What is this?

This is an RNG simulator. It's designed to run a scenario with random rewards until it's successful, start over and repeat it over and over again, and provide a statistical analysis of how long it actually takes to achieve success.

The purpose of this is to allow you to set your own expectations beforehand, as well as judging how lucky or unlucky you were after the fact. Nothing is absolutely guaranteed with RNG. However, there are statistically rigorous ways to set expectations and compare different scenarios, and that's what this page is designed to do.

Most of the drop rates used in the examples are from from the official Warframe PC Drops page.

Getting started

Select one of the pre-built scenarios from the Examples drop down at the top of the page.

Click the Start button to start the simulation.

You can stop the simulation at any time by clicking the Stop button.

The simulation runs the chosen scenario over and over again, up to a million times by default. While the simulation is running it will display a bunch of statistics. The most important ones are:

• Average: This is the average number of missions, enemies, Riven rolls, or amount of time that the scenario took to succeed. This is also known as the expected value.
• 90th Percentile and 99th Percentile: This is the number of missions, enemies, Riven rolls, or amount of time necessary in order to have a 90% or 99% chance of success.

Table of Contents

• Running
  • Stopping
  • Auto-stopping
• Results
  • Statistics
  • Distribution Chart
    • Saving a Chart
  • Querying the Data
• Scenarios In Depth
  • Mission Scenarios
    • Drop Event order
    • Defining the drop rate
    • Drop Count
      • Drop Count Range
    • Defining Success
      • Item Naming
    • Time
    • Bail Early
  • Enemy Scenarios
  • Riven Roll Scenarios
    • Riven Roll
    • Riven Criteria
    • Riven Generation Details
  • Saving The Scenario
• Notes on the Examples
  • Frames
  • Mods
  • Relics
  • Arbitrations
  • Rivens
  • Misc
• To Do
• Known Issues

Running

You can start the scenario at any time using the Start button.

Note that while the scenario is running you can't make any changes or load up another scenario.

The simulation itself is all in JavaScript, and runs entirely inside your browser. You can actually download this HTML file and run it offline; there are no required external dependencies.

Stopping

You can stop the simulation at any time using the Stop button.

If you create an extremely long scenario, like adding an Item with a 0.0001% drop rate, then the simulation can lock your browser tab while waiting for a single trial to finish. Just don't do that.

If you manually stop the simulation, and you haven't made any changes to the scenario, then you can start it again and it will resume where it left off.

Auto-stopping

The Stop after field allows you to specify a number of trials.

The simulation will automatically stop after hitting that number. By default this is one million trials, which seems to be a good estimate for when all the example scenarios settle in on something close to their true statistics.

If the simulation has been automatically stopped, and you haven't made any changes to the scenario, then you can start it again and it will run a second set of trials up to the Stop after number. In other words, if your simulation reaches a million trials and you start it again, then it will run a million more, for a total of two million. Starting again will run up to three million, etc.

Setting Stop After to 0 disables automatic stopping. The simulation will run until Stop is clicked or the page is closed.

Results

While the simulation is running, and after it has completed, it will update the statistical results at the top of the screen. If your scenario is a Mission Scenario that includes multiple missions, then each Mission will also get its own results section with statistics from just that mission's runs.

Each result section contains statistics for the total number of missions/enemies/Riven rolls. For a Mission Scenario this is a sum of all the missions that were run before success. For individual Mission results, this just counts the number of that mission that were run.

If time data was provided in a Mission Scenario, then each results section will have a second set of statistics for the time each run took. Again, the Mission Scenario time result contains the sum of the times for all the missions run, and the individual Mission time result only counts the time spent running that mission.

Each result section is further divided into two parts, Statistics and a Distribution Chart.

Statistics

The table on the right shows a number of statistics for the run:

• Average: This is the expected average of the distribution. This is the average number of mission, enemies, Riven rolls, or average amount of time until success.
• Median: The median is the midpoint of the data set, otherwise known as the 50th percentile. If the distribution has a non-zero skew then the median will be different from the average.
• 90th Percentile: The expected amount of missions/enemies/Riven rolls/time required in order to have a 90% chance of success.
• 99th Percentile: The expected amount of missions/enemies/Riven rolls/time required in order to have a 99% chance of success.
• Standard Deviation: Standard Deviation basically describes how spread out the data is from the average.
• Skew: Skew describes how slanted the data is towards one side or the other. A perfectly symmetric distribution has a skew of zero.
• Excess Kurtosis: Excess kurtosis basically describes how many extreme outliers there are compared to a normal distribution. A perfectly normal distribution has an excess kurtosis of zero.
• Trials: This shows the total number of trials that have been run to success.

• Best Case: This is the minimum result from the current simulation run. Below this statistic is shown the actual best case result, including the amount of missions/enemies/Riven Rolls/time, and how many of each Item ultimately dropped it the case of Mission and Enemy Scenarios, or what the actual Riven roll was in the case of Riven scenarios.

  The Best Case also includes a count of how many separate trials had that same result.

• Worst Case: This is the maximum result from the current simulation run. Below this statistic is shown the actual worst case result, including the amount of missions/enemies/Riven Rolls/time, and how many of each Item ultimately dropped it the case of Mission and Enemy Scenarios, or what the actual Riven roll was in the case of Riven scenarios.

The Best Case and Worst Case should be taken with a grain of salt. They just show the best and worst case from the current simulation run. Running the simulation over again with the same scenario can produce different best and worst cases.

• Last 5 Trials (Riven Scenarios only): This shows the last five Riven rolls that successfully passed the criteria.

• Criteria met (Riven Scenarios only): If there are multiple criteria on the Riven Roll, then this shows what percentage of the time each criteria was matched. 

  Note that these may add to more than 100% if the criteria are not mutually exclusive.

• Riven Probability (Riven Scenarios only): This shows the probability of rolling the Riven criteria given the data collected so far.

  Riven Rolls always boil down to a simple Negative Binomial Distribution, so this is simply 1 / Average.

Distribution Chart

The chart on the left shows a graphical view of the results. There are two possible views:

• Probability Density: This shows the probability distribution. The X axis represents possible results, and the height of the chart shows the relative frequency of that result.

  Most of the results tend to cluster on the left side of this chart, below the average, with a long tail of unlikely results extending very far to the right. The right edge of the chart cuts off at the 99.9th Percentile, but the Worst Case result shows you just how far it could extend if it was drawn out in full.

• Cumulative Probability: This shows the cumulative probability distribution. The X axis represents possible results, and the height of the chart shows the probability of getting that result or better, from 0% to 100%.

  In this chart the value increases and approaches 100% as it goes to the right. Like the other chart, this cuts off at the 99.9th Percentile to save space.

You can switch between these two views using the buttons above the chart.

In both cases, there is a highlighted vertical line which shows where the average falls on the chart.

The distribution of results basically follows a Negative Binomial Distribution, with Drop Events representing separate Bernoulli Trials. If the scenario only contains one Drop Event with one Item, then it's exactly equal to a Negative Binomial Distribution, and you can actually calculate exact values for Average, 90th percentile, etc. However, with more than one Item or Drop Event things get complicated to the point where it's easier to run a simulation than it is to try and calculate the exact statistics.

Saving a Chart

Most browsers should allow you to right-click on a chart and save it directly. However, if you simply click on a chart then it will be converted in-place to a PNG image, which you should definitely be able to save.

Querying the Data

Underneath the chart is a Query button. Clicking this button will display two fields:

• Result: A specific result, for example, how many runs you personally took to succeed at the chosen secnario
• Percentile: The probability of achieving the given result or better, as a percentage

If you enter a value in either of these fields and hit enter, it will calculate the other value. For example, if you have a particular result and want to know the percentile then you can fill in Result and let it calculate Precentile. Or, if you want to know a percentile other than the provided 50th, 90th, and 99th, then you can fill in Percentile and let it calculate the corresponding result.

Scenarios In Depth

There are three types of Scenarios:

• Mission Scenario: These scenarios are about running a mission or series of missions. These usually have some time duration associated with each mission.
• Enemy Scenario: These scenarios are about finding and killing a particular type of enemy.
• Riven Scenario: These scenarios are about rolling a Riven mod until its stats meet particular criteria

Mission Scenarios

A Mission Scenario is built up in three layers:

1. The Scenario contains one or more Missions. A Mission represents something that can be repeated as many times as necessary, like a Cetus Bounty or a Sabotage mission.
2. A Mission contains one or more Drop Events. A Drop Event represents something that triggers an RNG-based reward, like completing a survival rotation or finding all three Sabotage Caches.
3. A Drop Event contains one or more Items. An Item represents an individual object that the Drop Event can reward, like a blueprint or a mod.

Drop Event order

Each time a mission is run, each of its Drop Events is run in the order in which they are listed. A Drop Event can be repeated in the same mission by setting Events Per Mission. This is for when the Drop Event is something like an A Rotation that repeats twice per Mission.

The order in which Drop Events are listed should be the order in which they actually occur during the mission. See Bail Early below.

Defining the drop rate(s)

A drop rate is defined for each individual Item. This represents the probability of that Item being rewarded by the Drop Event containing it. The drop rate is a percentage, greater than 0% and less than 100%.

The sum of the drop rates for all the Items under a single Drop Event cannot add up to more than 100%. They can, however, add up to less than 100%. When the drop rates under a Drop Event add up to less than 100% then that means the Drop Event has a chance of not dropping anything we're interested it.

Drop Count

For each Item, Drop Count is the number of that Item that drops from a single Drop Event.

Most of the time you'll want to leave this at 1. For the vast majority of drops we care about analyzing, they drop one at a time.

Drop Count Range

There are a couple of "resource"-like items that drop many at a time, like Void Traces and Kuva. For this, you can actually specific a range in the Drop Count field. Range is specified as a minimum count and a maximum count, separate by a dash, like 6-30 or 550-700. Any number between the minimum and maximum, inclusive, has an equal chance of occurring.

You can tweak this by adding a comma and then an increment. This causes it to choose only numbers in that increment between the minimum and maximum, rather than every possible number. For example, 12-60,2 causes it to pick only even numbers between 12 and 60, and 500-750,10 causes it to pick only multiples of 10 between 500 and 750. In most cases this only makes a small difference, but it does make a difference.

Defining Success

An individual Item defines a Number Required. This is the total amount of that Item that must drop before that Item is considered successful. It's 1 by default, but you can make it higher to simulate how long it would take to get that many multiples of an item, or if the Items' Drop Count is something other than 1.

There are some more advanced success requirements you can set at the Drop Event and Mission levels. You probably won't have to mess with these.

• By default, if a Drop Event's Items Required field is left at zero, then we require that all of its Items under that Drop Event reach their required numbers. You can make Required Items greater than zero, in which case only that many Items have to be successful. Note that this is a count of unique Items, and does not count multiples of a single Item. The value of Required Items cannot be larger than the number of unique Items in the Drop Event.
• By default, if a Mission's Drop Events Required field is left at zero, then then we require that all of the Drop Events under that Mission reach their required Items. Like Required Items, You can make Required Drop Events greater than zero, in which case only that many Drop Events have to be successful. Note that this is a count of Drop Events, not individual Items. The value of Required Drop Events cannot be larger than the number of Drop Events in the Mission.

Item Naming

A Name is assigned to each Item, uniquely identifying it. A single Drop Event can't contain two Items with the same name. However the same Item name can appear in more than one Drop Event, and in more than one Mission. When this happens, an Item dropping in one Drop Event counts towards every Item in the Scenario with the same name.

Time

In addition to tracking the total number of missions it takes to get success, we can also assign a time duration to each mission and/or drop event. The time is given as a whole number of minutes.

• Mission time is used when a mission has some set amount of time it takes. This is the base amount of time a mission takes regardless of drop events/rotations.
• Drop Event time is used when a drop event itself has some set amount of time it takes to trigger, like a survival rotation. If a drop event repeats more than once then its time will be multiplied by how many times it occurred during that mission.

To determine the total amount of time a mission takes the base mission time, if any, is added to the times for each drop event that occurred, if any.

Some missions, like Survival and Onslaught, have a set amount of time they and their drop events/rotations will take. With other mission types you have come up with an estimate.

In the examples at the top of this page, I'm using the following time estimates:

• Exterminate: 5 minutes
• Capture: 5 minutes
• Assassinate: 10 minutes.
• Finding all three supply caches during Sabotage or Lua Crossfire: 10 minutes
• Spy if you're bad at Spy: 10 minutes
• Cetus bounty stage: 3 minutes
• Cetus Incursion, waiting for it to appear and then doing it: about 5 minutes
• Infested Salvage: 20 minutes to rotation C
• Onslaught/Elite Onslaught: 20 minutes to rotation C

Bail Early

If there are multiple Drop Events in a Mission, then it's possible for the later Drop Events to succeed during the run before the earlier ones. If there's a point in the mission after which all the Drop Events have already been successful, then we can bail early, ending the mission there and saving time.

This behavior is enabled by default in order to give us the shortest possible time. You can force a particular mission not to bail early by unchecking the Bail Early checkbox under the mission.

Prevented Consecutive Repeats

In some special cases (Kuva Liches) they have added the ability to prevent consecutive duplicate rewards. Setting this to 1 will enable that behavior in the simulation.

You can also set this value to more than 1, if you want to prevent duplicate rewards going back the given number of consecutive drops. Note that Prevented Consecutive Repeats must be less than the number of items in the drop.

Enemy Scenarios

An Enemy Scenario is basically a simplified Mission Scenario, where the mission and drop event are just finding an enemy and killing it.

An Enemy Scenario just has single Enemy to kill, and a single Item list under the Enemy. You still have access to Required Items to control how many of the Item drops need to succeed for the scenario to succeed, but there is no time duration or any other settings.

Riven Roll Scenarios

A Riven Roll Scenario tracks how many times a Riven needs to be rolled in order to meet certain stat criteria. A Riven-based scenario is built up in two layers:

1. A Scenario contains one Riven Roll. A Riven Roll represents a single Riven Mod that can be rolled as many times as necessary.
2. A Riven Roll contains one or more Riven Criteria. A Criteria represents a list of stats that are required or wanted as either positive or negative stats on the Riven mod.

Riven Roll

The only configuration for the Riven Roll is what kind of Riven it is, Gun or Melee. This determines the list of stats to choose from.

Riven Criteria

The Riven Criteria contains a few configuration fields followed by the full list of possible stats for the Riven type.

Required Positives and Required Negatives can be used if you want a specific number of positive or negative stats. For the positive stats, you can require 2, 3, or accept both 2 or 3. For the negative stats, you can require 0, 1, or accept both 0 or 1.

For the stats themselves, they are listed in alphabetical order. For each one, one of four options is available:

• ++: Required Positive

  This marks a stat that's required to be among the Riven's positive stats. If this is selected for three stats then the Riven is automatically required to have three positive stats, and all three of the marked stats must be present.

  You cannot select more than three required positive stats.

• +: Wanted Positive.

  This marks a stat that you want to be used in the Riven's positive stats. You can mark as many stats as you like this way; the only requirement is that the Riven's positive stats come from the set of Wanted Positives.

  If there are less than three total Required and Wanted Positives, then the remaining positive stats in the Riven can come from the Don't Care set.

• +: Wanted Negative.

  This marks a stat that you want to be used as the Riven's negative stat. You can mark as many stats as you like this way; the only requirement is that the Riven's negative stat comes from the set of Wanted Negatives.

  If there are no Wanted Negatives, then the Riven's negative stat can come from the Don't Want set, unless Required Negatives is set to zero.

  If there is only one Wanted Negative marked, then the negative stat on the Riven, if present, must be that stat.

• ?: Don't Care

  The Don't Care category is everything else. If there aren't enough Wanted/Required Positive stats or Wanted Negative stats to fill out a Riven, then the remaining stats can come from the Don't Care list. Otherwise, stats marked as Don't Care will not appear in accepted Rivens at all.

The Riven Criteria is only concerned with which stats appear on the Riven, not their actual numeric values.

Riven Generation Details

I've had to make a few assumptions about how Riven stats are generated, because the details are not published like other drop rates.

• Two vs Three Positive Stats: I'm assuming that the choice between two or three positive stats is just a 50/50 chance. My own my own sampling of about a hundred rolls on various Rivens comes close enough to 50/50 to make this likely.
• Zero vs One Negative Stat: I'm also assuming that the choice between zero or one negative stats is a 50/50 chance. Again, this is backed up by my experience.
• No correlation between the number of positive stats and the number of negative stats. I'm assuming that all four combinations of 2/3 positives and 0/1 negative are equally likely.
• Stat probability: I'm assuming that, on each roll, all applicable stats have the same probability of being chosen.
• Negative Stat Exclusions: I'm assuming that the following stats never appear as negative stats:
  • Cold Damage
  • Electric Damage
  • Heat Damage
  • Toxin Damage
  • Punch Through (This one has been confirmed)
  • Combo Duration

Saving The Scenario

Use the Generate Link button to build a link to the current Scenario.

This link can be copied and shared. Opening it will pre-fill the page exactly how it currently appears.

If you check the Auto Start checkbox then the link will automatically start the Scenario when opened.

Notes on the Examples

Frames

The frame scenarios are listed roughly in order of how long each one takes.

Just guessing, but I think the line between being faster to farm the frame directly and being faster to run Tridolons, sell Arcanes for Plat, and buy the frame, is somewhere between Nidus and Ivara. I still recommend spending some time farming Ivara, if only for the Spy practice. However, Khora is firmly on the plat side of that line, with very little else to get from that mission besides maybe ranking up other warframes.

Standard Boss Frame:	There's a lot of variance in how long the various assassination missions take, so I just guessed an average of 10 minutes.
Ash:	This is not something you can actively farm, but it's interesting that it doesn't actually take that many Manics to get.
Gara:	This is an example where the same Item can drop from multiple Drop Events, since each Gara part drops from multiple bounty stages with a different drop rate in each stage.
Garuda:	Garuda's grind is comparabe to Gara, at least for getting the blueprints. Getting the materials is another matter entirely.
Protea:	Each part takes nine tries on average, but there's going to be a lot of variance with the time depending on exactly which Corpus Ship missions you choose. I'm guessing five minutes for each run assuming you pick a capture or exterminate but have to wait for the Treasurer to appear and drop a coin each time.
Octavia:	Getting the Neuroptics from ODS is pretty straightforward. For the other two parts I combined the Lua cache hunt with the Music Room hunt, as you'll probably find the latter by accident while looking for caches. I took a wild guess that the Music Room spawn rate is about 25%. The scenario splits up the second mission into five minutes finding the caches and five minutes just doing the mission. So if the Systems drops from the caches first then it bails early and only spends five minutes running the mission and looking for the music room.
Revenant:	Like Gara, Revanant's parts drop from multiple bounty stages with different drop rates in each one. Revanant takes a little longer than Gara on average because of the longer bounties and the rare status of the third part.
Revenant (no bail early):	The Gara and Revanant scenarios above both enable Bail Early to cut down on time a little bit. This is an example of what happens when you disable Bail Early on these kinds of multi-rotation missions. In this case, bailing early seems to save about 10-15 minutes on the average, from 2.8 hours down to 2.6. Also, not bailing early makes the probability density chart for the total time look totally cool.
Gauss - 8 rounds:	Doing eight rounds of disruption gives you three times as many rotation Cs as other rotations, bringing this grind down to manageable
Harrow:	The Chassis is basically a given if you do any amount of relic cracking, and the Neuroptics isn't bad if you're decent at spy or can get a friend. The real killer with Harrow is the Systems, with an expected 3 hours in everyone's least favorite game mode.
Equinox:	Everyone's favorite example of a heavy grind is at most the fourth worst.
Gauss - 4 rounds:	If you do a minimal four rounds on each disruption mission then you only get 50% rotation Cs, making your expected grind time for Gauss slightly worse than Equinox. So do more than four.
Nidus:	It varies how long it actually takes to hit rotation C on infested Salvage, depending on how on top of it your team is. I gave it a nice round 20 minutes.
Ivara:	I'm estimating 10 minutes for a Spy mission but if you're good at Spy, or have a friend who's good at Spy, then they can be cut down to five minutes. However, even with the ten minute estimate Ivara is still second place.
Khora:	20 minutes to get to Rotation C is somewhat of an underestimate, but good enough. Disabling early bailing adds about 20 minutes to the expected average, bringing it close to an even ten hours.

Mods

There are a ton of very low drop rate mods. A lot of them drop from common enemies you'll run into thousands of times just through normal play, which somewhat lessens the impact of the low drop rate. Most of the others are Melee Stance mods.

The nice thing about mods is that trading is always available as a check on bad luck. On a long enough time line your luck will even out, and if you're unlucky with one mod then you might have already gotten lucky with a different rare mod and have an extra copy to trade. This is especially true with mod sets that all drop from the same place, like Corrupted Mods, Kela de Thaym's mods, Razorback, etc. By the time you hit the expected average number of runs you'll probably have enough to trade for what you're missing.

Argon Scope:	This one wouldn't be that bad if it wasn't time limited and only once a year or so.
All Mods From An Acolyte:	It takes only slightly longer on average to get all four mods from an Acolyte.
Fetch:	Fetch seems pretty benign, but no idea what the spawn rate of Orb Vallis Hyenas is.
Four Riders:	You have to kill a heck of a lot of Manics on average to get this stance mod.
Vengeful Revenant:	This shows why loot frames like Ivara and Nekros are almost mandatory when farming Conculysts.
Bullet Dance:	You have to find all three caches just to get a chance.
Tempo Royale (Bursa):	Is it faster to wait for Baro to bring this?
Tempo Royale (Incursion):	...Yes.
Condition Overload:	The simulation gets pretty slow when you start getting into the really rare drops like this.
Guardian Derision:	I still don't have this one.
Full Hunter Set (Ghouls):	A full set of these is going to take a little bit of time, if you can catch a Ghoul event.
Full Hunter Set (Plague Star):	It's much if you can manage to catch a Plague Star event, assuming you can kill four Hemocytes pretty quickly.
Full Synth/Mecha/Tek Set:	The Synth, Mecha, and Tek sets all have the same drop rates on different bounty rotations. You'll probably get a full set of each just running Orb Vallis bounties for other stuff.
All Corrupted Mods:	This is why trying to obtain all the corrupted mods without trading is crazy.
Profit-Taker: All Mods:	Getting all of the new mods from the Profit-Taker heist without trading is a pretty grim prospect. For time estimates, I'm just guessing rough best-case scenarios of 5, 2, 5, and 10 minutes for each phase, respectively.

Relics

The transition from the old Void Key system to Relics was always less about reducing grind and more about increasing gameplay variety. Personally, I don't miss spending twelve straight hours in T3 Survival, like, at all.

Giving newly released Relics a dedicated mission with high drop rates is great and I wish they did that with more Relics. It's way better than Void Fatigue Xini Fatigue.

That being said, Relic supply is still usually the bottleneck. You can stockpile Void Traces but you can't stockpile newly released Relics. So refining Relics to Radiant for the uncommon reward, which almost doubles the chances, is still worth it.

Radshare:	Radiant shares aren't bad on average. However, I've personally seen it take 20 radshares to get the rare, which put that particular run in the 99.98th percentile.
Solo Radiant:	And this is why you find Radiant shares.
Uncommon Intact Share:	If you just want an uncommon reward, is it still worth it to refine the relic to Radiant?
Uncommon Radshare:	...Probably.
Three Axi C3/G1 Relics (Bounties):	The new Axi relics have a very nice drop rate in the Cetus bounties, assuming you can catch the right rotation.
Three Neo G1 Relics (Bounties):	Some of the new Neo relics are also relatively easy to get in Cetus bounties.
Meso R1 Relic (IO Defense):	Getting one is not bad on average but the outliers are a killer.
100 Void Traces:	How long does it take to get 100 Void Traces?
1000 Void Traces with a Booster:	How about 1000 Void Traces with a 2x Resource Booster?

Arbitrations

The drop table and rotation pattern is the same for all Arbitration missions regardless of mission type. However, time estimates will vary based on which mission you're doing and how long you go for.

The endless C rotation means your time efficiency goes up the longer you run each mission, but you're going to have to pick some cutoff for the simulation to work with.

So far I've only tried survival and interception. Most of the other game modes seem to be bugged in some way that makes them take much longer to get the same number of rotations.

Arbitration Mods (Survival 30 minutes):	Doing a Survival for 30 minutes gives you one A, one B, and one C rotation.
Arbitration Mods (Survival 1 hour):	Doing a Survival for an hour gives you one A, one B, and four C rotations.
Arbitration Mods (Survival 2 hours):	Doing a Survival for two hours gives you one A, one B, and ten C rotations.
Arbitration Mods (Interception 1h):	With a good team you can get about 14 rounds of Interception done in an hour, which gives you one A, one B, and five C rotations.
Aura Forma (Interception 1h):	You're probably looking at around 6 hours of Arbitration per Aura Forma.

Rivens

Before you get too worked up over the outrageous averages on some of the Riven examples, bear in mind this fact:

Rolling one Riven 50 times has the same odds as rolling 5 similar Rivens 10 times each.

Given the results in some of the Riven Roll examples here, I highly recommend going the latter route: Just putting up to 10 rolls into Rivens as you get them and moving on, saving more than 50% of your Kuva. You're bound to get some "acceptable" Rivens you can use while waiting for that big roll.

90 Rivens:	How many Sorties does it take to fill your Riven collection to capacity? This one's interesting, as it's very close to a normal distribution
Riven for a Specific Rifle:	How many Sorties does it take to get a Riven for your favorite non-shotgun primary, out of 64 possibilities as of November 2018?
Basic Damage Gun Riven:	Just a basic damage Riven for a gun isn't too hard to roll.
Decent Melee Riven:	A decent Melee Riven is also not hard to roll if you're not picky.
Crit or Status Melee Riven:	When rolling a Zaw Riven, I get the best results by rolling until I get either good crit stats or good status stats, and then building a Zaw accordingly.
Status Gun Riven:	If you're picky about your status Riven then it may take a few rolls.
Crit Gun Riven:	Even more if you're you're picky about your crit Riven.
Godlike Lanka Riven:	This is why players pay a very high price for these.
Godlike Vectis Riven:	If you absolutely have to have negative magazine capacity.
Perfect Rubico Riven:	Getting super ridiculous now.
God's God's Lanka Riven:	Someone told me this is the Lanka Riven to shoot for. They are crazy.
Riven with Exact Stats:	Okay, so you're going to have to accept some wiggle room in the Riven stats you're looking for. If you absolutely must have a Riven with three specific positive stats and one specific negative stat, then you will be rolling it for an astronomically long time. Honestly, this is mostly here as a joke. It's just past the point where the simulation really starts bogging down on my laptop. Still, it's interesting to watch each data point appear one-by-one in slow motion after hundreds of thousands of rolls.
Riven with Exact Stats (Experimental Stat Locking):	This is totally an experiment, and does not reflect how Rivens actually work in the game. The idea is to roll a particular stat 10 times to lock it, and then roll another stat 10 times to lock it, etc, until you have rolled all three positive stats and one negative stat. There are some edge cases not accounted for in the scenario I've built here, but they only negligibly affect the final result. There are a number of experimental simulator features enabled for just this scenario, including stat locking itself. They don't have full documentation yet because they're pretty out there.

Misc

All Shedu Parts:	This has got to be wrong. I have 3/4 parts and I'm definitely not close to killing 200 of these suckers.
All Ropalyst Drops:	Depending on how fast you can do the fight, this isn't a total slog.
Miter:	I bet you forgot about farming all five pieces of this thing.
All Hydrolyst arcanes:	Honestly you're probably better off waiting for the next Scarlet Spear event.
Imperator Vandal:	Not that bad on average.
All Granum Void Drops:	The actual "mission" takes very little time, it's all the hanging around a Corpus ship waiting for the Treasurer that makes this a grind
Wolf Sledge:	The drop table is misleading because it's on top of just a 35% chance to drop any part at all.
All Razorback Drops:	The two ultra-rare mods that you can only obtain from the Razorback are what make this painful.
All Shadow Stalker Drops:	This particular scenario ends up having the highest Excess Kurtosis of all the examples, which explains the surprising number of players still missing some Stalker drops after playing for years.
Broken War Parts:	Again, bring a loot frame to farm Conculysts.
Braton and Lato Vandal:	Get ready to cry.
All Kuva Lich Weapons:	At at least a couple of hours per try, this is a challenger for heaviest grind.
Legendary Core:	It's legendary for a reason.
Xiphos:	Possibly the most ludicrous grind in the game.
All Necramech Mods:	Oh man there's a new contender for ridiculous grind, and it's the Necramech mod set. It takes 30+ minutes just to run all three tiers once.

To Do

• I'd like to be able to make some kind of scenario that combines Relic Farming, Void Trace Farming, and Fissures into one grand total, but the process of consuming items and going backwards in the mission sequence when you run out of something makes things more complicated. A scenario that combines Kuva Farming and Riven Rolling is tough for the same reason.
• When you switch Riven types I wish it would keep any selections under the stat list that are common between the two types rather than just wiping it all out and starting fresh.
• A whole bunch of performance optimizations, probably.
• I'm not sure the Items Required and Drop Events Required fields need to be here. I'm pretty sure they're always set to 0 on every example, meaning all Items and Drop Events are required.
• Make the statistics section update in place rather than rebuilding the entire section every time.
• Enhance the chart so it doesn't have to start at zero on the left. Right now, distributions with a high average and small standard deviation end up as indistinguishable spikes all the way to the right.

Known Issues

• The charts don't work in Microsoft Edge. Don't use Edge.
• Sometimes the Generate Link pop-up doesn't disappear correctly on Chrome if you switch tabs at the wrong time. Shrug Emoji.
• A Riven scenario runs somewhere between a hundred and a thousand times slower than a mission or enemy scenario with a similar average. It's Javascript, there's only so much optimization you can do.
• Running a lot of simulations on a mobile device will really tear your battery a new one. That's just the way it is.
• The simulation pauses if the browser tab doesn't have focus. I'm not using WebWorker because I want everything to be in one file.