[DE]Rebecca

Warframe Drop Rates Data

Recommended Posts

@[DE]RebeccaThanks so much for sharing the data!

A quick request: can you put a date/timestamp at the top of the page each time it's updated? A changelog would be *amazing*, but even that would be great.

  • Upvote 2

Share this post


Link to post
Share on other sites

I'm afraid in my experience, the observable rates in game grossly disagree with the data.

Since my last count of a certain Neo relic (B2) I have gained 29. It shares assumed equal rates in all listed locations I checked with its counterparts, including Neo B3 for the duration of the Vault opening.

I have seen zero Neo B3 relics.

Sadly I do not bear the statistical knowledge to confidently calculate what the combinatory odds of such an outcome is, but I know that I've done 15 two-wave runs of Xini today and seen at least 10 additional B2 relics and still zero B3 for those 30 potential drops at assumed-equal 9.09% rates.

 

Yes, that's anecdotal, yes it's plausible for a player to hit even the most abysmal odds, but if the veracity of this data is true, and the allegations of drop rates I've had argued before we had this data were also technically true for an equivalent source found through datamining (here I recall having to grind 8 hours between Axi V5s appearing when Valkyr Prime was fresh), then there has to be something algorithmically wrong with how the RNG creates and compares numbers against these supposed chances. One person can't anecdotally be subpar against the odds all the time without that cumulative probability becoming so infinitesimal as to be unrealistic.

  • Upvote 1

Share this post


Link to post
Share on other sites
16 hours ago, EDYinnit said:

I'm afraid in my experience, the observable rates in game grossly disagree with the data.

Since my last count of a certain Neo relic (B2) I have gained 29. It shares assumed equal rates in all listed locations I checked with its counterparts, including Neo B3 for the duration of the Vault opening.

I have seen zero Neo B3 relics.

Sadly I do not bear the statistical knowledge to confidently calculate what the combinatory odds of such an outcome is, but I know that I've done 15 two-wave runs of Xini today and seen at least 10 additional B2 relics and still zero B3 for those 30 potential drops at assumed-equal 9.09% rates.

Neo B3 has probability of 0.09%, that is less than 1 in a 1000. This means that your probability of getting Neo B3 *at least once* in n tries is 1 - (1 - 0.0009)n. This 0.0009 is not a mistake: you want 100% to be 1 so 0.09% is 0.09/100 (0.0009). With 30 potential drops this gives only a 0.0266 (2.66%) change of getting Neo B3. In contrast to this the Neo B2 with its probability of 6.32% will drop *at least once* in 30 tries with around 86% probability. Chance of getting 10 Neo B2 relics in 30 tries is rather low but I have to assume two things. First of: you may be exaggerating. :) Second of: there may be other mechanisms at play that change the odds (DE may decide to temporarily boost probability of certain drop for the benefit of the ecosystem or there may be warframe/mod/whatever that boosts probability of certain drop whether it's exposed in the codex or not).

Edited by repoocelad
  • Upvote 1

Share this post


Link to post
Share on other sites

Could we get updated data ?

I can understand not updating for every single hotfix if you have things to check manually, but a hotfix like this one affects *many* mission rewards.

 

On 31/07/2017 at 11:01 PM, EDYinnit said:

I'm afraid in my experience, the observable rates in game grossly disagree with the data.

Since my last count of a certain Neo relic (B2) I have gained 29. It shares assumed equal rates in all listed locations I checked with its counterparts, including Neo B3 for the duration of the Vault opening.

I have seen zero Neo B3 relics.

Sadly I do not bear the statistical knowledge to confidently calculate what the combinatory odds of such an outcome is, but I know that I've done 15 two-wave runs of Xini today and seen at least 10 additional B2 relics and still zero B3 for those 30 potential drops at assumed-equal 9.09% rates.

 

Yes, that's anecdotal, yes it's plausible for a player to hit even the most abysmal odds, but if the veracity of this data is true, and the allegations of drop rates I've had argued before we had this data were also technically true for an equivalent source found through datamining (here I recall having to grind 8 hours between Axi V5s appearing when Valkyr Prime was fresh), then there has to be something algorithmically wrong with how the RNG creates and compares numbers against these supposed chances. One person can't anecdotally be subpar against the odds all the time without that cumulative probability becoming so infinitesimal as to be unrealistic.

Personally, I've had no problem particular problem dropping Neo B3. But I haven't farmed Xini, so I can't comment on that particular node.

With a 9.09% probability, and 30 tries, that makes a 5.7% chance of getting no Neo B3. So nothing special, just unlucky.

Even my most awful unlucky streak, 66 radiant Neo B1/B2 without getting the Banshee prime neuroptics has only slightly under 1/1000 chance to happen. Are streaks like that extremely annoying? Yes. But sadly not statistically aberrant. And lucky streaks do not erase the frustration it generates.

 

7 hours ago, repoocelad said:

Neo B3 has probability of 0.09%, that is less than 1 in a 1000. This means that your probability of getting Neo B3 *at least once* in n tries is 1 - (1 - 0.0009)n. This 0.0009 is not a mistake: you want 100% to be 1 so 0.09% is 0.09/100 (0.0009). With 30 potential drops this gives only a 0.0266 (2.66%) change of getting Neo B3. In contrast to this the Neo B2 with its probability of 6.32% will drop *at least once* in 30 tries with around 86% probability. Chance of getting 10 Neo B2 relics in 30 tries is rather low but I have to assume two things. First of: you may be exaggerating. :) Second of: there may be other mechanisms at play that change the odds (DE may decide to temporarily boost probability of certain drop for the benefit of the ecosystem or there may be warframe/mod/whatever that boosts probability of certain drop whether it's exposed in the codex or not).

You're looking at the capture probabilities. For interception T3, they are both equal to 9.09%.

Edited by SubtleOaf
  • Upvote 2

Share this post


Link to post
Share on other sites
13 hours ago, SubtleOaf said:

Personally, I've had no problem particular problem dropping Neo B3. But I haven't farmed Xini, so I can't comment on that particular node.

With a 9.09% probability, and 30 tries, that makes a 5.7% chance of getting no Neo B3. So nothing special, just unlucky.

Even my most awful unlucky streak, 66 radiant Neo B1/B2 without getting the Banshee prime neuroptics has only slightly under 1/1000 chance to happen. Are streaks like that extremely annoying? Yes. But sadly not statistically aberrant. And lucky streaks do not erase the frustration it generates.

Again, I'm aware that confirmation bias is a thing and it's purely anecdotal. But all the same, I can scarcely remember the last time I was relic farming and got equivalent odds for the expectations when looking for the freshest batch despite 'equal' drops.

8 hours of Xini grind between Axi V5 drops. Some 4 hours between Axi B1, I believe. An entire unvaulting period in which a certain relic can just not show up while one of its counterparts in equal chance does so 29 times.

Oberon's relics came in just over par, but it was a welcome reprieve from the arduous tasks of the previous primes.

 

I'm just voicing concerns here; either the given data is strictly misleading and as @repoocelad suggestd, there are ancillary factors that further affect the actual, functional rates... Or it's entirely possible that the awarding algorithm is non-simple enough to have bugs.

DE will not divulge the inner workings of that, understandably as it could lead to exploitations, but if the code is old enough and there is space for a problem to arise (Bad seeding on the generator? Some spurious effect on the arrangement of potential rewards that affects the actual likelihood in use?) then no amount of happily plugging in internally decided odds to the flawed-but-assumed-sacrosanct system is going to give players a satisfactory experience.

I think we'd all appreciate it if we could at least be assured there are regular veracity checks in place to ensure that something so core to the experience as the reward RNG is not likely to have been bugged and corrupted during various patches. It's not even unprecedented. Remember the Endless Void having that "Every reward is Forma" period? I do, I'm still sitting on a sizable pile of shiny jigsaw pieces from it.

 

1 hour ago, repoocelad said:

EDYinnit didn't mention interception in his post, did he?

I did state Xini, which is Interception T3.

  • Applause 1
  • Upvote 1

Share this post


Link to post
Share on other sites

You're right, I missed that, sorry!

FWIW: bad random function is extremely unlikely to be the cause. I remember back in the day when (I think) Catan released on Xbox 360 people accused AI opponents  of cheating because they were getting what was perceived as impossibly lucky/unlucky streaks of rolls. Developers released pieces of code pertaining to randomness in the game for people to audit and, unsurprisingly, code looked very much correct (i.e. nothing shady was going on). Bugs are obviously possible but probably not in something this basic. There are probably stacks of lua code related to drops though and it is possible that game has bugs there.

Share this post


Link to post
Share on other sites

Please include drop location and rates for Broken War components (wiki claims these drop from Conculysts, but the droprate seems extremely low for such a rare enemy).

Share this post


Link to post
Share on other sites

First off, because I don't think I've said this here:
Thanks for this.
It's not perfect (yet), and - while acknowledging that I only have a partial picture of the situation - I'm still not happy about how the datamining shutdown situation shook out,
but given that this was done, the shutdown doesn't seem to have been an attempt to escape scrutiny/being 'kept honest', so I'm comfortable with holding my concerns in abeyance for the time being.

-----

A request that ought to be simple, and no risk of spoilers (and therefore automatible with no curation needed):

Please add a list of the weapons in the Riven pool.
No numbers, modifiers, whatever. Just which weapons are eligible.
(And ideally their disposition.)

Edited by Chroia
  • Upvote 1

Share this post


Link to post
Share on other sites

The current table needs to be updated, and if this is going to only happen with Major updates (e.g. U21, U22, etc.) then it needs to be more often, like with sub-updates (21.1, 21.2, etc. more along the release times of the Tenno Reinforcements) so people can reliably find things they need instead of looking at old drop table data that may or may not be accurate on many levels. Much like myself with 2 parkour bugs I've been following (one coming out with Octavia, another arriving with Harrow on PC), it lends to an unreliability in the entire system, which casts this effort to show transparency in a bad light.

One major reason they need to be updated is that the list has items in it that no longer drop (Vaulted relics, particularly), and they're making a lot of the tables appear inaccurate and not able to be wholly trusted. On top of that, if there's ever been a change to a drop rate for anything, people wouldn't find out until long after the change because major updates happen a few times a year instead of every week or two. This is especially a problem if there's another one of those "whoops" cases where a new item (like that new Machete stance) wasn't correctly put into the mission reward table it was supposedly added to, so people waste hours upon hours farming for something they literally cannot acquire.

If this is just a script to run, I presume it can be run relatively quickly, and that it reasonably only needs to be run every time there's a content drop (new weapons, vaultings/unvaultings, major feature additions, and so on), so it would take very little effort keeping the table up to date and maintaining a good image of transparency, while only needing to run it, realistically, once every week or two.

 

And as a side note, adding resource drop chances for both planetary resources (like Argon Crystals and Ferrite in the Void, for example) and crates/lockers (which also drop Nav Coordinates and Ayatan stars, for example) would be a welcome addition to the table, if such things are still being considered.

Edited by MrBubbleSS
  • Upvote 1

Share this post


Link to post
Share on other sites

Thank you for finally giving us infos on droprates like that.

However, in regard on how you've handled recent things with a fellow dataminer, i will still rely on my own data instead.

One more reason for me doing this is:

1 hour ago, MrBubbleSS said:

The current table needs to be updated, and if this is going to only happen with Major updates (e.g. U21, U22, etc.) then it needs to be more often, like with sub-updates (21.1, 21.2, etc. more along the release times of the Tenno Reinforcements) so people can reliably find things they need instead of looking at old drop table data that may or may not be accurate on many levels. Much like myself with 2 parkour bugs I've been following (one coming out with Octavia, another arriving with Harrow on PC), it lends to an unreliability in the entire system, which casts this effort to show transparency in a bad light.

One major reason they need to be updated is that the list has items in it that no longer drop (Vaulted relics, particularly), and they're making a lot of the tables appear inaccurate and not able to be wholly trusted. On top of that, if there's ever been a change to a drop rate for anything, people wouldn't find out until long after the change because major updates happen a few times a year instead of every week or two. This is especially a problem if there's another one of those "whoops" cases where a new item (like that new Machete stance) wasn't correctly put into the mission reward table it was supposedly added to, so people waste hours upon hours farming for something they literally cannot acquire.

If this is just a script to run, I presume it can be run relatively quickly, and that it reasonably only needs to be run every time there's a content drop (new weapons, vaultings/unvaultings, major feature additions, and so on), so it would take very little effort keeping the table up to date and maintaining a good image of transparency, while only needing to run it, realistically, once every week or two.

 

And as a side note, adding resource drop chances for both planetary resources (like Argon Crystals and Ferrite in the Void, for example) and crates/lockers (which also drop Nav Coordinates and Ayatan stars, for example) would be a welcome addition to the table, if such things are still being considered.

and

On 1.8.2017 at 10:59 PM, redeyedtreefrog said:

@[DE]Rebecca The drop chance resource appears to still contain the relics for the now-vaulted rhino and mag relics; will it be updated and if so, when?

 

You've stated:

On 28.6.2017 at 11:34 PM, [DE]Rebecca said:

Hello Tenno!

-snip-

 

This list will be maintained by an automated process combined with manual publishing with Updates (not all Hotfixes).

-snip-

Why not run a simple cronjob or run a script everytime you ship out an update/a hotfix?

 

Again: Don't get me wrong! Thank you for finally giving us this resource! It is really appreciated.

Share this post


Link to post
Share on other sites

i like that you released that list, ther is just on little problem, some of those drop chance valuse are confusing as hell, like the dropchance of blueprints.

Share this post


Link to post
Share on other sites
On 8/2/2017 at 10:26 PM, repoocelad said:

bad random function is extremely unlikely to be the cause

Not that unlikely at all.  Random number generation functions are as prone to bugs as any other aspect of a game (and can even be skewed by the way the functions are handed a seed) but vastly more difficult to detect as even a bugged random generator will look like it is working correctly while it really is not.  Even a single miss-keyed number, character or function input can vastly change the behavior of an otherwise working random number generator (RNG).

For example, a 50/50 test on a bell curve RNG will give the exact same 50/50 test results as that of a linear RNG, but clearly the bell curve isn't giving the same distribution of results as the linear RNG, despite it actually working.

Edited by Loswaith

Share this post


Link to post
Share on other sites
On 8/3/2017 at 1:34 PM, Zaeronine said:

Please include drop location and rates for Broken War components (wiki claims these drop from Conculysts, but the droprate seems extremely low for such a rare enemy).

indeed (and i asked about this some pages back too, i think).

when i'd do a wild-guess about it here, i would say DE is planing to put those two parts into a new sentient enemy type, which we will likely get with the plains addon - maybe even the big-boss there. my reason is, that if the numbers are correct on the droptable, there is no more room for any missing drops - not even for such ultra-rare ones like those two parts. I also remember that the chances were much smaller when lua (and therefore war and the conculysts) was released. on the other hand, since the drop chances on the mods (which is 10% for a mod) are still the same as before (mods, resources, ammo and bp's are all handled seperatly and can drop side by side too) - so it might indeed be only a case of "missing" information (we don't be told the chances for resources or ammo afterall too).

if someone would indulge us with a confirmation that the conculyst is still dropping those parts (and maybe with a screenshot too) it would be nice... I never got ANY bp from a conculyst (and i sure tried a lot with teams of ivaras nontheless).

Share this post


Link to post
Share on other sites

@[DE]Rebecca Can you possibly make a url like warframe.com/drop-rates, that redirects to the page you have? As it is, some people get (understandably) suspicious when they see that URL.

Thanks!

  • Upvote 1

Share this post


Link to post
Share on other sites

I am fairly new to this game and i am wondering what rotations are? Rotation A/B/C? does a rotation happen after completing the mission or during? This resource is ungodly helpful especially for a new player like myself trying to find argon crystals all this time.

EDIT: I got many wiki links to multiple ways to access the specific resources im trying to earn. Thanks everyone. :D

Edited by CharizardBud

Share this post


Link to post
Share on other sites
3 hours ago, CharizardBud said:

I am fairly new to this game and i am wondering what rotations are? Rotation A/B/C?

For Interception:
First message decoded, rewards are taken from Rotation A
Second message decoded, rewards are taken from Rotation A again
Third message decoded, rewards are taken from Rotation B
Fourth message decoded, rewards are taken from Rotation C
This reward pattern (A, A, B, C) would then be followed after every succeeding message decoded.

For Defense:
After the first 5 Waves, rewards are taken from Rotation A
After Next 5 waves, rewards are taken from Rotation A again
After Next 5 waves, rewards are taken from Rotation B
After Next 5 waves, rewards are taken from Rotation C
This reward pattern (A, A, B, C) would then be followed every 5 waves of defense.

For Survival:
First 5 minutes, rewards are taken from Rotation A
After next 5 minutes, rewards are taken from Rotation A again
After next 5 minutes, rewards are taken from Rotation B
After next 5 minutes, rewards are taken from Rotation C
This reward pattern (A, A, B, C) would then be followed every 5 minutes of survival.

For Defection:
First 2 squads, rewards are taken from Rotation A
After next 2 squads, rewards are taken from Rotation A again
After next 2 squads, rewards are taken from Rotation B
After next 2 squads, rewards are taken from Rotation C
This reward pattern (A, A, B, C) would then be followed after every 2 squads evacuated.

I hope this helps.

Edited by (PS4)Feox_PH

Share this post


Link to post
Share on other sites

I want to point out just one little thing... These DATA are Obsolete...

There are still the AXI R1 relic drop table, and so it is clearly outdated from 2 updates...

I simply hope that DE don't update this drop table only at Major updates, because otherwise people will be higly upset and disappointed

https://n8k6e2y6.ssl.hwcdn.net/repos/hnfvc0o3jnfvc873njb03enrf56.html

 

Just click on the link and look for AXI R1...

Share this post


Link to post
Share on other sites
On ‎30‎.‎07‎.‎2017 at 10:16 PM, Raf42 said:

A quick request: can you put a date/timestamp at the top of the page each time it's updated? A changelog would be *amazing*, but even that would be great.

@Raf42 maybe this can help.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now