Real Legendary drop rates and other statistics for 9785 chars [Spreadsheet]

[Snackwiches]

The neck is so undervalued it's hilarious. It can prevent death several times in an encounter, it could save you 200 wipes in Mythic progress. It's a very good second Legendary for an entire raid team to use IMO.

Dont you know that healer mana is infinite and all their spells are instant cast? They should just heal everything!

[Reyzzz]

Some conclusions:
- The odds to get warforged or titanforged from mythic+ loot is exactly 50%

That droprate is skewed since the chance to get warforged item from the weekly chest is 100%.

[Pheros]

The worst thing about RNG is seeing the effort I put in compared to others. Since getting my hidden artifact appearance, I have done 52 Mythic plus runs. (I'm tracking them with my mod), did enough BGs to get the 1000 kills appearance and cleared normal and heroic Emerald Nightmare. Everyone in my Mythic Plus group has 1 or more legendaries. (The DK and Druid both have 3 and I've been present for all their drops).. so I get really salty when a Hunter in our guild who only logs on to raid gets a Legendary in our normal EN clear..

Where's my bad luck protection..?

The point is, rng doesn't work that way. Every chance is a separate roll of the dice. People might be starting NH early next year and still not have one by then. Even if your pity timer gets to like 20%, you could still conceivably not get one after 100 tries.

[Deleted]

So yes, in one sense you are correct, Pure RNG can not be predicted, but a computer is unable to produce a purely random number, They can use several methods to simulate it, from using atmospheric noise, to an algorithm that produces selection of seed numbers fed into another algorithm to produce a "random" number.

With enough time and data, (unless they are using some sort of complex interaction with the real world like random.org does) then it will be possible to work out the algorithm and thus predict it. With that being said, Blizzard will no doubt be tweaking and changing the system in the background at various points, which means any data collected previous to the changes that they make will make all the previous work invalid.

Computers not being able to produce random numbers has very little to do with this. Everytime you clear content that can drop a legendary, the computer rolls a virtual die. How many sides the die has can be predicted with data. What the next roll will be is impossible to predict. The data is interesting because it can help us determine what content has the die with the least amount of sides and whether it's worth farming.

[Beace]

Thanks for putting this data together, OP.

On another note, it always frustrates me when people who think they understand statistics and chance, try to argue with people who actually understands it. You tend to confuse both yourselves and others with weird convoluted examples. Just because something is random, doesn't mean you can't determine the chance that it'll occur.

[Toppy]

You do realize those number are scewed due to legends dropping at different rates in different content. More and more people are mythic raiding and doing mythic dungeons, which both have higher drop rates. The numbers were expected to go up.

[Deleted]

10000 from 10 million subs 0.1% sample, and people actually believe it like the holy bible.

google "sample size calculator", with this sample it fits in 95% cases (+/- 1%)

[Deleted]

Computers not being able to produce random numbers has very little to do with this. Everytime you clear content that can drop a legendary, the computer rolls a virtual die. How many sides the die has can be predicted with data. What the next roll will be is impossible to predict. The data is interesting because it can help us determine what content has the die with the least amount of sides and whether it's worth farming.

My post had nothing to do with which content is better to farm for them. it was regarding the point that "Pure" RNG in a computer system being is actually predictable with enough time and data. Even without the information in the spreadsheet it should have been plainly obvious that the harder the content the better the probability of getting a drop.

[Ornyx]

Interesting statistics, thanks for spending the time putting this together. It's one realm on EU, but would be an interesting comparison with other realms on a day-to-day to see the variance. Even then, it's not necessarily the whole picture of how Legendary drops work - after all, a percent-to-get is still encompassed by a percent-to-not-get every time, so not too much weight to be shifted there.

[Magemaer]

Awesome initiative and work Krawken.

One thing that bugs me is that 1-2% chance from mythic+ seems way higher than i expected. I expected it to be bellow 0.5%.
You excluded people that got loot from heroic or mythic raids, which has a high drop chance. Therefore this number (133 of 6414) includes drops from normal raid and everything else (cache, heroic dungeon, world drop etc).

This little poll (http://www.mmo-champion.com/threads/...your-legendary) seem to suggest that roughly a fifth of legendaries come from raids, but it's missing mythic raid and lfr as options. Maybe it's roughly 1/4. More than half of these legendaries-from-raid were excluded from your sample, those from heroic and mythic mode. So maybe 1/10 to 1/15 of these 133 came from normal and lfr?

This poll seem to suggest that almost a third (31.5%) of legendaries came from m+ chests. This number will be slightly higher since you excluded heroic and mythic raiders.
If we suppose 33% to 40% of those 133 legendaries came from mythics+, that would give a drop chance, very roughly, from 0.69% to 0.82%. Higher than i expected, but lower than 1%.

edit: a 0.75% legendary chance regarding m+ drops would give roughly one legendary every 104 3chest runs. This sounds about right to me, anecdotally.

[Aggrophobic]

The neck is so undervalued it's hilarious. It can prevent death several times in an encounter, it could save you 200 wipes in Mythic progress. It's a very good second Legendary for an entire raid team to use IMO.

It's a huge DPS loss though so it's a bit pointless.

You are better off deleting that neck, no doubt.

[impending doom]

That droprate is skewed since the chance to get warforged item from the weekly chest is 100%.

I guess you're right, that 50% just sounded so nice round number. I could've filtered out the "jackpot" chest, now it's just included in all mythic+ loot. So the WF drop rate from actual dungeons would be little lower. Although the weekly chest should count for something too.

One thing that bugs me is that 1-2% chance from mythic+ seems way higher than i expected. I expected it to be bellow 0.5%.
You excluded people that got loot from heroic or mythic raids, which has a high drop chance. Therefore this number (133 of 6414) includes drops from normal raid and everything else (cache, heroic dungeon, world drop etc).

This little poll (http://www.mmo-champion.com/threads/...your-legendary) ...

Yes the last statistic of 133 legendaries contains some from other sources too, but I tried to minimize these other sources by counting only drop rate of players who focused more on mythic+ than other sources. Also large part of people who answered that poll were probably below ilvl 847, and didn't farm mythic+ as much. So I'd still guess that over 50% of these 133 were from mythic+. Nevertheless that 1-2% is just rough estimate.

I admit that some of the numbers at the end part of these statistics aren't as straightforward as others, and with them need to be more careful how to interpret them. They just show average drop rates for people in different kind of activity groups, but do not completely remove some other criterias. Due to limitations of armory data, it's not very easy to calculate drop rates directly for some more specific activities.

[avoidconfusion]

I completed a normal mythic neltharions lair run yesterday with a PUG and the Warlock in the group got TWO Legendaries in that run. I shit you not. Talk about being lucky!

[Magemaer]

Yes the last statistic of 133 legendaries contains some from other sources too, but I tried to minimize these other sources by counting only drop rate of players who focused more on mythic+ than other sources. Also large part of people who answered that poll were probably below ilvl 847, and didn't farm mythic+ as much. So I'd still guess that over 50% of these 133 were from mythic+. Nevertheless that 1-2% is just rough estimate.

I admit that some of the numbers at the end part of these statistics aren't as straightforward as others, and with them need to be more careful how to interpret them. They just show average drop rates for people in different kind of activity groups, but do not completely remove some other criterias. Due to limitations of armory data, it's not very easy to calculate drop rates directly for some more specific activities.

That's why i took into account the self-reported source from legendary drops from that other thread. We can't know where exactly they dropped from unless we follow activity feed closely, and even then it's not accurate. Try to get a larger sample size of reports from players asking where they got their first legendary if they have it, the second and so on. Crossing this with data such as the one you gathered might give some us some valuable insight.

A major doubt i have is if it's there's a higher drop chance the higher the m+, and if it is how much? How does it scales? Is it more efficient than low+ 3chests? There is no sign whatsoever of an answer to that.

[Deleted]

This is great information, please make more!

[Typhoria]

The neck is so undervalued it's hilarious. It can prevent death several times in an encounter, it could save you 200 wipes in Mythic progress. It's a very good second Legendary for an entire raid team to use IMO.

The neck is dogshit. On most of the fights in EN you have a perma dot rolling on you or some other constant damage. At best you get one shield proc at the start of the fight. I have it, it sucks.

[Tech614]

The neck is dogshit. On most of the fights in EN you have a perma dot rolling on you or some other constant damage. At best you get one shield proc at the start of the fight. I have it, it sucks.

The shield is regularly 15% of my healing(as a Vegeance DH who does 150-200k HPS depending on the fight). That is as a tank. You're wrong, flat out The only fight you could argue it's bad on is Cenarius, if it can manage 10-15k hps to a tank on Ursoc it should be up all the damn time on any dps. The only fight with any kind of constant damage is Cenarius. Even then on Cenarius if wisps are cleanses you can easily sit in a small circle for 5 seconds to proc it.

If you're taking damage every 4 seconds on any other fight you're bad and doing something wrong.

[Typhoria]

The shield is regularly 15% of my healing(as a Vegeance DH who does 150-200k HPS depending on the fight). That is as a tank. You're wrong, flat out The only fight you could argue it's bad on is Cenarius, if it can manage 10-15k hps to a tank on Ursoc it should be up all the damn time on any dps. The only fight with any kind of constant damage is Cenarius.

Nythendra constant rot dot. Il'gynoth random targeted damage. Dragons of Nightmare, Permanent dot from being near a dragon. Elerethe Renferal random targeted debuff. Xavius soaking meteor. Apparently you are clueless about fight mechanics.

The shield doesn't reapply instantly if the old shield isn't consumed, there is still a 5 sec wait period. It is not reliable, it is not good. Its even worse as dps, since health pools are much smaller.

[Supercool]

You can drop the Dragons debuff if you're part of a portal group.

But ultimately if you're thinking in terms of healer mana saved, well, you'll save more with strong DPS legendaries by finishing the fight quicker.

[Tech614]

Nythendra constant rot dot. Il'gynoth random targeted damage. Dragons of Nightmare, Permanent dot from being near a dragon. Elerethe Renferal random targeted debuff. Xavius soaking meteor. Apparently you are clueless about fight mechanics.

The shield doesn't reapply instantly if the old shield isn't consumed, there is still a 5 sec wait period. It is not reliable, it is not good. Its even worse as dps, since health pools are much smaller.

20% of healing on Il'gynoth
22% of healing on Nythendra
17% of healing on Xavius

Grats you got one right, it was only 4% on dragons.

If we switch to Mythic from heroic, then Dragons % gos up and Nythendra % go's down. The rest are pretty universal.

- - - Updated - - -

You can drop the Dragons debuff if you're part of a portal group.

But ultimately if you're thinking in terms of healer mana saved, well, you'll save more with strong DPS legendaries by finishing the fight quicker.

Except no one is saying it's strong for saving healer mana. People are saying it's strong for saving you when you screw up. Much like the reasons you used to bring disc priests before this expansion even when their raw HPS was lower then other healers. Absorbs save lives and save wipes.

It's literally adding 225-275k effective health(or more like 450-500k if a tank) to you any time you don't take damage for 5 seconds, every 30 seconds. If you don't understand the value of that you probably don't raid lead or understand why you wipe 40 times on a hard boss before downing it.