Predetermined Pulls

pheregas

I will preface this by playing devil's advocate and stating that the reason why Predetermined Pulls exists is to most likely keep people from cheating and artificially inflating their 5* pull rate.  I understand the devs need to keep cheaters in check.

What I will say is that at as a long time vet (day 1876), I get really disgruntled that my pull rate is below the advertised 15%.

Being a vet who is post-iso, the only real progress in the game for me is gaining 5* champs and levels.  This is a slow process, but that is understandable and to be expected and I think I've put in the appropriate effort to get here.

Back before I was post-iso, I used to track every single one of my LT draws (including which store they came from).  I just double checked my google doc and the last time I updated it was 12/2016.  At that point I had logged around 400 LT pulls and my 5* draw rate was 13%.  Not great, sure, by probably within statistical variation.  (and all other stores were also less than 15%).  And probably enough pulls to get a decent sample size.

It did not feel good though to know that the in-game code selected me to be the lower draw person and randomly selected someone else to be the 17% puller.  So I stopped logging.  I'm reasonably sure I didn't start pulling higher than 12-13% because my 5s started getting harder and harder to cover with me barely champing them before they would be banished to classic tokens.  Is this bad odds or lack of feeders, who's to know?

That was, until I started hoarding because by some miracle, I got 20 Kingpin covers before Cable even left tokens.  Of course, I pulled only Kingpin and nobody else, so don't start thinking I got above a 15% pull rate.  So I decided to join the hoarding club and wait for KP to be banished.

On April 16th, I dropped my hoard.  It wasn't the greatest of hoards, but in the less than a month since I started hoarding, I had almost 1,400 CP and 22 LTs.  I've earned more CP and progression rewards since then, so as of today, April 24th, I've pulled and catalogued 118 pulls.

My pull rate for 5*s is 9.3%.  In that time I've gotten 4 BH 4s and ZERO 5*BH.  (I've  only ever gotten four 5*BH since BH got introduced.)  I'm also on a 25 pull run of no 5s.  That does not help my feelings of discontentment and lack of progress.

I've seen mentioned several times a streak breaker.  I can't get behind this enough.  But if not that, at least something has to done.  I'm really not sure what I did to make the game select me as the poor slob who is on the lower end of the average, but as I said before, in a game-state where the only progress is the one thing I am being denied (RNG hates me), it makes for a less than fun experience.

TLDR: My draw rates are horrible.  I can only get so much satisfaction out of rotating 2/3* (and entering 4*) farms.  Why has RNG forsaken me?

Apologies if this came out as whining or a rant.  I just want to raise the subject that as a vet, I am getting less satisfaction out of the game due to lack of progress.  Taking a break does not fix this issue since taking a break lowers my CP income (and even after 1876 days, I'm not burned out yet.)  And this is an issue that can be addressed.  Feeders for all 5s, streak breakers, paying CP/HP for an LT reroll, increasing 5*BH odds.  

Anyway, anyone that would like to comment, please feel free.  Try and keep the roasting to a minimum please.

wymtime

Pull rates can be really frustrating especially when you see a 5* you really want.  One thing I would ask is how many pulls did it take you to get 20 king pin covers.  If you add in your 5* draw rate from when KP entered packs to what you have gotten today how would that adjust your 5* draw rate?

i am not sure how the predetermined 5* draw rate works if it is when you make pull number X it will be a 5* or will it be you have a 15% chance for a 5* and if you get a 5* it will be X character.

Bowgentle

wymtime said:

Pull rates can be really frustrating especially when you see a 5* you really want.  One thing I would ask is how many pulls did it take you to get 20 king pin covers.  If you add in your 5* draw rate from when KP entered packs to what you have gotten today how would that adjust your 5* draw rate?

i am not sure how the predetermined 5* draw rate works if it is when you make pull number X it will be a 5* or will it be you have a 15% chance for a 5* and if you get a 5* it will be X character.

Pull X will be a specific 5.
That 5 will change when new 5s enter tokens.
X will stay the same.

Borstock

I mean.... you state that when you were recording every one of your pulls, you were only off by about 2%, and you admit that's within reason. It doesn't seem like you have an issue there.

So, it feels like your main issue is with the 9.3% pull rate for 78 pulls. And that feels like an issue that could easily correct itself within a handful of pulls.

pheregas

Bowgentle said:

wymtime said:

Pull rates can be really frustrating especially when you see a 5* you really want.  One thing I would ask is how many pulls did it take you to get 20 king pin covers.  If you add in your 5* draw rate from when KP entered packs to what you have gotten today how would that adjust your 5* draw rate?

i am not sure how the predetermined 5* draw rate works if it is when you make pull number X it will be a 5* or will it be you have a 15% chance for a 5* and if you get a 5* it will be X character.

Pull X will be a specific 5.
That 5 will change when new 5s enter tokens.
X will stay the same.

And I  don't have a second device for previewing the new release stores, so I stay away from them completely anymore and wait for the new release to enter Latest.

bluewolf

Should we form a 13% support group?

Before I broke my (relatively small 200 pull) hoard after Kingpin left, I had made 2720 pulls and gotten 359 5's, or 13.2%, lifetime.  My 200 pulls yielded around 24 5's, so it was about a 12% rate there.  Obviously that didn't increase the % but I haven't formally calculated it out.

I guess the answer (if they were to give an official one) is that your string of results, hard coded into your store you pull from, would eventually, after nearly infinite pulls, get you a pretty close to 15%.  (Anyone have a million dollars they don't need that I can use to test this?)

And maybe if you play another 30 years you'd see your overall rate creep up to 14.8% or something.  It takes a LOT of pulls to statistically create enough to show you aren't where you are supposed to be and/or prove that you actually will be there.

Meanwhile Mr 17% might see his rate drop down in those same 30 years and be upset because he's not getting 5's left and right anymore.

Someone in my ally said he got 4 BH 5's out of his 200 hoard break last week, where I got one.

I mean, it sucks.  We don't like it?  blah blah blah no guarantees random blah blah blah...

Sigh.

If it makes you feel any better, having a 450 or a 460 champ doesn't really matter all that much in the long run, especially for anyone who's not the meta.

My only suggestion is stop pulling garbage 5's...hoarding is boring but better than having 20 Kingpin covers.  I stopped mine at 453.

pheregas

Borstock said:

I mean.... you state that when you were recording every one of your pulls, you were only off by about 2%, and you admit that's within reason. It doesn't seem like you have an issue there.

So, it feels like your main issue is with the 9.3% pull rate for 78 pulls. And that feels like an issue that could easily correct itself within a handful of pulls.

I get what you mean, but then again, what is reasonable?  Is it within d3's error rate so reasonable to them?  Should my pull rate have evened out over time (though it hasn't appeared to)?   I said reasonable  because, while I was certainly not happy about it, I didn't want the STATS patrol to jump out and snipe me.  400 seemed like a statistically significant N.  I didn't run a student's T-test on it or anything.

If I had gotten the advertised 15%, that would have been about another 8 5*s over that period.  Those rewards add up and as 4* power is completely irrelevant to me while, those 8 4s I got aren't worth as much with their progression rewards.  That would have also been 8 more chances at a BH.

pheregas

wymtime said:

Pull rates can be really frustrating especially when you see a 5* you really want.  One thing I would ask is how many pulls did it take you to get 20 king pin covers.  If you add in your 5* draw rate from when KP entered packs to what you have gotten today how would that adjust your 5* draw rate?

I'm honestly not sure.  I pulled as many CPs and LTs as I earned over that time frame.  I got exactly 17 cable covers and Doom had 4 while KP had the 20.  I average about 16 covers per character this way since most of my 5s are between level 452 and 454 before they leave for classics.

I'm sure someone has already run the numbers on average CP/LT earn rates running CL8 PVE, and CL9 PVP to 900.

Kojubat

In an environment where the pulls are static (you cannot 'reroll'), it seems reasonable that the population algorithm could be designed in such a way to guarantee a 15% pull rate over X number of draws. Perhaps it is even designed that way for some large value of X, but let us assume not.

So, that suggests it is designed to play into mobile game monetization strategies (gambling). I see two psychological aspects at play here:

1. Sunk cost fallacy: if you are below the advertised rate, the thought is that you have already spent so much, the better results must be just about to happen. 
2. Lucky streak: if you are above the advertised rate, the thought is that you must be on a run of positive results, so why stop now?

Both theories encourage spending (not just of currency, but of time, effort, etc.), and are not necessarily evil or unique. I am not suggesting anyone has fallen prey to either, just that it could be a reason to have the stores work this way.

The other thing to consider is that every store is a different set of results. Doing poorly in a release store does not make it more likely you will do better in a special event store, and the inverse is also true. In a way, it is sort of like playing lottery scratch-off games - if all of the desirable prizes are still available, your odds are slightly better playing the same game more than once instead of playing an assortment of different ones.

HoundofShadow

As for odds, mathematically speaking, the sample rate has to be pretty high (10,000?) pulls in order to be closer or equal to the odds.

I remember reading an unplanned mini-discussion between the devs and a player regarding streak-breaker. It was mentioned that if streak breaker were introduced, the odds of drawing 5* has to be increased in order to even out the draw rates (?).  So, players who enjoyed good pulls will dislike this change and players like you might welcome this change. Streakbreak is a double edged sword.

I don't see how feeders could be a solution since it gives only 6 covers. It might help to reduce the number of pulls needed to champ a 5*. On the other hand, you have to focus your 4* BH. But RNG comes in again. I'm not sure how many years it will take to start double feeder. Probably the next one or two years?

Another alternative would be to decrease the number of days to get 5* cover as login rewards. Currently, it's 1 every 90 days. 

pheregas

My only suggestion is stop pulling garbage 5's...hoarding is boring but better than having 20 Kingpin covers.  I stopped mine at 453.

Maybe it is a failing of mine, but I want every 5* champed.  I don't care if they get up in the high 460+ range.  Just champing is enough.  And Cable was lacking covers.  It was the dreaded 2/5/5.  I suppose I could have just turned in my 3 saved covers and then champed.  But using saved covers in that manner feels like such a WASTE, especially when you're in the 13% club.  Maybe I should get over that.  

So I kept pulling from that group of Latest hoping to get that last Cable cover, but I kept pulling KP.  

pheregas

Kojubat said:

In an environment where the pulls are static (you cannot 'reroll'), it seems reasonable that the population algorithm could be designed in such a way to guarantee a 15% pull rate over X number of draws. Perhaps it is even designed that way for some large value of X, but let us assume not.

So, that suggests it is designed to play into mobile game monetization strategies (gambling). I see two psychological aspects at play here:

1. Sunk cost fallacy: if you are below the advertised rate, the thought is that you have already spent so much, the better results must be just about to happen. 
2. Lucky streak: if you are above the advertised rate, the thought is that you must be on a run of positive results, so why stop now?

Both theories encourage spending (not just of currency, but of time, effort, etc.), and are not necessarily evil or unique. I am not suggesting anyone has fallen prey to either, just that it could be a reason to have the stores work this way.

The other thing to consider is that every store is a different set of results. Doing poorly in a release store does not make it more likely you will do better in a special event store, and the inverse is also true. In a way, it is sort of like playing lottery scratch-off games - if all of the desirable prizes are still available, your odds are slightly better playing the same game more than once instead of playing an assortment of different ones.

I like your synopsis.  And knowing that each store has it's own pull rate is why I never put CP into new release stores.  Unless the regular Latest store resets odds every time the 5* contents change.

pheregas

I don't see how feeders could be a solution since it gives only 6 covers. It might help to reduce the number of pulls needed to champ a 5*. On the other hand, you have to focus your 4* BH. But RNG comes in again. I'm not sure how many years it will take to start double feeder. Probably the next one or two years?

My theory on why feeders would be a great benefit is that you have the ability to potentially get covers before a character leaves for classics.  For instance, Danvers is set to get a feeder on Friday.  In my case, I get 6 Danvers covers.  Suddenly my 1/2/3 Danvers is a 3/4/5 Danvers and I only need one more non maxed color to champ.  This means I can stop pulling tokens from that LT-group until Danvers leaves, so my CP per needed 5* value goes up.  The feeder eliminates up to 150 cp worth of pull requirements to Max-Cover. (Edit: The CP needs are probably much greater since those 6 required covers won't take 6 pulls to get but probably 6/.15 = 40 pulls, so 1,000CP if I did that math correctly.)

bluewolf

pheregas said:

Kojubat said:

In an environment where the pulls are static (you cannot 'reroll'), it seems reasonable that the population algorithm could be designed in such a way to guarantee a 15% pull rate over X number of draws. Perhaps it is even designed that way for some large value of X, but let us assume not.

So, that suggests it is designed to play into mobile game monetization strategies (gambling). I see two psychological aspects at play here:

1. Sunk cost fallacy: if you are below the advertised rate, the thought is that you have already spent so much, the better results must be just about to happen. 
2. Lucky streak: if you are above the advertised rate, the thought is that you must be on a run of positive results, so why stop now?

Both theories encourage spending (not just of currency, but of time, effort, etc.), and are not necessarily evil or unique. I am not suggesting anyone has fallen prey to either, just that it could be a reason to have the stores work this way.

The other thing to consider is that every store is a different set of results. Doing poorly in a release store does not make it more likely you will do better in a special event store, and the inverse is also true. In a way, it is sort of like playing lottery scratch-off games - if all of the desirable prizes are still available, your odds are slightly better playing the same game more than once instead of playing an assortment of different ones.

I like your synopsis.  And knowing that each store has it's own pull rate is why I never put CP into new release stores.  Unless the regular Latest store resets odds every time the 5* contents change.

The Latest store odds never reset.  What happens is the covers that you would pull, change. So, when Hela entered, you couldn't pull any more Kingpins, but could now pull Hela.  That didn't change at all when your next 5* pull was going to happen.  (maybe it was 7 pulls in, that stayed the same, just (maybe) changed who you would pull at that point)

Bowgentle

OJSP said:

What I'm interested to know is how far is it predetermined?

I've had someone pull 150+ on some miracle device, so... pretty deeply.

Vhailorx

OJSP said:

What I'm interested to know is how far is it predetermined?

Isn't it predetermined out to infinity?  I don't mean that Demi/d3 are storing an infinitely long list of pulls for each player/store.  Just that each store has a fixed algorithm for pulls and each player has a unique key.  There is no randomizing factor like a date value, so we can all pull against that store-specific alogorithm forever and because player-unique key is the starting/seed value, we will always get the same results in the same order every time.

pheregas

Vhailorx said:

OJSP said:

What I'm interested to know is how far is it predetermined?

Isn't it predetermined out to infinity?  I don't mean that Demi/d3 are storing an infinitely long list of pulls for each player/store.  Just that each store has a fixed algorithm for pulls and each player has a unique key.  There is no randomizing factor like a date value, so we can all pull against that store-specific alogorithm forever and because player-unique key is the starting/seed value, we will always get the same results in the same order every time.

If this is true, where does the sequence repeat?  Has anyone actually tracked all their pulls and have something that they could plug into a pattern discoverer?  Is that even a thing?  #notaprogrammer

Borstock

pheregas said:

Borstock said:

I mean.... you state that when you were recording every one of your pulls, you were only off by about 2%, and you admit that's within reason. It doesn't seem like you have an issue there.

So, it feels like your main issue is with the 9.3% pull rate for 78 pulls. And that feels like an issue that could easily correct itself within a handful of pulls.

I get what you mean, but then again, what is reasonable?  Is it within d3's error rate so reasonable to them?  Should my pull rate have evened out over time (though it hasn't appeared to)?   I said reasonable  because, while I was certainly not happy about it, I didn't want the STATS patrol to jump out and snipe me.  400 seemed like a statistically significant N.  I didn't run a student's T-test on it or anything.

If I had gotten the advertised 15%, that would have been about another 8 5*s over that period.  Those rewards add up and as 4* power is completely irrelevant to me while, those 8 4s I got aren't worth as much with their progression rewards.  That would have also been 8 more chances at a BH.

I don't know what a Statistics major would say, but I'd imagine you'd have to go to at least a few thousand pulls before you (general "you", not you "you") could reasonably point to a low pull rate and say with statistical certainty that you're getting an unfair pull-rate.

moss04

pheregas said:

Vhailorx said:

OJSP said:

What I'm interested to know is how far is it predetermined?

Isn't it predetermined out to infinity?  I don't mean that Demi/d3 are storing an infinitely long list of pulls for each player/store.  Just that each store has a fixed algorithm for pulls and each player has a unique key.  There is no randomizing factor like a date value, so we can all pull against that store-specific alogorithm forever and because player-unique key is the starting/seed value, we will always get the same results in the same order every time.

If this is true, where does the sequence repeat?  Has anyone actually tracked all their pulls and have something that they could plug into a pattern discoverer?  Is that even a thing?  #notaprogrammer

The standard for stuff like this is for the sequence to be so large that even if you pulled a cover every second non stop for 100 years it still wouldn't repeat

Colognoisseur

I’ve tracked every pull I’ve done from the beginning of legendary tokens. Over the long run things even out. As of this post my overall rate is 15.4%. It is going down a bit because over the last 200 pulls I’m at 9%. What has generally happened when I look back at my history these bad streaks are followed by a short streak where I pull 18-19%. I am a single data point but the percentages even out for me.

Sm0keyJ0e

I think the focus should be on the negative aspect of a bad streak, not the actual math behind it. How a streak-breaker is introduced or implemented I'll leave up to the wizards and magicians, but it really needs to be implemented.

The devs have showed empathy for things that make the player feel bad or cheated a la the cascade limiter they implemented for PVE. I can only assume people complained that the "game cheated" and walked away from it, especially early on. There is no doubt in my mind that as resources for pulling from the various legendary stores have increased, players are leaving because of that, too. Actually it's a fact.