Diablo 3 Legendary Drop Rates Revealed – Updated

We’ve got the Diablo 3 Legendary drop rates for all set and legendary items in the game. Thanks to site reader Childe for helping us assemble them, via a Korean site where they first leaked. These are the *weighted* item drop rates, much like the old Diablo 2 Treasure Classes. These figures do not tell you your chances of finding a legendary item. What they tell you is IF the game rolls a legendary or set item for a given slot, WHICH of the possible items will appear.

We’ve got all the data in three forum threads, and you’ll want to spend some time crawling over them, since it’s a huge trove of useful data. Note that 15% of item drops are un-Smart Drops, which is why your characters will sometimes find items “only” for other classes. (Including legendaries.)

The different threads for difficulty level are due to the odds changing with DiabloWikiTorment-only gear. For instance, there aren’t any Torment-only Fists, or Spears, or Swords, etc, so the drop odds for the legendary items of those types are the same on all difficulty levels. Most types of armor though have several Torment-only items, though.

For example, there are five legendary helms that can drop for five classes (Barb is the exception), on all difficulty levels. They are the DiabloWikiBroken Crown, DiabloWikiLeoric’s Crown, DiabloWikiBlind Faith, DiabloWikiAndariel’s Visage, DiabloWikiMempo of Twilight, and the DiabloWikiDeathseer’s Cowl, and all have a 16.67% chance of dropping. (That’s 1/6, even odds for each.) The Barbarian is different, since the DiabloWikiSkull of Resonance can only drop for him. Thus his odds are 15.38% for the other 6 helms, and 7.69% for the Helm of Resonance. (That’s 2/13 for the first 6 and 1/15 for the Skull of Resonance.)

Things get more complicated when you take your Barbarian into Torment+, since then the basic 6 helms can still drop (11.76% each), or the Helm of Resonance (5.88%), but you might also get one of the Torment-only helms: DiabloWikiImmortal King’s (5.88%), Raekor’s Will (5.88%), DiabloWikiCrown of the Invoker (5.88%), or DiabloWikiEyes of the Earth (5.88%). The odds for the normal helms drop since the total odds have to equal 100%, and there are more Torment-only helms being added into the chances. The other classes see their odds for the different helms change in Torment as well, but somewhat differently than the Barb’s since they all have a different number of possible set/leg helms to choose from.)

Again, those are the percent chance that a given set/leg helm will drop IF THE GAME ROLLS A LEGENDARY OR SET HELM for your drop. (Just to emphasize that point again, since it was always misunderstood when we were talking about Diablo 2 unique drop odds back in the ago.)


Now I Understand!

Looking over the numbers, I had a number of light bulb moments. Now I see why Blackthorn’s Set items are so damn common! They’re weighted the same as legendary items, while most Set Items are weighted half as common. DiabloWikiBlackthorn’s Jousting Mail (pants) are 30.3% chance to be the set/leg pants you get if you are playing a Barb/Crus/Monk/Wiz. Or 23.62% if you’re a DH or WD, since those classes have DiabloWikiSwamp Land Waders dropping, which changes the odds for all pants.

This is why we can't have nice things.

This is why we can’t have nice things.

And now I know why I’ve found every other bow in the game but not a DiabloWikiKridershot. If the game decides to drop a legendary bow (there are no set bows) the DiabloWikiUskang, DiabloWikiThe Raven’s Wing, DiabloWikiEtrayu, DiabloWikiCluckeye, or DiabloWikiWindforce all have a 19.61% chance of rolling, while the much-coveted Kridershot is down at 1.96%. (The bow weighting odds odds are the same on Torment+, since there are no Torment-only bows.)Hand Crossbows do the same thing to the DiabloWikiCalamity. On Normal-Master you’re looking at 17.86% odds for the DiabloWikiIzzuccob, DiabloWikiBalefire Caster, DiabloWikiDanetta’s Spite, or DiabloWikiDanetta’s Revenge, with the DiabloWikiK’mar Tenclip at 8.94%, and the Calamity at 1.79%. Yes, that’s 1/10th as likely to drop. It’s even worse on Torment+, when DiabloWikiNatalya’s Slayer starts to drop. All the others shift to 16.4%, Nat’s and K’mar Tenclip get a 8.2% chance, and Calamity decreases to 1.64%. Good luck with that!

Most other weapons and several types of armor have similar weighting issues, where the one or two items you really want in that slot are MUCH less likely to drop. That’s not a surprise of course, but seeing the actual figures can put the “I’ll never find X” into a harsher reality.



These figures may not be final, as the info is still being correlated from the Korean info leak and some of the items listed (all with 0% drop odds) aren’t enabled in the game. There may be additional factors to consider as well. For instance, the whole thing with the game assigning 5-7 items of each slot to drop in each game. That’s a real thing, confirmed by a lot more testing results I’ve seen since we posted that article, but it’s not known how those items figure in with these drop odds.

Also, I’m skeptical about the figures for Set Items. They are listed as weighted at 1/2 the frequency of legendary items (in most cases) but my experience and that of other players I’ve talked to is that sets are *much* less common that that. Maybe 1/3 to 1/5th (to 1/10th?) as frequently found as legendaries. That’s partially explained by the fact that many sets are Torment-only, and the fact that there are zero set items in many weapon slots, so all those times you find a spear or a bow there’s a 0% chance of it being a set item. That said, it’s possible that other code in the game sets limits that impact these drop odds.

This appears to be the master weighting list, but there may be other lists in the code that further cut the odds of set items, or raise or lower specific items in frequency for all classes, for all monsters, for some specific monsters, etc. We know that sometimes classes with a 0% listed rate for certain items will find them anyway (violating their Smart Drop). And we know that legendary items get a drop bonus in Nephalem Rifts, and Blizzard said they’re about to buff legendary drop rates from gambling. They could just as easily put in buffs for specific items or item types, from certain monsters, etc. For instance, bow-using monsters had special weighted odds to drop legendary bows in Diablo 2.

That said, it’s great info and I enjoyed poking through it. What do you guys notice that seems surprising, or that’s not balanced correctly, or that confirms your darkest suspicions?

Update: By request, here are the ultra-rare items, most of them set to drop at 1/10th the standard rate:



5 May List updated

As many of you have pointed out there were some omissions in the tables and they have once again been updated with the following additions on all three lists.


Update:  The DiabloNut Armory has now been populated with :


Tagged As: | Categories: *Featured, Datamining, Diablo 3, Legendary Items


You're not logged in but can still post comments. Register or login to remember your details.
  1. Not at all surprising. It’s very clear the enchanting has properties weighted in a similar fashion, with the most desirable properties having miniscule chances to show up. Make a set of level 70 rares and see how much you spend even getting one reduced level requirement affix to roll, let alone on all pieces or a useful number on all pieces.

  2. Whelp. This information made me feel really awesome, and really shitty, some of the legendaries I’ve found are among the rarest in the game, and yet, the ones I still want, are still the rarest.


    Confirms my suspicions though.

    • You know this is wrong right?

      This list is flatout WRONG.

      Some notable exceptions (from a confirmed, PERSONAL evidence.)

      I play ONLY a Barb, and the list says it’s 0% drop chance for some of the items.

      But as a Barb I’ve gotten:

      * The Eye of the Storm (monk hat)
      * Kassar’s Retribution (crux 1h flail)
      * Skycutter (1h sword)
      * Cape of Dark Knight (dh cloak)
      * Cloak of the Garwulf (dh cloak)
      * Natalya’s Embrace (dh cloak)
      * Blackhand Key (wiz wand)
      * Natalya’s Bloody Footprints (dh/monk boots)
      * Eberli Charo (crux shield)
      * Hellcrack (dh crossbow)
      * Danetta’s Spite (dh bow)
      * Natalya’s Slayer (dh bow)
      * Jekangbord (crux shield)

      Note: These are listed as 0% for Barbarian, but I’ve gotten at least 1 of those listed items. Drop list is not accurate.

      • Did you know that smart loot is not 100% drop chance? So before you cry again saying it’s 100% incorrect, note that Blizzard designed the game to not always drop smart loot so you will in fact see off class items drop. That’s why you see all those items you found as a barb. Please think before you shout blasphemy next time.

        • I added a note after the post to clarify this. Smart drops are designed to function 85% of the time, thus 15% of your finds *should* be random. It’s why chars can find items out of their class, and why the 100% and 0% in the fields in these drop weighting tables are not absolute.

      • Blizzard dropped those items so you would come out of your Barb shell and play other chars.

  3. I’ve had wd set logs drop on monk and also dh boots when it says on this list they are 0% sorry i’m calling bs lol

    • Agreed. I’ve found Natalya’s Bloody Footprints while being on a Barbarian.. and I ONLY play Barbarian. This list isn’t 100% accurate, at all.

      • “Note that 15% of item drops are un-Smart Drops, which is why your characters will sometimes find items “only” for other classes. (Including legendaries.)”

        Please learn how to read before you make yourself look like a fool.

        • To be fair, when we first posted this we didn’t have that note, so people who didn’t know about the (little-documented) 85/15% smart/unsmart ratio were making a valid objection. Hopefully everyone’s read the disclaimer now.

          Also, look at the 3rd link for the smart loot info; it specifies every legendary and which classes it will smart drop for, i.e. the cases in which the listed percentages apply. If your char finds one of the legendaries that are listed with 0% chance for them it’s not a bug.

          If you’re playing a Barb you’ll find around 15% of items that are nto smart-dropped. They’ll have INT or DEX for the mainstat, or will be items like Monk fists or DH xbows. They might even be set or legendary quality. But we don’t yet know anything about how the game weights the probability of those finds or how it determines their item quality.

    • This list is for when you get a smart loot drop. We know that smart loot is not 100%, as designed by Blizzard. So you can see WD items as a monk. So, now you understand why it says 0%. Not BS as you say it is

  4. So legendaries are simply sorted by three tiers of rarity:

    – Tier 1 = Full probability to drop
    – Tier 2 = Half the full probability to drop
    – Tier 3 = 1/10th the full probability to drop

    Else all items in the same category are simply added up evenly to 100%?

    Hmmm… I sure hope Blizz has other systems in play, as this doesn’t really sound very creativ and substantial. On the contrary: It gives off quite the vibe of a tight bottleneck on balancing chance to drop vs. itempower in conjunction with class x/skill y.

    • It further provides suspicion to all secondary gamesystem designs possibly only using euclidic coordinate systems. If so, the lack of depth I perceive in them ain’t really surprising, as the magic providing the possible depth I hunger for would only come into play through interjunctions of euclidic and/or humugenous coordinate systems with polar coordinate systems, centered in a single global or local one.

  5. The table is not 100% correct. I have found Vigilance polearm on my WD even though the table says drop chance is 0%.

  6. What Toxify says is true. The only character I have played so so far in torment has been my Witch Doctor. I have only found one of the torment only set items, that item being Tal Rasha’s Unwavering Glare. According to the chart I have a 0% chance of finding it. This info is plainly wrong at least to some degree.

    • 15% of the drops are not SMART thus the char specific table doesn’t apply (probably a weighted master table that is neutral in regards to char type)

  7. Childe here– The percentages are just approximations that it should be noted specifically cannot account for the exact math behind Smart Loot (or the occasional lack thereof leading to non-Smart Loot drops) because that math isn’t known. I think in some of the excitement and the relaying of information between multiple people that may have been lost, but that was communicated.

    But moreover, the data also included the raw numbers that those approximate percentages were based on, separate from the classes. Hopefully Flux can share some of that information as he and the other staff have a chance to sort through it and get it onto the site/Wiki to alleviate the confusion. Those numbers are the weighted drop rates Flux alluded to at the beginning of the post, which seem to work just like in Diablo 2.

  8. This list is probably correct when smart drop is enabled. However smart drop is not always factored in so it’s quite possible to find something listed as 0% for any given class.

    Personally, if this list is correct, I have been really lucky.

  9. “For instance, the whole thing with the game assigning 5-7 items of each slot to drop in each game.”
    It’s looks like it’s not true at all. Besides that idea is just plain stupid.

    • So provide a counter-argument. Do you have data to counter the data that the claimers of the 5-7 argument have made?


      Shut the fuck up and go back to playing, Mortalo.

      • Common sense 😛

      • Actually, we don’t need to provide a counter-argument. If someone made a statement that’s based on an observation it’s their job to provide irrefutable evidence for that claim. If you make such a claim the burden of proof is on *you* as the claimant. Your claim is not mine to disprove.

  10. None of the links are working, I guess too much load

  11. to all n00b fanboys who say this is fake because u got a barb item on your wizard , hey n00bs, this loot table is for when the smart loot is not active, aka 15% of the time something will drop.
    Educate urself before posting, n00bs, this shows just how idi0ts fanboys are

  12. All I have to say is that I hope this makes Blizzard move now that the cat is out of the bag. I want to find my damn Kridershot before 2029 (yeah and we all know it has a very high chance of rolling shitty affixes and ranges)

    I wouldn’t mind the ultra rare items IF the stats on them would be guaranteed very good.

  13. jeez, looking at the comments i think you should rewrite this article so that every other sentence points out that you are talking about smartdrops. maybe put those sentences in caps too…

    anyway, interesting read!

    oh and \legendary bows in Diablo 2\? they were called unique! keep your shit together flux 😀

    • No kidding, here, reddit, battle.net, pretty much anywhere: “Well that site is just building a house of lies ’cause I got this item on a class that they say had a 0.0%!”

      It’s as if none of them ever look at the yellows/blues/whites they get and see all the cross-class/stat non-smart loot drops they get every single game.

  14. Ok, with these odds – or allow us to trade or allow us to gamble for SPECIFIC weapons. Or both.

  15. For the set item … I think the problem is “the game roll what it will roll at the very beginning”. The result is if we dont re create game drop chance = 0 forever.

    Even it rolled it may not drop, but if doesnt roll at the very begining no matter how 7/24 you play an no chance to get any

  16. Skull of Resonance dropped for me from a normal cache while playing with my Wiz, and it dropped with strength and an avalanche proc, so….

  17. whoa, so that ‘loot table per game’ conspiracy theory was true after all?

    • No. This does not mean that. And I’m not convinced that Blizz did something like that.

      • >There may be additional factors to consider as well. For instance, the whole thing >with the game assigning 5-7 items of each slot to drop in each game. That’s a real >thing, confirmed by a lot more testing results

        • I don’t consider it true until we get an official word from Blizzard, or at the very least until some convincing test results are shown.

          • Well I got 3 Tal’s belt drops in a single session so that proves it!

            Oh wait, no, we need more information than that. Darn. I thought my anecdotal evidence would be sufficient for second there 😛

            To be honest though, that triple Tal’s belt drop really did make me think there may be some truth to the allegations, but it could also just have been RNG.

  18. I’m assuming this list is a “work-in-progeress” because it’s not complete. It’s missing at least one thing – Death Watch Mantle in pauldrons (which I’m hunting right now) and the percentages are not adjusted to incorporate it.
    I don’t know what is the source of this data (I’m guessing data mining) but if there is one such omission who’s to tell there aren’t more. This makes this list much less useful.
    One other thing. Providing the theory about drop assignment at game creation is correct, how does changing difficulty mid-game influences this assignment? I can imagine situation when a legendary of the desired slot drops on normal so one goes back to town, changes to torment 6 and starts gambling.

    • White names are PH names or:

      The white names are codenames for the “missing items”. They were left on code with the codename instead of the final name – I think, based on D2 MPQs and TXTs.

      Excluding Seven Sins/Corruption (crafted) and Pauldrons of Skeleton King (cache), there are 5 legendary pauldrons:

      Homing Pads – on the list
      Spaulders of Zakara – on the list

      Profane Pauldrons – on the list (the important one, by the way, with lower rate)

      “Evil Monitor” – Probably a place holding name for Vile Ward (I’m assuming it due to similarity)

      “Dying Shawl” – Probably a PH name for Death Watch Mantle (exclusion)

      Since the Evil Monitor and Dying Shawl rates are the same, as in “not set, not Profane Pauldrons” (20% torment drop; 22,22% kadala gamble), what PH name is which really isn’t that deal breaker imo.

      • Yeah, I haven’t thought about it. I just assumed these were some new legendaries I haven’t seen yet but it might be just bad Korean-English translation.

        • Could be that, too, but sometimes MPQs/txts (or whatever the filesource type was) have mispellings and PH names on the final version and they left it unchanged on the files, because if you change X, you end up having to change Y to avoid bugs.

          Maybe the dataminer just copy + paste the names/rates.

          Looks like it’s updated now, so no more missing names, I think. 🙂

  19. I found a Fist of Az-Tarrask with my WD 🙂 My conspiracy theory now is “game is trolling me”.

    But yep, no WD sets (other than zuni helm), usable voodoo masks, starmetal kukri, etc. Got. The. Rarest. Usable. Monk. Fist.

    Jokes aside, this table looks like Pindle RNG in game, more or less, imo – you don’t get what you want, but once in a while you get a uber rare thing you don’t want – like a TC 3 from pindle.

    I matches my results and the results I’ve been reading around the interwebs.

    • Good comparison. In D2 there was the option to do meph or andy runs, thus changing TCs and essentially controlling your droprates to hunt for specific items. (At least to a small degree.) That freedom is totally lost to the player according to this data.

  20. Well, I got the Hexing Pants a week or two ago, so I guess that covers my ultra-rare drop for the month. I was underwhelmed. On the other hand, I’ve actually been surprised how many T-only set items drop – and the numbers are more generous than I would have expected. I’ve found about 1 per week since I started running T1. Do we know what the odds are of the game dropping each item type, i.e. a sword vs a helm? I’ve been hankering for a legendary potion and haven’t found one, and I wonder how rare they are vs other drops in torment.

    • We don’t know, I suppose. But based on the above and playing experience I would say that it’s as evenly a distribution as seen in the above, though propably without any tiering in rarity.

  21. The one thing I notice is that Magefist does not appear to be anywhere, nor does there appear to be a missing percentage for it.

    • The set ring Wailing Host is also nowhere on the list, and it is the one I find most of all rings. I’ve found at least 6 of them so far, more than any other ring.

  22. Pox Faulds are listed at between 3.77% and 5.26% to drop depending on class Certainly hasn’t felt that rare to me. Of the 6 set/legendary pants I’ve had drop so far, 4 were Pox Faulds, 1 was Hammer Jammers, and 1 was Blackthorne’s. All had terrible rolls of course.

    Is this just random being random or have others had a similar experience?

    • At this point it seems safe to say, that it’s definitely the accumulation process(/the clustering) not being rightfully balanceable due to the design being too simple to allow for enough methods to filter the raw randomisation it provides. So a clear “Yes!” on both options as given by your question.

      Iirc: Pox Faulds – none; Hammer Jammers – 3; Blackthorne’s – 1. (Not 100% sure, as I’m only keeping the best rolled ones, if having the same mainstat…)

  23. I’m inclined to think these drop rates were different or have been modified by Blizzard between the big 2.0 item patch and the release of ROS. Or I guess I was just the luckiest bastard ever when I managed to find 3 different Zuni set items and another 2 duplicates in the first few days of 2.0 patch (prior to ROS release).

    Since ROS launched, I’ve found 1 zuni set item in the many hours of farming.

    Nothing to back it up, purely anecdotal, just seems odd I found so many of the level 60 zuni set items prior to ROS, then only 1 set item in at least 5 times the hours of farming post ROS.

  24. \For instance, the whole thing with the game assigning 5-7 items of each slot to drop in each game. That’s a real thing, confirmed by a lot more testing results I’ve seen since we posted that article, but it’s not known how those items figure in with these drop odds.\

    Nonsense. The loot tables have been comprehensively debunked. As this data, BTW, CLEARLY demonstrates.

    • This data is only the inner axxis of the complete drop system. For itself it doesn’t exclude the possibility of such a scenario.

      But what the data tells us is that there can only ever be put two additional systems in conjunction with it. Adding just one layer more would already result in the systems blocking out each others usefulness. (Smartdrops obviously being one of the systems.)

      And it leads to the conclusion that none of the two can be put above the displayed droptables in the grand design, as this would additionally extrapolate the clusterization. But as two systems accounting for finetuning does not pan out well using only euclidic coordinate systems, it has probably been done anyways…

  25. This is bull shit and just retarded after 600H in game play as a barb with no fulminator / tunderF / Soh /uniti and i can list mb almost all the good items

    i got nothing and like i did say i invested 600H in too the game. there are ppl with 50 H and got full set Green / Wepons and hes not even playing the game .

    So i wanna say thanx blizzard and …. , i think is a good day to quit this good game this is too mutch for me .

  26. It’s funny that it looks like DHs have the best chance to find good items – they have a base 0% chance to have 1H, 2H, Shields, etc to drop (discounting the 15% un-smart loot drop chance, which I may add is a nice big middle finger to players)

    As stated above, you’re better off playing on Master when looking for certain items since switching to T1 adds even more items to the list of loot drops.

    Also, maybe it’s just me but I did find a lot more rare legendaries at the start of the release, after which it’s switched over to Blackthorne’s and belts primarily….

  27. Several items missing from list, otherwise seems to confirm my own play experience. Looks like now is a good time to quit and uninstall. Hopefully the next diablo game is much more realistic in regards to drop rates.

  28. Just to add something,

    I got a Tasker and Theo as a drop, and immediately reported to clan mates. Personally I spent over 700 shards in the instance, and others spent plenty as well… Nothing no gloves from Kadala… at all.

    This is very, very far out on the curve so I am guessing something else is afoot. Perhaps the drop odds somehow affect odds with Kadala, as well.

    On a side note, some of the rarer drops give over the top power when paired together. This probably affected Blizz’s decisions.

  29. You FORGOT to add Rechel’s Ring of Larceny in the ULTRA RARE rings @ less than 0.50% AND has a range of 4560%… NUTS! Hands-down the best speed farming item in the game though most are unaware it even exists!

    • hey, I found that! 🙂

      Looked like an affix I could use sometime in the future so I’ve kept it in my stash luckily.

  30. Correct me if I’m wrong, the layers of RNG is like this?

    1.%chance to drop legendary
    2.rolls what kind of gear is it(bracer,rings)
    3.rolls if its smart loot or not
    4.if smart loot then roll with the chances listed in the table
    if not smart loot then roll all of them
    5.roll stats

    • Would put this way, …:

      I. 1. % chance, to drop as legendary, (rare6, rare5, …)
      II. 2.-4.
      III. 5.

      , … but yeah, that would essentially be the jizz of it, though not accounting for the probability of systems providing finetuning.

  31. I like to keep it simple my data is saying, You have 99,99% trash.

  32. So let’s combine this with that tealdawn’s theory. If I host a lvl11 normal game and gamble for boots with a lvl70 toon. the only legendary in the table would be lut socks and the crudest boots? If so then wow. Need people to try is this out.

    • PS:And each of them is 50%?

      • Looks like it.

        Even at the times I’ve practically only was critizing simplicity and lack of depth in D3s design, I never in my wildest dreams had the thought of it being so da** linear on top of it.

    • No, that is not what it means. Your adventuring/non-gambling drops will be limited to those boots IF they are on your loot table, but Kadala can still give you items above the game’s level (i.e. all the legendary boots that are on your loot table). Tealdawn’s theory says that you can determine if an item is on your loot table by speed clearing lower level T6 games and if it is, THEN you should gamble for it.

      Example #1
      You start a level 11 T6 game and your loot table is generated. On it are:

      1) Crudest Boots
      2) Irontoe Mudsputters
      3) Ice Climbers

      When you play, the only legendary boots that will drop for you will be Crudest Boots. However, when you gamble Kadala could potentially give you Crudest Boots, Irontoes, or Ice Climbers since she will provide items up to level 70. Since Lut Socks aren’t on your loot table, you will never obtain them from Kadala.

      Example #2
      Your loot table contains:

      1) Lut Socks
      2) Crudest Boots
      3) Irontoe Mudsputters

      In this case, when playing in your level 11 T6 game, the only legendary boots that will drop are Lut Socks and Crudest boots. By speed clearing rifts, many legendaries will drop which will let you determine if Lut Socks are on your loot table or not. After playing for 2+ hours, if a level 11 Lut Socks drops it means you can now gamble for the level 70 version from Kadala. However, Kadala can still give you level 70 Crudest Boots or Irontoes.

  33. I’d called D3 “just being an action-adventure, but no arpg” before. But after this info I have to retract my statement:

    D3 is just an hack&slay-accident mora than excelling in primary h&s gameplay, graphics and animations and allover polishing, but practically being botched in every other way by design, making it technically inferior to most of its competition on the arpg market. Sad as this definitively is, it raises the level of respect on the team, for bringing the game to a level of providing so much fun to play, that it still elevates the game above most of them.

  34. I can’t accept that Mempo and Andy’s have the same drop-rate as Deathseekers Cowl and Blind Faith. I have seen a DOZEN of each of the latter and ZERO of the former, outside Kadala.

  35. The title is quite misleading, I was hoping to learn the ACTUAL legendary droprates. I want to know how many items, on average, have to drop before I get a legendary. Is it one every 100, 500, 1000?

    • “I. 1. % chance, to drop as legendary, (rare6, rare5, …)”

      This is the only part that Iirc is still unknown standing in the way of calculation. (Well … besides the possibility of existing rarity tiers in other layers of the design behind drops.)

    • That depends greatly on your MF and the difficulty you are playing on. How are you playing; killing every last monster? Clicking all the dead bodies? Destroying all the destructibles? Also where you are playing; Rifts get a big bonus drop rate to leg items over other parts of the game.

      And sadly, we’re nowhere near being able to calculate that yet since most of the info is hidden on the B.net end of the client-server interaction. We knew exactly how items were generated in D1 and D2, but D3 remains mostly unknown.

      You could track data yourself and extrapolate, of course. Do 10 rifts on the same difficulty, count all the rares and legs that drop, and figure out a ratio. That’s way too small a data set for robust conclusions, but if you average 40 Rares per Rift and 1 leg every other rift, you’ve got something to measure and estimate from, at least.

  36. i got a The Star of Azkaranth on hardcore, apparently the rarest item in the game. am i just lucky or is it not the rarest item?

  37. I really like this table,but it is not correct or the game is bugged. I had Bul kathos solemn vow dropped for my wizard. At least,it had strength on it…

  38. As a barb there is a 23,36 chance for Blackthorn’s pants and 11,63 & for Weight of the Earth, according to this list.
    I’ve had Blackthorn’s pants 15 times. yes, 15. And 0 EQ pants. RNG!
    I’ve been farming like a manic for those pants. I have to admit it’s frustrating.
    Is torment only items going to be included in gambling in next patch, or is it only bounty caches that will be updated to drop torment legs too?

  39. I’m a barb and I have gotten a legendary daibo once and 2 spirit stone when in a rift, so it is not 0% lol

  40. Can anyone explain how these numbers were obtained in more detail than ‘datamining’?

    Was the .exe un-compiled(IDK proper term)?

  41. Sadly, rarer doesn’t always equate to better. I’ve gotten some of the rarest items on that list and they were junk that I never used while I am still searching for some of the less rare items. (actually, not searching until they increase legendary drop rate globally, I am done with a game and a game developer that goes out of its way to punish the players for no apparent reason.)

  42. When this list will get 100% accurate it could be a fairly useful resource, but without knowing the exact algorithms that use these numbers we are still mostly in the dark.

    Not that that’s necessarily a bad thing in this case. I’m still not sure if it’s better for us to know exactly how to get what we want. That could change and narrow our ingame behavior considerably. What ultimately killed D2 for me was the endless boss runs after I learned at which boss do I have the best chance of finding the item I want.

  43. About what I expected.

    Some of the rates are a bit wonky. Lacuni can’t really roll that well anymore, so there’s no need it would be so rare. Just as well off with Steady Strikers.

    There are a lot of legendary items that could really use updating so they play better with the 4/2 split. :/

  44. I think we have some evidence of functional illiteracy.

  45. So this means, if you can get a legendary to drop at all, and it’s a helm, it has an equal chance of 16.67% to be an Andariel’s Visage?

    Does this apply to gambling from Kadala? My personal sample size may be small but I have spent almost 8K shards on helms with no Andy’s at all. 3 Mempo’s, 7 Blind Faith, 5 Deathseer’s, and 5 each of Leoric’s/Broken at least…just bad luck then? Because out of all my drops so far, I should’ve rolled at least 3 or 4 Andy’s by now…again I know the sample size is small, but damn.

  46. QUOTE

    So this means, if you can get a legendary to drop at all, and it's a helm, it has an equal chance of 16.67% to be an Andariel's Visage? 
    Does this apply to gambling from Kadala? My personal sample size may be small but I have spent almost 8K shards on helms with no Andy's at all. 3 Mempo's, 7 Blind Faith, 5 Deathseer's, and 5 each of Leoric's/Broken at least...just bad luck then? Because out of all my drops so far, I should've rolled at least 3 or 4 Andy's by now...again I know the sample size is small, but damn.

    It’s been about 15 years since I took stats but I believe “confidence intervals” will answer your question.

  47. there are some issues with this paragraph:

    “For example, there are five legendary helms that can drop for five classes (Barb is the exception), on all difficulty levels. They are the DiabloWikiBroken Crown, DiabloWikiLeoric’s Crown, DiabloWikiBlind Faith, DiabloWikiAndariel’s Visage, DiabloWikiMempo of Twilight, and the DiabloWikiDeathseer’s Cowl, and all have a 16.67% chance of dropping. (That’s 1/6, even odds for each.) The Barbarian is different, since the DiabloWikiSkull of Resonance can only drop for him. Thus his odds are 15.38% for the other 6 helms, and 7.69% for the Helm of Resonance. (That’s 2/13 for the first 6 and 1/15 for the Skull of Resonance.)”

    there are five legendary helms > there are six legendary helms

    That’s 2/13 for the first 6 and 1/15 for the Skull of Resonance > That’s 2/13 for the first 6 and 1/13 for the Skull of Resonance

Comments are closed.