out socket stat for the affix <>

So you will have to kill

at least(the percentages above areVERYgenerous) to have anaverage change(fifty-fifty) of finding the item: 1 * (1/2) * (1/0.1) * (1/0.5) * (1/0.2) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.3) * (1/0.05) * (1/0.05) * (1/0.05) * (1/0.1) * (1/0.1) * (1/0.5) = 109,739,368,999 packs.Good hunting!

So when was that said bug fixed? Because the item was looted on the day I also posted it.

That number is so high that people will most likely not realize the true meaning of it.

I took the freedom to manipulate it a bit so people may realize the size of farming required.

109,739,368,999 packs required

Let's assume the average farmer with decent gear can manage to kill (on average) 35 packs/hr

--> 3,135,410,543 hours required.

------> This equals 130,642,106 days (farming 24/7).

---------> which also equals 357,924 years (farming 24/7) for just 1 person.

(this means that if the average lifetime of a person is ~75 years of which you are capable of farming for ~70, you would need at least 5,000 lifetimes).

Put another way:

It would take ~358,000 people approximately a year of farming 24/7 to farm a perfect item. That population is equal to ½ of the population of Detroit in Michigan!

you can obtain an exact copy of an item from another mob randomly, it is just a very small chance.

If a great item somehow dropped every 10 minutes, you wouldn't be able to sell a single one of those items on the AH. No one would bother to buy them, because all they have to do is play for 10 minutes to get a great item on their own.

To be fair, that is the chance to find a PARTICULAR perfect item. Since there are several perfect items (i.e. other item slots, in some slots different affixes might be desirable depending on your build / character), the actual chance would be a bit higher. It would still be astronomically small, but slightly bigger.

Yes - but the calculation was made on chances that were VERY generous and hence (even though it depends on build and/or character), the true chance is actually much much smaller

then off course MF wouldn't exist.

Weather an items is usable or not, is often a matter abut the the right stat and the right combination of these.

There are a bunch of (don't know the exact number) different affix, and the chance to get the right affix-combination, with hight enough stats for inferno, is very low - probably lower than the 5% you mentioned in the OP. You also have to take into account, weather an item is good or bad differ from class to class. Loot that you find trash-material can for some people be quite good....