CC, please impliment a mercy rule when opening boxes.

Discuss anything related to Shop Heroes gameplay.
Niexist
Posts: 10
Joined: Fri May 05, 2017 11:57 pm

CC, please impliment a mercy rule when opening boxes.

Postby Niexist » Sat May 06, 2017 7:51 am

I've opened 30 magic, and 60 gold chests since I last got a blueprint from one of those chests. This is why players quit your game guys. Give me a mercy roll every 10 failures to a draw a blueprint and an automatic blueprint and I would not be so incredibly frustrated with this game.

User avatar
Lionstar
"Great" Mastery Member
Posts: 194
Joined: Wed Feb 03, 2016 10:53 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby Lionstar » Sat May 06, 2017 8:40 am

We've already asked for this numerous times, maybe with the new CM there will be some response.

I had some really long streaks with no BP as well but 30 magic chests seems like little too far stretched, just saying. = )

User avatar
Shiroe
"Legendary" Mastery Member
Posts: 2330
Joined: Sun Aug 23, 2015 11:05 am
Location: the Netherlands
Contact:

Re: CC, please impliment a mercy rule when opening boxes.

Postby Shiroe » Sat May 06, 2017 9:05 am

Would need to be implemented per chest type then; can't have people opening wooden chests until the "mercy" gets activated, then open a primal chest for that one...
as of 2016-09-11: Player level: 44, City: Eolythes, Blueprints: 517, Mastered: 419, Crafted: 78.61K
(except for tier 1 and some tier 2 artifacts mostly running my shop/gearing self sufficient)

User avatar
Malric
Posts: 45
Joined: Mon Oct 10, 2016 5:16 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby Malric » Sat May 06, 2017 1:25 pm

Lionstar wrote:We've already asked for this numerous times, maybe with the new CM there will be some response.

I had some really long streaks with no BP as well but 30 magic chests seems like little too far stretched, just saying. = )


My two worst streaks were with gold and iron. I've had 100 key requests on both and towards the end I've gotten 2 from gold and 3 from iron.

Basically that's 5 BP from 200 chests, was pretty pissed, if I'd paid gems for those chests I'd probably have rage quit but amazingly when I open chests with gems I never have that bad of a streak.

I wholeheartedly agree there should be a mercy rule, per chest type to avoid abuse. Especially on higher chest types like Primal that can only be opened with gems at 750 a pop.
As of 06/01/2017 Level 58, 344.56G Net Worth, 633 Blueprints

VeNoM0619
"Great" Mastery Member
Posts: 138
Joined: Wed Apr 19, 2017 2:07 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby VeNoM0619 » Sat May 06, 2017 3:10 pm

Might want to post this suggestion in the sticky request for suggestions.

Also, as stated in the blueprint drop rate thread, RNG isn't really random. Just figure out where the current state is, and run from there.

Here's the tips I can share:
  • Never open less than 10 chests at a time (20 is better if you choose to ignore other rules), I've never opened more than 15 in a row with nothing. Given the odds are about 5-20%... you should never plan on going "empty handed". Rolling one at a time isn't great either. You don't go to casinos and play a game once and give up, expecting to win it big.
  • Always stop opening after getting a bp (take a 2-3 minute break if you insist on streaks). 3 Separate occasions I've gotten 3 in a row back to back taking breaks in between.
  • Open bags to check the server's current RNG internal state (I open purple bags) About 5-10 is good enough to tell. If you get a bunch of blues/artifacts, don't open chests. If you get flawless, sure go for it. Epics/Legendaries, you better open some chests.
  • If you are opening a lot of chests back to back, and it takes you well over 7-8 chests to get a bp (talking magic+), then give it an hour, and repeat. It does appear the ebb and flows are about 1-2 hours apart...Expect your luck to turn given that timeframe or less


Feel free to call it superstition and dismiss this completely... I don't care, but good "luck".


Given that adding a "mercy" rule actually increases the odds on bps, they might tweak the odds lower sadly. I've always felt that "loot pools" are the way to go when it comes to important "rng expectations", however, I've never seen an implementation in any game (why write a module, when you can just call MT() every time eh?) but omg would it solve so many issues with RNG expectations.

Wraith
Posts: 29
Joined: Sat Feb 11, 2017 6:22 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby Wraith » Sat May 06, 2017 6:42 pm

VeNoM0619 wrote:Might want to post this suggestion in the sticky request for suggestions.

Also, as stated in the blueprint drop rate thread, RNG isn't really random. Just figure out where the current state is, and run from there.

Here's the tips I can share:
  • Never open less than 10 chests at a time (20 is better if you choose to ignore other rules), I've never opened more than 15 in a row with nothing. Given the odds are about 5-20%... you should never plan on going "empty handed". Rolling one at a time isn't great either. You don't go to casinos and play a game once and give up, expecting to win it big.
  • Always stop opening after getting a bp (take a 2-3 minute break if you insist on streaks). 3 Separate occasions I've gotten 3 in a row back to back taking breaks in between.
  • Open bags to check the server's current RNG internal state (I open purple bags) About 5-10 is good enough to tell. If you get a bunch of blues/artifacts, don't open chests. If you get flawless, sure go for it. Epics/Legendaries, you better open some chests.
  • If you are opening a lot of chests back to back, and it takes you well over 7-8 chests to get a bp (talking magic+), then give it an hour, and repeat. It does appear the ebb and flows are about 1-2 hours apart...Expect your luck to turn given that timeframe or less


Feel free to call it superstition and dismiss this completely... I don't care, but good "luck".


This is definitely superstitious belief, no offense intended but I want to debunk these myths right away before people start to put so much effort into opening chests and then become even more frustrated when they don't get the expected results.

First, unless you know the RNG algorithm, there is nothing suggesting that some rolls (i.e. one roll = opening one chest) have bigger odds than others to yield a BP. As far as we know the odds are the same every time. If there were brackets of times (say, periods of 5 minutes) when the server is more likely to drop BPs, or if the rolls were predictable in any other way, then that would mean the RNG algorithm is very badly designed. I don't know what algorithm Shop Heroes uses, but I doubt it's an algorithm that Cloudcade wrote themselves because software devs rarely write their own. Without getting too technical, developers usually use a default RNG algorithm that come with the coding library they use, and these algorithms are carefully designed by experienced professionals to give every number an equal chance of dropping and to ensure the results can't be predicted by humans.

Second, some of your tips contradict each other. For example, you say there are times when the server is more likely to drop BPs and at the same time you advise people to stop opening chests after getting a BP. But if there really were better periods of time than others to open chests, then logic would say that if you win a BP, it's more likely the server is on a good time and you should keep opening chests.

Third, there's a lot of psychology research on superstition (or more precisely: there's research on why people are superstitious):

Avoiding streaks as you advise is called the gambler's fallacy: it's the mistaken belief that you can't get the same result multiple times in a row. An example is roulette players in a casino who see the ball land 10 times on black in a row, and who then assume the the next time it MUST land on red so they go all in. Nope, the odds of red and black (or in our case: the odds of getting another BP) are still the same.

There's also something called selection bias: that's when you only pay attention to the things that prove your assumptions, and you tend to ignore or forget the things that contradict them. For example, you say it's better to open 20 chests in a row than separately. That's probably because when you open 20 chests separately (let's say 1 chest per day), you pay more attention to every failure ("Oh no, failed again! It's been 2 weeks since I got a BP!"). On the other hand, when you open 20 chests in a row, you pay less attention to every failure because you just keep opening another chest every time; not only that but if you get a BP while opening 20 chests in a row, you'll focus on the fact that you won a BP and you will pay even less attention to all the failures.

Research also shows that it doesn't take many coincidences for humans to become superstitious. For instance if you only open chests after getting good loot from bags, and if this strategy happens to give you something good 3 times in a week, that's likely all it takes for most people to believe the strategy works and it isn't just a coincidence. In fact that's often how most superstitious beliefs begin.

I'm not saying all of this to stop you from believing what you want to believe about RNG rolls, if you are really convinced that you are right about this then good for you, but I would strongly advise other people to take these claims with a grain of salt. There's no evidence that the RNG algorithm is humanly predictable, and all the anecdotal claims that it is are easily explained away by actual research on superstitious belief and various cognitive fallacies that humans are naturally prone to. Personally I wouldn't bother putting any effort into trying to game the system, it seems like a wasted effort.

With that said, if you really wanted to prove your ideas, you could open a few hundred chests while following your tips, and then open an identical number of chests at random times. It would take time, but it's the only way to prove you're correct about what you say.

VeNoM0619
"Great" Mastery Member
Posts: 138
Joined: Wed Apr 19, 2017 2:07 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby VeNoM0619 » Sat May 06, 2017 7:48 pm

Wraith wrote:First, unless you know the RNG algorithm, there is nothing suggesting that some rolls (i.e. one roll = opening one chest) have bigger odds than others to yield a BP. As far as we know the odds are the same every time.

MT https://en.wikipedia.org/wiki/Mersenne_Twister

Wraith wrote:If there were brackets of times (say, periods of 5 minutes) when the server is more likely to drop BPs, or if the rolls were predictable in any other way, then that would mean the RNG algorithm is very badly designed.

Period lengths of MT are known to have dead streaks, and get caught in them.


Wraith wrote:I don't know what algorithm Shop Heroes uses, but I doubt it's an algorithm that Cloudcade wrote themselves because software devs rarely write their own. Without getting too technical, developers usually use a default RNG algorithm that come with the coding library they use, and these algorithms are carefully designed by experienced professionals to give every number an equal chance of dropping and to ensure the results can't be predicted by humans.

MT, used by Unity (black box beyond who's implementation they used though), which is usually the same standard algo thrown around. There have been many variant improvements, However, they are variants, and never merged.


Wraith wrote:Second, some of your tips contradict each other. For example, you say there are times when the server is more likely to drop BPs and at the same time you advise people to stop opening chests after getting a BP. But if there really were better periods of time than others to open chests, then logic would say that if you win a BP, it's more likely the server is on a good time and you should keep opening chests.

True, some could be superstition there, this is just my personal experience getting 3 in a row, 3 separate times. It may not even matter here, my best theory is that in blackjack, if you have 52 cards, you pull a King, what are the odds you pull another king on that next card? When you wait until the deck is reshuffled (other players rolling, since the assumption the internal state is server side), are those odds higher? Feel free to ignore that if you wish, and let me know.

Wraith wrote:Third, there's a lot of psychology research on superstition (or more precisely: there's research on why people are superstitious):

Avoiding streaks as you advise is called the gambler's fallacy: it's the mistaken belief that you can't get the same result multiple times in a row. An example is roulette players in a casino who see the ball land 10 times on black in a row, and who then assume the the next time it MUST land on red so they go all in. Nope, the odds of red and black (or in our case: the odds of getting another BP) are still the same.

Assuming the RNG is:
Perfect
and doesn't have flaws: http://www.pcg-random.org/ plenty listed here

Wraith wrote:There's no evidence that the RNG algorithm is humanly predictable

Except all PRNG actually IS predictable. It's called pseudo for a reason, and computers are logical - NO randomness. If you give it 1+1, you are always getting the same result. Which is why all devs are told to seed by time() to help it "look more random" when opening an app.
If you know the internal state, you can figure out what's next in any RNG. MT only has 1 stream of numbers as well, so you just figure out where that pointer for the internal state is. https://jazzy.id.au/2010/09/22/cracking_random_number_generators_part_3.html You only need 624 numbers from it to determine the internal state.
The only reason it is not usually predictable in games, is because its server side. However, for client side RNG: there are people who write programs to look for the internal state to determine what the next rng will be for their next "moves".

Wraith wrote:With that said, if you really wanted to prove your ideas, you could open a few hundred chests while following your tips, and then open an identical number of chests at random times. It would take time, but it's the only way to prove you're correct about what you say.

Love to, donate me the gems.

I can only give ALTERNATIVES to their current methods. Worst case scenario, their luck doesn't improve right? "oh well"? Best case, they feel better,
which is half the goal here (as noted in your selection bias paragraph, which I don't see as a bad thing).

How much is superstition and how much has a basis/research behind it... who really knows. I'm sharing my personal findings, and I'm open to other testimonies, but to say its completely baseless... well, feel free to read those links.
Last edited by VeNoM0619 on Sat May 06, 2017 9:18 pm, edited 1 time in total.

User avatar
cc_Darkis
Cloudcade Staff
Posts: 234
Joined: Mon Mar 20, 2017 1:50 pm

Re: CC, please impliment a mercy rule when opening boxes.

Postby cc_Darkis » Sat May 06, 2017 7:59 pm

Whoa.....what I just step into!

Noting this thread guys and listening to the feedback across the board about this.
"Cauldron is like a box of chocolates. You never know what you're gonna get."~river#44249 :D
"Gems of Shop Heroes are like diamonds of real life."~Prometheus :o

Wraith
Posts: 29
Joined: Sat Feb 11, 2017 6:22 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby Wraith » Sat May 06, 2017 9:19 pm

VeNoM0619 wrote:MT https://en.wikipedia.org/wiki/Mersenne_Twister

Period lengths of MT are known to have dead streaks, and get caught in them.


Ah I'm really sorry, since you didn't mention the algorithm at all I thought you relied entirely on your personal impressions of the outcomes you got when opening chests. If you know the algorithm (and hence it's flaws) this changes things entirely. I'll read up about it, but to be honest a lot of it is way above my knowledge.

True, some could be superstition there, this is just my personal experience getting 3 in a row, 3 separate times. It may not even matter here, my best theory is that in blackjack, if you have 52 cards, you pull a King, what are the odds you pull another king on that next card? When you wait until the deck is reshuffled (other players rolling, since the assumption the internal state is server side), are those odds higher? Feel free to ignore that if you wish, and let me know.


I'm not sure if I understand what you are trying to explain, but I'll try answering anyway:

First I just want to clarify something in case there might be a misunderstanding: as far as I know there's a difference between blackjack and Shop Heroes: in blackjack the odds of pulling a card change after you pulled another card, because the first card you pulled is no longer in the deck. So if you pull a king, there's only 3 kings left in the deck and your odds of pulling another king will be reduced. From what I heard so far, in Shop Heroes your odds of winning a BP never change no matter how many BPs you have already won. Casino roulette is a much better comparison in my opinion.

If you pull a card from a deck, then put the card back in, shuffle the deck and pull another card, the odds of pulling a king are the same every time because you always have 4 kings in the deck. Pulling two kings in a row seems unlikely (that's the gamblers fallacy), but the odds of pulling a king are actually the same every time you pull a card because the fact that you pulled a king previously has no effect at all on the next card you pull.

To put it simply: drawing two kings in a row is not less likely than pulling a king, another card, then a king a gain on the third draw. Or pulling a king, five other cards, then another king on the seventh card draw. All these possible configurations have equal odds of happening. If you're thinking "I just earned a BP on the first chest, there's no way I'll get a BP on the second chest" then my response would be "You're not more likely to get a BP on the third, fourth, fifth, etc. chest than on the second one".

With that clarified and out of the way, I'm not sure what you mean by "(other players rolling, since the assumption the internal state is server side)". Do you mean that the algorithm is affected by the number of players who have opened chests during a specific time period? If so:
- Could you please try explain to me in simple terms how this works exactly? Sounds like a flaw in the algorithm so I'm curious to know about it.
- Gambler's fallacy probably no longer applies, since the act of drawing numbers (opening chests) and maybe even the outcome of each previous draw affects the odds of the next draws (and perhaps your blackjack analogy was adequate after all).

Except all PRNG actually IS predictable. It's called pseudo for a reason, and computers are logical - NO randomness. If you give it 1+1, you are always getting the same result. Which is why all devs are told to seed by time() to help it "look more random" when opening an app.
If you know the internal state, you can figure out what's next in any RNG. MT only has 1 stream of numbers as well, so you just figure out where that pointer for the internal state is. https://jazzy.id.au/2010/09/22/cracking_random_number_generators_part_3.html You only need 624 numbers from it to determine the internal state.
The only reason it is not usually predictable in games, is because its server side. However, for client side RNG: there are people who write programs to look for the internal state to determine what the next rng will be for their next "moves".


When I say it's not humanly predictable, I mean that a human alone can't really predict when the algorithm is more likely to give a specific outcome. I know that PRNG isn't actually random and that if you know all the variables an algorithm relies on and what it does with them then you can predict outcomes. But the important question in practice is: how easy is it for a person (or for the majority of users) to predict the algorithm's outcome?

But anyway, as you pointed out the MT algorithm has flaws that seem exploitable to some degree by the average user (provided they learn how), so that makes the algorithm predictable at least to a certain degree: you can't predict it accurately enough to earn a BP every time, but you can increase your odds.

VeNoM0619
"Great" Mastery Member
Posts: 138
Joined: Wed Apr 19, 2017 2:07 am

Re: CC, please impliment a mercy rule when opening boxes.

Postby VeNoM0619 » Sat May 06, 2017 11:04 pm

First, I need to correct myself, I ended up digging more, and finally found that Unity is using the Marsaglia's Xorshift 128 https://en.wikipedia.org/wiki/Xorshift(this was only mentioned in 1 post, by a non Unity member) so.. my new assumption is based on that (assuming they used the extended variant as well, but no idea if its the XorShift+ or *, lets assume they are using the + now since that came out in 2014, all bets are off if they don't however). Which is still leaps above MT (both performance/statistically speaking), and is acceptable for modern use (however still not thoroughly tested/reviewed sadly).

Regardless, no PRNG is perfect, and new tests keep coming out, finding holes in previous ones (and basically completely invalidating some). Given these new assumptions, I could lean more towards superstition on most things. However, as stated before, half of my reason is that these are alternatives and something to believe in/try. Not proof of anything, but I never denied some being superstitious already, and I welcome others to give me their experiences.


Now for the response, the kings example is based on the fact that RNG's have periods, and "loop" around so to speak (shuffling). All other players are hitting that "deck" with their rolls.This assumes very very small periods of course, and I already admitted that's probably the most superstitious part, however I provided that as a "best guess" explanation why I had such crazy pulls. Not sure how many people get back to back doing immediate consecutive pulls, but I didn't get any back to backs until I started doing it. Sounds crazy? Yep. Works? Who knows. Makes me happy? You bet, my city mates always ask me how I keep getting 3-4 bp every 7-10 pulls.

I will also explain what I have read about their chest system here. There are 3 pools, each pool has its own odds
(NOT ACTUAL CHANCES: 20% Bp, 40% artifact, 40% item) These odds supposedly never change, you will always have those odds at pulling that category, regardless of how many bps you already have (because it does another roll to see which bp/quality item you have afterwards). So those odds are "set in stone". Each chest has its own chances however. Another thread here is guessing that magic+ have higher chances at bps, while <gold is half(?) the chances.

And yes... this isn't about perfect manipulation of the rng... It's simply about knowing when you have better odds, this applies to ALL PRNG. Knowing when the RNG is in a state that gives 53% low numbers, vs when its giving 48% low numbers, doesn't cheat the system completely. But understanding that it does happen is what needs to be understood here. Finding out when, is the difficult part of course.

TLDR: Try it, report back, I'm only attempting to help.


Return to “General Discussion”



Who is online

Users browsing this forum: Bing [Bot] and 4 guests

©2015 Cloudcade, Inc. All Rights Reserved.