Riven Slots Post-19.3.2 Feedback

[Xzorn]

54 minutes ago, Mudfam said:

Are you even serious? Are you just trolling or do you actually believe this?

Well, all I can say is have fun in make believe world. Your imagination is your only limit, I guess.

 

Instead of pretending you're right and saying I live in fantasy or I'm talking nonsense why don't you prove me wrong.

 

29 minutes ago, P0Pz said:

Without it, it is a hardcore gambling system.

 

There's a reason Casinos kick you out if they catch you card counting.

Probability is a proven statistical means to project your odds of success. Even layers of odds like the Riven system.

I'm not trying to say the odds are favorable. I'm saying your time is not fully wasted. If you know the range of chance each stat in the available pool of stats and the range each stat can yield you can very much set a desired result and formulate an estimated probability and time on that result. Each trial you make towards that projected result is progress.

RNG only exist if you attempt something once. The moment you attempt it more than once it becomes probability.

[Mudfam]

11 minutes ago, P0Pz said:

Read my post above. I think DE tells the truth... there is a lot happening just for 1 riven.... a lot...

Nope, there isn't. In computer terms it's close enough to zero that's it's not even worth counting. It doesn't matter if you multiply this number by 100 slots for 50 million accounts, it's still nothing.

Someone said that the mods are stored extremely inefficiently, as string literals. Even so, there's no way this can be a problem for modern server capacity. If it were really that bad, then they should just fix it because it's ludicrous and pointless. At most a riven mod in a binary format should be taking up 12 bytes (read: nothing).

[Arkvold]

3 hours ago, Tizodd said:

It's this thinking that will keep allowing consumers/customers/gamers to be screwed over by companies.  We pay [DE]'s bills.  We put food in their stomachs.  They owe us everything.  But as long as we feel they don't owe us anything, we'll continue being the crying little victims.  Yes they make a good product, but without us logging in, buying plat and prime access, [DE] would be bankrupt.  We can cry and complain all we want, but until we cry and complain with our wallets, [DE] can drink our tears all the way to the bank.

It's similar to gamers complaining about release day DLC, real money cash shops in-game, and always-online DRM.  We complain, but people inevitably buy the product.  Actions speak louder than words.  The way to get a company to listen, is by talking with your wallet.

TL:DR - [DE] is a company.  Companies exist to make a profit.  Companies listen to consumer complaints when it affects the bottom line.  When people stop buying Prime Access, Platinum, and logging into the game...[DE] will listen.  Until then, they will do whatever results in the most profit.

Careful, man.

I got a thread locked for saying this kind of stuff on the forums.

[Mudfam]

5 minutes ago, Xzorn said:

Instead of pretending you're right and saying I live in fantasy or I'm talking nonsense why don't you prove me wrong.

Simple facts don't need proof. This isn't theoretical, there's nothing to be proved or debated.

Probability is exactly that, probability. If you were to attempt something 50 times you would have a certain probability of success. If you actually attempt it once, you still have the same exact probability for your next 50 attempts. The same probability doesn't somehow magically apply to your next 49 attempts instead, it's still 50 just as before.

It doesn't matter how many times you've tried in the past, it doesn't matter how many times you failed. The odds remain exactly the same as before. None of this somehow counts towards improving your odds, there is no progress made, there isn't a higher power tracking your failures to finally reward you.

This should be really obvious, I don't know why I'm explaining something like this.

[P0Pz]

34 minutes ago, Mudfam said:

Nope, there isn't. In computer terms it's close enough to zero that's it's not even worth counting. It doesn't matter if you multiply this number by 100 slots for 50 million accounts, it's still nothing.

Someone said that the mods are stored extremely inefficiently, as string literals. Even so, there's no way this can be a problem for modern server capacity. If it were really that bad, then they should just fix it because it's ludicrous and pointless. At most a riven mod in a binary format should be taking up 12 bytes (read: nothing).

Might be true. But compared to other "MODs" it is with 4 RNGs running on each riven a hell of a MOD.

I dont know the specs from WF servers but as you said...it should not be a problem...

And IF, why did they pushed it out, if it is so critical ?! 😉

[sh00chu]

What would make paying for riven mod slots palatable for me would be to:

1) Make it so that the mod is attuned to the weapon you have it equipped on when you complete the challenge on the mod. 

2) Remove the ever increasing amount of kuva required to reroll stats.

Too many layers of rng are involved with these mods to make it worth the real cash required to use them. When you get 2 weapon slots for just 12p, it's hard to justify paying 20p for a single mod that may never be useful based on your luck.

[P0Pz]

1 hour ago, Xzorn said:

There's a reason Casinos kick you out if they catch you card counting.

Probability is a proven statistical means to project your odds of success. Even layers of odds like the Riven system.

I'm not trying to say the odds are favorable. I'm saying your time is not fully wasted. If you know the range of chance each stat in the available pool of stats and the range each stat can yield you can very much set a desired result and formulate an estimated probability and time on that result. Each trial you make towards that projected result is progress.

RNG only exist if you attempt something once. The moment you attempt it more than once it becomes probability.

The reason why they care is: "THEY DO NOT WANT TO REWARD YOU".

As i said, with 4 rngs, each with a own droptable, which do not add sonething to your next roll is endless gambling.

Just 4 x 4 ( 4 stats with each 4 items-stats in a list, lets ya gambling like crazy. Even if you gamble 1 positive at that very moment another stat that was decent is gone and can be negative afterwards. Unlimited possible.

Thats not rewarding. It is gambling. And gambling does not want you to be rewarded at all.

Which is in our case, InGame, a Fail-rewarding mechanic for grinding.

The progress of first grind is getting a riven

The progress of gambling on the riven stats is endless expensive scaling and not rewarding in doing your "additional 2nd grind invest to pay the next roll" in account into it. It fails to deliver the core point of this game...grinding~rewarded.

The chance of endless gambling does not allow it with 4 RNGs running positive and negativs.

Edited December 14, 2016 by P0Pz

[Xzorn]

12 minutes ago, Mudfam said:

Simple facts don't need proof. This isn't theoretical, there's nothing to be proved or debated.

Probability is exactly that, probability. If you were to attempt something 50 times you would have a certain probability of success. If you actually attempt it once, you still have the same exact probability for your next 50 attempts. The same probability doesn't somehow magically apply to your next 49 attempts instead, it's still 50 just as before.

It doesn't matter how many times you've tried in the past, it doesn't matter how many times you failed. The odds remain exactly the same as before. None of this somehow counts towards improving your odds, there is no progress made, there isn't a higher power tracking your failures to finally reward you.

This should be really obvious, I don't know why I'm explaining something like this.

 

When you project probability of 50 attempts then attempt 10 times. Do you then project your probability of the next 50 attempts or you do keep counting?

There's no point in projecting probability of anything unless you know how many attempts on avg it will take for a result. What you're pretending is that the result of your projection will remain the same regardless of attempts when that's completely contradictory to the entire purpose of calculating probability.

When you project 50 attempt then attempt once, your new projection for 50 attempts would be 51 attempts as you've already attempted once. At the same time if you project 99% chance of one or more result in 50 attempts then fail 25 times. You can consider yourself halfway done with your projected result. You could get it sooner and yes there's even a chance you don't get it but it's so incredibly unlikely it's not worth considering. The next 25 attempts will finish your projected 99% result.

Attempting something once is not probability. It's chance.

 

[Mudfam]

30 minutes ago, Xzorn said:

 yes there's even a chance you don't get it but it's so incredibly unlikely it's not worth considering.

And the first time you fail you have taken your first definitive step towards that unlikely outcome.

Quote

The next 25 attempts will finish your projected 99% result.

*sigh*. Your previous attempts are irrelevant. The moment you start actually realising them, probability no longer applies. Those are actual results, what really happened as opposed to what is probable. You can calculate a new probability for your next 25 attempts, or make 50 additional attempts. What has gone before has gone, it has no place in probability whatsoever, it's a just history which has no means of interacting with your entirely hypothetical projected probability future.

Anyway, I'm not going to reiterate this again. I've ran out of different ways of saying the same thing.

As for actual probabilities in riven mods, you may want to take a look at just how many factors are involved in rolling an "ideal" riven mod. Someone on Reddit calculated that to reach a 50% probability of rolling the stats you want (without considering their values) it would take ~180'000 cycles. Feel free to convert to lifetimes of grinding.

Good luck.

Edited December 14, 2016 by Mudfam

[bubbabenali]

4 signed shorts (64 Bit), one name composed out of 3/32 syllables(24 Bit) and a list of the corresponding effects, a polarity (8Bit) and MR value (8 Bit)
(=104 Bit - put in some overhead => 128 Bit)

times 15 for each player
(=1920 Bit)

say, Warframe has 10 million active players
(~19.2 GBit = 2.4 GB of data)

Guys, it is time to get into panic mode; DE probably has financial issues.

[Mudfam]

12 minutes ago, bubbabenali said:

4 signed shorts (64 Bit), one name composed out of 3/32 syllables(24 Bit) and a list of the corresponding effects, a polarity (8Bit) and MR value (8 Bit)
(=104 Bit - put in some overhead => 128 Bit)

times 15 for each player
(=1920 Bit)

say, Warframe has 10 million active players
(~19.2 GBit = 2.4 GB of data)

Guys, it is time to get into panic mode; DE probably has financial issues.

Yeah, or even:

4 x 24 possible modifiers (20 bits). 4x different value modifiers (say 32 bits). Name is implicit (0 bits). Polarity (2 bits). Mastery rank (3 bits).

Total 57 bits, padded to 64. Even at half we could still store everything needed.

 

DE, we can buy you a USB thumbdrive for Christmas and solve the riven capacity problem.

[S074]

If I were to save a riven on disk in the most efficient way, I'd save all its stats (minus level and number of rerolls) as a seed, the simplest seed that comes to mind for me is the time of unveiling server side.

The seed would not weight more than 64 bit being generous. And once ran in a script it that seed would always produce the same riven.

I'd calculate the riven value at login and keep it like that only in the user's ram.

So yes, a seed, the level, the rerolls. It's probably as efficient as any other mod they have in their database.

 

[Gunstray]

23 hours ago, (Xbox One)Nittymaster said:

This best thing was to award these slots during sortie missions

instead of the boat loads of Endo.

Last datamine showed they increased the reward drop of 2000 endo.

 

Can;t find it atm but Im surprised nobody called them out on it.

[(PSN)Stealth_Cobra]

19 minutes ago, Gunstray said:

Last datamine showed they increased the reward drop of 2000 endo.

 

Can;t find it atm but Im surprised nobody called them out on it.

That would explain alot. When War Within came out on PC, I got four or five riven mods in my first week. When War Within came out on PS4, I got endo , endo, endo, endo, lens, endo...

[Siekier]

2 hours ago, Arkvold said:

Careful, man.

I got a thread locked for saying this kind of stuff on the forums.

Locked because of "encouraging other players to boycott the game"? Dafuq?

 

So telling people to simply vote with their wallets is "encouraging other players to boycott the game"?

Wow.

This is some serious bs.

[LSG501]

21 hours ago, xRufus7x said:

You are assuming weapon tiering isn't intentional, which it obviously is.

Still doesn't mean that all weapons on the same tier are equal, let alone scaling as mods are applied or the player/enemy increases in levels....  The simple fact is that there are far too many weapons that are basically MR fodder and riven mods are supposed to be the system to fix this, yet we're now going to be charged to fix the issue that shouldn't be there in the first place, the odds are that the players riven mods are aimed at has more than 15 weapons that could take riven mods....  

ALL weapons should be able to kill high level enemies with enough investment in forma, that isn't the case at the moment because of the way that enemy and weapon scaling currently works. 

[xRufus7x]

1 hour ago, LSG501 said:

Still doesn't mean that all weapons on the same tier are equal, let alone scaling as mods are applied or the player/enemy increases in levels....  The simple fact is that there are far too many weapons that are basically MR fodder and riven mods are supposed to be the system to fix this, yet we're now going to be charged to fix the issue that shouldn't be there in the first place, the odds are that the players riven mods are aimed at has more than 15 weapons that could take riven mods....  

ALL weapons should be able to kill high level enemies with enough investment in forma, that isn't the case at the moment because of the way that enemy and weapon scaling currently works. 

Depends on how you define high level. That is the tricky part of Warframe. The thing is with an appropriate investment in Forma and a proper load out any weapon is already sortie viable with the exception of the third stage with specific resistances or eximus.

[(XBOX)Squidzbusterson]

You know I'm cool with a cap of 60 I doubt I use even thirty weapons often and even then I only really want Rivens for my more favorite weapons that don't quite cut it at top tier, but sixty plat is pretty steep Im just a console player so I haven't had the system as long but if you made it like 10 plat or somethingIld buy the crap out of em

[GoneBlank]

@Mudfam Your efforts in explaining Gambler's fallacy were admirable.

If there are people reading this who truly believe that the quantity of their previous fails will improve their odds of a success in the the future (when conducting independent draws) then I would strongly encourage you to read https://en.wikipedia.org/wiki/Gambler's_fallacy and associated references. We also went over this topic extensively in a previous post called "The Math of Farming".

[Xzorn]

 

6 minutes ago, GoneBlank said:

@Mudfam Your efforts in explaining Gambler's fallacy were admirable.

If there are people reading this who truly believe that the quantity of their previous fails will improve their odds of a success in the the future (when conducting independent draws) then I would strongly encourage you to read https://en.wikipedia.org/wiki/Gambler's_fallacy and associated references. We also went over this topic extensively in a previous post called "The Math of Farming".

 

It's not Gambler Fallacy. I'm not trying to say the chance is better on the next attempt because you failed the first X amount.

I'm saying that when you have a the probability of a result in X attempts, then completing that trial can be seen as progression and therefor it is highly unlikely you will be wasting your time by going through those attempts. Repeated attempts will yield a better chance of a desired result overall, not on the next attempt. That's the difference.

 

 

 

[GoneBlank]

@Xzorn Your use of the word "progression" is.......creative.

Most people would generally interpret that word is saying that their previous efforts would in some sense improve the chance of success on the next draw. For example, imagine if with Rivens that there were a finite number of combinations and that once you had unveiled a particular combination then that combination could not be experienced again; i.e., that combination would be removed from the feasible set (drawing without replacement). If a particularly bad combination was experienced and could therefore not be experienced again then I think most people would call that progression. However, this feature does not exist and you see many people re-roll rivens and experience similarly bad combinations roll after roll. Hence, the drop tables appear to be fixed. So previous fails do not improve your chances of good roll. Every riven roll is completely independent of history.

I agree with you that the primary method of generating a good roll is to roll a LOT. In the same way that buying 1000000 lotto tickets will increase your chance of winning the lotto relative to buying just one. However, having 999,999 not win the lotto has absolutely no influence on the chance of the millionth being the winner. Similarly, if I was checking my 1,2,...,999999 failing lotto tickets I would not consider myself to be progressing.

That being said....we all look at life differently.

[FlinttheImmortal]

6 hours ago, Xzorn said:

When you project probability of 50 attempts then attempt 10 times. Do you then project your probability of the next 50 attempts or you do keep counting?

1 hour ago, Xzorn said:

Repeated attempts will yield a better chance of a desired result overall, not on the next attempt. That's the difference.

Seems to be a bit of confusion here.  Specifically, the difference between statistics and probabilities.  They are not the same.

Take dice for example.  Clean dice - not loaded ones, obviously.  Also assume using a single die, for the sake of simplicity.  For each toss, the probability that you will roll a one (your desired result) is exactly one in six.  Roll that die ten times, 1000 times, a million times, your chance to roll a one is still exactly one in six.  Your previous rolls do not in any way weight the chances of rolling a one on the next roll.  Statistics, on the other hand, merely tallies the outcomes of all the individual rolls.  Over the short run, the statistics may show that a roll of two has been prevalent.  Over a larger data set, the statistics will show that each roll outcome occurs one sixth of the time.  Same goes for a coin toss, or a royal flush in a hand of poker.

Rivens are a bit like dice.  Apart from locking in the weapon on the first "roll," each subsequent roll is its own distinct event and not dependent on the previous one(s) in any way whatsoever.  You are no more likely to get a "good" Riven on the tenth Cycle than you are on the first.  And a player can no more "project" the likelihood of what their next Cycle will look like anymore than they can predict what the next toss of a die will be.  All that can be said is that on the next toss, the player has a one in six chance of rolling a (insert number).  That's it.  Why?  Because those are the initial conditions and they don't change from toss to toss.  And all that a player can say about their next Riven cycle is that the outcome will be entirely randomly determined.  Why?  Because those are the initial conditions and they don't change from Cycle to Cycle.

Unless, of course, DE has built a weighting mechanic into Rivens.  Like the 2k endo thing, for example.  If 2k endo drops for a player 3 times in a row, the forth will exclude that drop from the list of possibilities.  Near as I can tell though, no such mechanic exists for Riven.

[P0Pz]

6 hours ago, FlinttheImmortal said:

Seems to be a bit of confusion here.  Specifically, the difference between statistics and probabilities.  They are not the same.

Take dice for example.  Clean dice - not loaded ones, obviously.  Also assume using a single die, for the sake of simplicity.  For each toss, the probability that you will roll a one (your desired result) is exactly one in six.  Roll that die ten times, 1000 times, a million times, your chance to roll a one is still exactly one in six.  Your previous rolls do not in any way weight the chances of rolling a one on the next roll.  Statistics, on the other hand, merely tallies the outcomes of all the individual rolls.  Over the short run, the statistics may show that a roll of two has been prevalent.  Over a larger data set, the statistics will show that each roll outcome occurs one sixth of the time.  Same goes for a coin toss, or a royal flush in a hand of poker.

Rivens are a bit like dice.  Apart from locking in the weapon on the first "roll," each subsequent roll is its own distinct event and not dependent on the previous one(s) in any way whatsoever.  You are no more likely to get a "good" Riven on the tenth Cycle than you are on the first.  And a player can no more "project" the likelihood of what their next Cycle will look like anymore than they can predict what the next toss of a die will be.  All that can be said is that on the next toss, the player has a one in six chance of rolling a (insert number).  That's it.  Why?  Because those are the initial conditions and they don't change from toss to toss.  And all that a player can say about their next Riven cycle is that the outcome will be entirely randomly determined.  Why?  Because those are the initial conditions and they don't change from Cycle to Cycle.

Unless, of course, DE has built a weighting mechanic into Rivens.  Like the 2k endo thing, for example.  If 2k endo drops for a player 3 times in a row, the forth will exclude that drop from the list of possibilities.  Near as I can tell though, no such mechanic exists for Riven.

Agree...

Well,...

In our case with riven...we have 4 dices !

With more then just 6 sides!

Now, a loot table (droptable) can contains not only 1 "item-side"  on our dices. Lets say we got 10 "item-stats" but our loot table for 1 dice contains 100 item-stats. What?  A stat can apear 10 times in the table, another 5 times and anotherone just 1 time. This is the RNG of 1 item. In this case item-stat.

So we have 4 dices with a lot of sides, and some sides-stats apears more often then others on the dices sides.

Now, THIS allows you to throw your 4 dices endless with the possibility to never hit your reward/goal(s).

1 Dice which symbols 1 RNG, is what we have regular in grind games. Also here in Warframe.

With 1 RNG you CAN make progress. With 3 RNGs within 1 item you have the possibility to never hit the rewards. We have 4. Thats the fail point 1.

Another fail...riven got no BeP to reward your investment. Which would be like:

After your 5th roll, the riven stat nr1, stays positive +% no matter how often the player keeps on to roll. This excamble symbols the "break-point". You had 5 times the chance to get already better stats but it fails. Your investment of your 2nd grind for your riven (scaling roll-cost) is rewarded to get at least at your 5th roll your 1st stat locked to only positive sides. Ect. This would reward the player within thus endless gambling and "breaks the endless possibility".

Edited December 15, 2016 by P0Pz

[Sziklamester]

On 2016. 12. 13. at 9:54 PM, Sophista said:

Because its not pay to win. Its pay to have BIGGER selection and some Cosmetic Glitter.
more slots will not make your stuff perform better.
Player with 15 slots have the same chance of pulling great riven as player with 60.
Slots also dont make you win faster. Boosters kinda do, but no one complains about them, ever.

Saying all that, 60p for 3 slots is about 3 times too much. DE knows we collect. 

It is stll in terms of pay to win because with this you win something which have value.

[Ciaus]

20 hours ago, Xzorn said:

One failed roll simply increases the probability of your next being a success

That's not how independent probability works. 
Say I want to roll a mod with top 1/10 quality. 

From any starting point, the more rolls, the probability of getting what I want converges to 1. But on any single roll, assuming the rolls are independent, which they should be in any purely RNG system, the probability of getting what I want is the same. Having done 100 rolls or 1, the probability of getting what I want next roll (roll 101 or roll 2) is THE SAME. So you can easily get stuck in a seemingly endless loop of bad luck just by chance. There's no light at the end of the tunnel that helps to pull you out of that.