Jump to content
Dante Unbound: Share Bug Reports and Feedback Here! ×

The Current Drop System And How To Improve It


Notso
 Share

Recommended Posts

Quick links:

01 - The Current drop dystem

02 - Alternative drop systems

03 - A happy compromise

04 - Appendix

 

 

00 - Introduction

 

Notso, why should I read this wall of text?

In it, I will be analysing the current drop system; how it works, its pros and cons, and then I will look some alternative systems found in other games, and looking at there pros and cons. At the end of the day, I hope you'll all help me to refine my solution you'll find in section 03.

 

That's nice Notso, but why isn't this thread in Gameplay Feedback?

First of all, this topic is a very important one, juding by the sheer volume of threads complaining about; having to do 100 runs for a blueprint; how DE is killing the game by making two new items obtainable only through the Alerts system or with Pt, two I tell you!0

Edited by Notso
Link to comment
Share on other sites

01 - The current drop system

 

In the current system every drop has a fixed probability of occurring1; for example, the chance for a boss to drop a Chassis Blueprint is 1/15. This also means the chances of getting the BPs or mods you need depreciates as you get the BPs or mods you need2.

 

Positives


  • It's simple and intuitive for new players
  • Uncertainty in how long it will take to get a drop incentivises players to spend Pt
  • The alert system creates diversity in the ways players can acquire weapons or Warframes - this is good because some methods may suit some players more than others
  • BPs for the same 'frame drop from the same boss - this reduces the amount of time needed to get the BPs for a single 'frame

Negatives


  • The same uncertainty that increases revenue frustrates many players in a "bad" way3, causing them to uninstall the game, quite obviously reducing revenue
  • There will always be some players who end up needing 3 or 4 times the expected number of drop events to get the BPs or mods they need - this is not acceptable
  • The alert system feels far too random for many players, again leading to "bad" frustration
  • Mods drop equally from all sources4 - this hugely increases the uncertainty and time needed to get a specific mod, or to farm mods of similar or the same type for mod fusion
  • After the server decides a mod should drop, the mod that is picked up is different for each player - this leads to:
  • Frustration - if a player in a mission with me gets a Thunderbolt mod, from an enemy I killed, and I get a Sentinel Redirection for the umpteenth time
  • Time-wasting - I have to go searching after every mod just in case it's a good one. If the mod was the same for all 4 players then the first player to pick it up could say guys it's just another Sentinel Redirection]
  • All the BPs for a single 'frame being dropped from the same boss increases boredom/repetativeness for the player
  • Subconsiously, players who have a spate of bad luck will always worry that there is some hidden system they're not aware of and therefore not able to utilize, causing frustration
Edited by Notso
Link to comment
Share on other sites

02 - Alternative drop systems
 
Token System
This system is one whereby completion of certain events grants players a certain type of token which can later be exchanged for different goods in the Market. For example, killing a boss might grant the player a Completed Assassination Contract [CAC], 10 of which can be exchanged in the Market for a Warframe Chassis, while completing an alert mission might grant a certain number of Fast Response Tokens [FRTs] - the rewarded number of FRTs could vary by Alert.
 
Positives

  • Eliminates uncertainty in drops, where it is applied - taking the Alert example, it allows a player to see an Alert and think "Ah, if I complete this Alert I will have X% progress towards getting that Vauban Chassis Blueprint!"
  • Eliminates the possibility that a player may be cursed by the Random Number Gods Generator
  • Creates variety in gameplay - players can experience all game modes while still working towards the same goal
 
Negatives
  • Elimation in uncertainty reduces "immediate revenue" for DE
  • Not suitable for all drop types5

 
 
Positive feedback loop
Believe it or not, this Random Number God Generator is benevolant. Taking BPs as an example; as the number of times a player has killed Jackal without getting a Rhino Chassis Blueprint, so does the probability of him getting a Rhino Chassis Blueprint.
 
P5zvvmU.png
 
You can see that in this instance, after about 40 runs of Jackal without getting a Rhino Chassis Blueprint, the probability of him getting one is about 98%. The trade off though is that after this many runs, the probability of the other BPs dropping is split between that 2%, making the probability of getting the other two BPs 0.13% each6.
 
This particular curve is sigmoidal7, but at the end of the day the exact formula or type of curve used to generate the probaility increase doesn't matter - I picked this one because I find it easiest to see how it would work.
 
Positives
  • Completely eliminates players having to run 100+ times for "that Goddamned Ember Helmet" [and others]
  • Similar enough to fixed drop system that players will not be alienated by it
Negatives
  • Much more complicated to get working properly
  • More server intensive
 
 
Faction-specific drop tables
Certain factions or planets to be limited to dropping certain things from the resource and mod categories only. For example Attack Polarity mods drop only from Mars, Uranus and Pluto.
 
Positives
  • Allows players to focus their mod farming for the purposes of mod fusion [as same-polarity mods are great for fusion]
  • Allows players to focus their mod farming for acquisition of specific mods, or resource farming for specific blueprints.
  • Does not comepletely remove random aspect from the game; revenue maintained for DE
  • Information available in-game on which items drop from which planet reassures players that they're not wasting their time

 
Negatives
  • While redicing it, this solution does not completely remove the aspect of minority of players having to kill tens of thousands of enemies to get the mod they want, or having to open hundreds of lockers before getting the next Rubedo stack

Edited by Notso
Link to comment
Share on other sites

03 - A happy compromise [TL;DR]

 

Firstly, let me say this - any solution has to end up at the end of the day with DE still making money. We cannot comepletely remove the grind from this game. With that said, let's take a look in turn at the four most important "drop" types found in Warframe - Blueprints, Mods, Resources, and Alert ? Rewards.

 

Blueprints [excluding ones acquired from ? Alerts only]

In this instance, I would impliment the positive feedback loop. Firstly, it does not change the amount of time the average player has to spend in order to farm all three BPs for his or her next 'frame. However if this was to be implemented, a problem arises when we have two players in the same assassination mission who both haven't got their last BP, but their last BPs are different. In this case I would simply make BPs follow the way the current mod system works - the BP dropped is determined uniquely for each player.

 

Mods

Make certain planets drop certain polarity types - one [for high-level planets], two [for medium-level planets] or three [for low-level planets] types for each planet [except Mercury, which should drop all types]. This not only makes sense with the lore of the game [the lore states that Polarity types are linked to the 5 Orokin clans - building on this each clan's territories could have consisted of different planets, so naturally you'd find their mod types there], but it allows players to focus their mod-farming and feel that they're doing the farming in the correct manner. This information should also be clearly stated in-game, just as it is with resources.

 

? Alerts

Undoubtedly, this is the least controversial mechanic in the entire game. I say it's uncontrovsial because there doesn't seem to be a single person [DE included] who is happy with it in its current state. That said, its basic principle is not only fun [as it allows players to acquire weapons/Warframes they want while experiencing all possible game modes], but also practical, as it allows players with limited play-time to still get new weapons or Warframes through their own efforts rather than simply buying it [the former is a lot more satisfying].

 

My solution [which I came up with by myself, although I now realise it has been suggested on these forums already quite a few times (what rock have I been living under?)] is to make alerts give out a random number tokens which can be redeemed in the market for three levels of reward specification. The first level is the cheapest [dirt cheap], and grants a random reward from the entire alert pool. The second level allows the player to specify the type of reward [from Artifacts (cheapest), Warframe Blueprints Weapons, Helmets (most expensive)]. The third level is to specify the exact reward - while this removes all probability from the equation, the trade-off should be that it costs more tokens than if you'd done got it from a level 1 reward8.

 

Another thing to be taken into consideration with the token system is the need to keep inflation down9, as people will end up with so many tokens that they never need to spend Pt. To avoid this, we can either put an expiry date on tokens, or make people wait 2 weeks or so after an item is released before it is available to buy with tokens. Or both. Clever calculation on the average rate of token income could also reduce inflation.

Edited by Notso
Link to comment
Share on other sites

04 - Appendix

 
1 To determine what drops, each possible drop is assigned an interval of numbers, and then a random number generator spits out a number, and whichever item is assigned to that number is dropped. For example, in Jackal's loot table, Rhino Chassis may be assigned to numbers 68 to 134. The RNG spits out a number from 0 to 1000, giving Rhino Chassis a 1/15 chance of being picked.
_____________________
 
2 The number of trials, N, we expect to need before getting all the drops we want can be calculated using this formula:
 
N = P(Drop)^-1 x n x Hn
 
Where P(Drop) is the probability of a drop, Hn is the harmonic sum from 1 to n, and n is the number of drops we want, assuming that all drops of  the same type drop equally often.
 
For Blueprints:
N = P(Blueprint is dropped)^-1 x H3 = (1/5)^-1 x 3 x (1 + 1/2 + 1/3) = 27.5
N = 28
_____________________
 
3 Games will always lead to frustration due to difficulty. However there are two types of difficulty fuelled frustration in games, one of which is good and one of which is bad. The good type is the kind of frustration from difficulty which players can do something about. For example, a particularly difficult boss fight may cause a player frustration intially, but this frustration can be overcome by the player practicing and/or learning the weaknesses of the boss.
 
The bad type of frustration is that which arises form difficulty which cannot be overcome. Usually this is from being cursed by the Random Number Gods Generator - there is nothing the player can do about this, expect keep trying. As we've seen on the forums time and time again, sometimes they have to keep trying for far far too long [for example the player who had killed 24000 infested on Xini and not got a single Thunderbolt mod].
_____________________
 
4 I remember Steve saying that mod types shouldn't drop more form one place than another, however I cannot for the life of me remember where. If anybody has more up to date information please link it. Not the wiki though - that thing is full of unconfirmed BS and what amounts to the particular author's "feeling" of how often a mod drops.

_____________________
 
5 Mostly unsuitable for mods, but the same problem is encountered with BPs too. Making one type of token for all mods would reduce the amount of time to max all the mods you want far too much [due to being able a specific mod with a token acquired from anywhere]. If we want to limit this we either need to make mods more expensive in terms of tokens, which increases the time to collect all mods too much, or make more different token types for different polarity types, which is cumbersome and user-unfriendly.
_____________________
 
6 It sounds like it should be 1% each, but remember that the probability of a specific BP dropping is actually 1/15:
 
2% x 1/15 = 0.13%
_____________________
 
7 If anybody is interested, here is formula I used to get this curve:
 
V71PUga.png
_____________________
 
8 What do I mean by this? If in the first level reward tier the expected number of tokens it is necessary to spend to get, say, the Glaive is 500, then to specify the Glaive as a reward in level 3 would cost ~750 tokens.
_____________________
 
9 Credit to Mickey1779 for pointing out the inflation problem with a token system.
Edited by Notso
Link to comment
Share on other sites

Just read the entire thing and it is informative.

Most of the suggested changes I would be happy with.

+1 all around the board.

Only thing I will say is that I personally hate the Token systems, I know they make a lot of sense but something about them just rubs me the wrong way. Would be better than the current alert system, but I think bosses should not be a part of it, the positive feedback loop seems like a much better way to handle those.

Link to comment
Share on other sites

Just read the entire thing and it is informative.

Most of the suggested changes I would be happy with.

+1 all around the board.

Only thing I will say is that I personally hate the Token systems, I know they make a lot of sense but something about them just rubs me the wrong way. Would be better than the current alert system, but I think bosses should not be a part of it, the positive feedback loop seems like a much better way to handle those.

 

I think that's what I said in the conclusion? Or are you agreeing? o.0

Link to comment
Share on other sites

I think that's what I said in the conclusion? Or are you agreeing? o.0

Agreeing. But just stating it again because you in one of the sections mentioned also using tokens from bosses to buy Warframe parts.

Or maybe I misread/understood that part and you were just talking about the Vauban and possibly the Banshee

Link to comment
Share on other sites

Agreeing. But just stating it again because you in one of the sections mentioned also using tokens from bosses to buy Warframe parts.

Or maybe I misread/understood that part and you were just talking about the Vauban and possibly the Banshee

 

No in that part I was just using it as an example of things it could be used for

: ]

Link to comment
Share on other sites

No in that part I was just using it as an example of things it could be used for

: ]

Glad for the clarification, and I do really hope this thread gets more views/ a DE response, I feel this would help improve the game quiet some.

Link to comment
Share on other sites

You definitely put a lot of thought into this.

I know we will not see this, but it would be interesting to see how much actual income DE has generated from the Glaive and Vauban these last weeks.

I do not mean people that had platinum sitting around collecting dust and spent them on these items in marketplace from being a Master or Grandmaster, but actual NEW sales of platinum people made in order to acquire these new items.

That is what DE needs to weigh in against the implementation of a token system.

Also, as you stated, the bottom line is there still has to be some income generation for DE. Nobody expects to work for nothing and we all enjoy supporting and playing this game.

Free to play business model makes money by offering people convenience and time savings for their money. You cannot save people so much time that it removes revenue and kills the game.

Token system thoughts:

Eventually people will have a huge stockpile of tokens and be able to buy items the date they come out and nobody will need platinum anymore except for slots, so either

1 - tokens need to expire to keep inflation down. Downside, time stamping every token causes huge overhead. Also people will get into a spend them or lose them mentality.

or

2 - new items on the marketplace are available for purchase with platinum initially and become available for token purchase X weeks later.

Hypothetically using Vauban and 3 week delay...

So PlayerA wants Vauban now and decides he will pay the 300 platinum for it, while PlayerB decides he will wait the 3 weeks and farm up the 1500 tokens needed before Vauban BPs are available in the marketplace.

Marketplace can already have the BPs up with the pricing and release date the day Vauban is released so people can decide for themselves if they want to wait or not, and if they decide to wait how many tokens they will need and on what date.

Edited by Mickey1779
Link to comment
Share on other sites

Good post. Question, though. You state that the probability of getting an item increases after X amount of times, right? I've only taken basic stats, but a coin that's given you 50 heads in a row still has a 50/50 chance for heads or tails.

 

...then again, if what you're stating is the probability of getting that 50 heads in a row... then yeah, it makes sense. Getting 50 heads in a row is definitely lower probability than 49 heads in a row.

 

Yes this is a completely off topic and not even game related question but I'm curious. =P

Link to comment
Share on other sites

Good post. Question, though. You state that the probability of getting an item increases after X amount of times, right? I've only taken basic stats, but a coin that's given you 50 heads in a row still has a 50/50 chance for heads or tails.

 

...then again, if what you're stating is the probability of getting that 50 heads in a row... then yeah, it makes sense. Getting 50 heads in a row is definitely lower probability than 49 heads in a row.

 

Yes this is a completely off topic and not even game related question but I'm curious. =P

 

Sorry which part of the post are you talking about? The positive feedback loop?

Link to comment
Share on other sites

The positive feedback loop increases the probability based on the number of times you didn't get the drop, so that eventually it's impossible to not get the drop after a certain number of times of not getting it in a row. Obviously something like tossing a coin where each event is random and independent  would not change probability based on the outcomes before, but in the positive feedback loop the outcomes are not independent.

Edited by Notso
Link to comment
Share on other sites

I miss the "Cons" part of your own ideas in "03". Besides that good feedback, constructive criticism and nice analysis.

 

I try to get some cons, maybe you can tweak the suggestion to work around those cons:

 

Blueprints

- Everyone will be sure to get what they want after a maximum of runs. The gambling factor which is really fun (e.g. borderlands2 slot machine) is destroyed. You could cap the chance to get one after x-attempts to 50%. But I don't like items that are too easy to obtain. The harder to obtain something the more special feel players who did it.

 

Mods

This will cause all players to only run specific planets while other planets won't be played at all. You can already see it on Defense and Mobile Defense Mission. These are the only ones that are run to farm mods. If you now concentrate everything on the high level planets everyone will just run these. No one will ever run other planets, because you get more exp on high level planets AND now you get all mods you want from these planets. Only players who have no friends and not unlocked the high level planets will play on other planets.

 

? Alerts

Tokens give steady progress. Which is done in MMOs mostly. I think they are very bad. Again everyone gets everything with spending time. There is no lottery. A casual or new player should be able to get a super unique item while a long-time or hardcore player didn't have the luck. This is what makes games like D2 so great. If you implement tokens the long-time players always will have the advantage!

Link to comment
Share on other sites

I miss the "Cons" part of your own ideas in "03". Besides that good feedback, constructive criticism and nice analysis.

 

I was actually running out of time when I was finishing that - thanks for saving me some time though, I will edit it in when i get the chance.

Link to comment
Share on other sites

Very nice post.

To extend on what Mickey is saying, the token price could start high and then decrease over time. The first week could be platinum only, the second week would have it available at a 2x token cost. The third week would have 1.5 cost and the fourth week would be base price.

Link to comment
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
 Share

×
×
  • Create New...