Author Message

Maineiac12

Rank 1
Joined
26 Aug 2010
Posts
231
Location
Australia
PostedAug 07, 2013 2:57 am
sunday.mourning wrote:
Maineiac12 wrote:
I maintain if there is a 70% chance of success then over 14 tries IF that is in fact the % chance of success then there would have had to have been at least 1 success.  


momommo wrote:
If the odds that any given attempt to enhance will fail is 30%, then the odds that any given stretch of 14 attempts in a row will ALL be failures is simply 30% raised to the 14th power. Just open your computer's calculator app and pop the numbers in.

0.3^14 is 0.0000000478, or 1 in 21 million.

So your definition of "very probable" and mine differ considerably.  


It's the problem of odds versus probability. The odds of failing 14 times in a row are very low, but the probability that the "next try" after the 13th attempt will fail is still 30% because we're discussing independent events that just seem to happen sequentially. Focusing on the idea that one's luck is due after so many tries is what's called the Gambler's Fallacy, because it's playing the odds and disregarding the fact that each event has its own mutually exclusive probability.

...So, OK, I got a bit over-focused on semantics and not the intent of the conversation. I'll admit to that. Smile

However, I don't know how much LivePlex may have tweaked the numbers when converting QB to SB. Or if they're using a "fair coin." And from what I understand, most of the numbers used in the main help thread that lists percentage of success were pulled from QB code, not the translated version. We already know that some graphics were altered, do we know that enhancement and lucky box outcomes weren't also altered? Are we certain the percentages listed in the forums are correct according to Scarlet Blade? I don't even know what the formula they're using to gauge the percentages is--is it (for our 70% example) a random number between 1 and 100 and anything over 30 is considered a success or is the possible pool even larger? If it's the latter, than the percentage given even affects the overall likelihood of multiple "streaks" of success or failure. Is the random number seed based on datetime stamp, as many seeds are? If so, then when failures and the given length of time between attempts also becomes a factor.

For those of you who have played XCom: Enemy Unknown (the latest release), the random number seed is actually generated at the start of the turn, so if you save and make the same attempt in the same order, the exact same sequence of success/fail will happen every time--the game doesn't actually calculate per turn. I'd be curious if SB operates in a similar manner, tasking a number of pre-calculated success/fail ratios long before attempts are even made.

Again, I'd personally prefer it if multiple attempts did factor into the possible success of the next chance. Admittedly, that's for my own selfish reasons--I've spent enough on lucky boxes from the Item Mall and Consignment Shop alike!  


I think what we are all saying is that the percentages given in SB have no relation to reality. Maybe they figure that since it's a fantasy game so can the %'s be fantasy.
Advertisement

momommo

Rank 1
momommo
Joined
06 Jun 2013
Posts
362
Location
United States
PostedAug 08, 2013 9:11 am
sunday.mourning wrote:
Maineiac12 wrote:
I maintain if there is a 70% chance of success then over 14 tries IF that is in fact the % chance of success then there would have had to have been at least 1 success.  


momommo wrote:
If the odds that any given attempt to enhance will fail is 30%, then the odds that any given stretch of 14 attempts in a row will ALL be failures is simply 30% raised to the 14th power. Just open your computer's calculator app and pop the numbers in.

0.3^14 is 0.0000000478, or 1 in 21 million.

So your definition of "very probable" and mine differ considerably.  


It's the problem of odds versus probability. The odds of failing 14 times in a row are very low, but the probability that the "next try" after the 13th attempt will fail is still 30% because we're discussing independent events that just seem to happen sequentially. Focusing on the idea that one's luck is due after so many tries is what's called the Gambler's Fallacy, because it's playing the odds and disregarding the fact that each event has its own mutually exclusive probability.

...So, OK, I got a bit over-focused on semantics and not the intent of the conversation. I'll admit to that. Smile

However, I don't know how much LivePlex may have tweaked the numbers when converting QB to SB. Or if they're using a "fair coin." And from what I understand, most of the numbers used in the main help thread that lists percentage of success were pulled from QB code, not the translated version. We already know that some graphics were altered, do we know that enhancement and lucky box outcomes weren't also altered? Are we certain the percentages listed in the forums are correct according to Scarlet Blade? I don't even know what the formula they're using to gauge the percentages is--is it (for our 70% example) a random number between 1 and 100 and anything over 30 is considered a success or is the possible pool even larger? If it's the latter, than the percentage given even affects the overall likelihood of multiple "streaks" of success or failure. Is the random number seed based on datetime stamp, as many seeds are? If so, then when failures and the given length of time between attempts also becomes a factor.

For those of you who have played XCom: Enemy Unknown (the latest release), the random number seed is actually generated at the start of the turn, so if you save and make the same attempt in the same order, the exact same sequence of success/fail will happen every time--the game doesn't actually calculate per turn. I'd be curious if SB operates in a similar manner, tasking a number of pre-calculated success/fail ratios long before attempts are even made.

Again, I'd personally prefer it if multiple attempts did factor into the possible success of the next chance. Admittedly, that's for my own selfish reasons--I've spent enough on lucky boxes from the Item Mall and Consignment Shop alike!  


The question here isn't whether the odds of consecutive rolls are independent--nobody here has even raised that as something to consider except you.

It's whether 14 failures in a row is just "bad luck" at the posted success probabilities, or a sign that the posted success probabilities are incorrect in some way.

"One second. I think my Commander fainted again." Lachon, Andromeda Server

sunday.mourning

Rank 0
sunday.mourning
Joined
07 Apr 2013
Posts
18
Location
Chicago United States
PostedAug 08, 2013 10:07 am
momommo wrote:
The question here isn't whether the odds of consecutive rolls are independent--nobody here has even raised that as something to consider except you.

It's whether 14 failures in a row is just "bad luck" at the posted success probabilities, or a sign that the posted success probabilities are incorrect in some way.  


It's funny that we both seem to actually agree with the OP, but we have devolved into internal debate anyway. I definitely can shoulder a large part of that blame, as I was the one who derailed the conversation re:probability vs. odds.

As I already admitted, I got stuck on the semantics of the word and not the intent of usage. The word "probability" was being bandied about without clarity of definition. I wasn't disagreeing with the notion that the odds seem to be skewed in the LivePlex/Aeria's favor. I was, however, pointing out that since it's statistically probable to have multiple failures in a row, it doesn't necessarily reflect that the posted percentage chances of success are incorrect.

I, too, would like confirmation that the percentages posted in NinjaKitten's Gameplay Guide in are correct. I'd also like to know where those numbers came from and how the provided percentages in general are being calculated. Because it's really hard to judge the given percentages as accurate or inaccurate or even properly weigh the odds without a better look under the hood.

I also agree that a possible solution to resolve the frustration for enhancements would be to provide an increased chance of success after each failed enhancement attempt. I don't believe that providing such a bonus would even undercut Aeria's profits, considering failures at higher levels can result in the loss of an enhancement and that even with Narak shards, success with a high-level enhancement seems drastically (even unfairly) unlikely.

TheNinjaKitten

Rank 5
TheNinjaKitten
Joined
28 May 2011
Posts
4136
Location
United States
PostedAug 08, 2013 12:09 pm
sunday.mourning wrote:
I, too, would like confirmation that the percentages posted in NinjaKitten's Gameplay Guide in are correct. I'd also like to know where those numbers came from and how the provided percentages in general are being calculated. Because it's really hard to judge the given percentages as accurate or inaccurate or even properly weigh the odds without a better look under the hood.  


Those numbers came from the Korean Queen's Blade forums which were taken from the client files. There have been no alternations to success rates between versions. I can not provide any better confirmation than that so I apologize.

Maineiac12

Rank 1
Joined
26 Aug 2010
Posts
231
Location
Australia
PostedAug 08, 2013 8:43 pm
TheNinjaKitten wrote:
sunday.mourning wrote:
I, too, would like confirmation that the percentages posted in NinjaKitten's Gameplay Guide in are correct. I'd also like to know where those numbers came from and how the provided percentages in general are being calculated. Because it's really hard to judge the given percentages as accurate or inaccurate or even properly weigh the odds without a better look under the hood.  


Those numbers came from the Korean Queen's Blade forums which were taken from the client files. There have been no alternations to success rates between versions. I can not provide any better confirmation than that so I apologize.  


There have been so many changes between versions, all to benefit Aeria, then why should we assume those haven't been changed too? I'm sorry but those success rates simply do not add up.

KinkyJoy

Rank 0
KinkyJoy
Joined
19 Jun 2013
Posts
4
Location
United States
PostedAug 09, 2013 1:54 pm
[quote="Maineiac12"]
giljs wrote:
Whether this practice is right or wrong is irrelevant. It's the way "free to play" games EVERYWHERE are run, and is standard practice. And really, why should Aeria change the game to make any part of it easier for free players? What does Aeria get out of the deal?  


more players, meaning more advertisement... one person plays the game> they like it> they tell 10 more people about it.
10 people play it and like it> they tell 30 more people.
multiply that by about 10 million people and you got an active AND free advertisement system. Aeria would actually save money by doing this without the need for advertisement costs via webpage banners, fast food money cards (yes, this can be considered advertising as well), etc...
It's pretty amazing how none of these companies have even thought of this yet Laughing

DivinoN_65634

Rank 0
Joined
03 Mar 2012
Posts
7
Location
Philippines
PostedAug 12, 2013 1:09 am
seriously? you just failed at 14x at 2-3 upgrades and yet you complain? as far as I know you're not the only person or player who have experienced in that way. Whether it is Aeria games or whatsoever. BUT if that happens at 50x fails then it must mean something. But dude.. its only 14? just because you failed 14x on 2-3 means you're going to spend more AP on much high upgrades. It's a 50% probability. even at 1% success it could happen. I have played alot of MMOs and im not only talking just because i experienced it but also from other people. So those kind of things are just normal. What would happen if all players will gets +10 and max set so easy? can u picture out how boring it will be if all people have +10 stuff? there will be no hardship. (just because you got money you're favored to have high+ gears)
and btw if the ratio of people having a hard time as with you is about 40% then probably it will be noticeable that there's an alter on the success rate. but dude there are people who made it not spending too much.

IMO you just want yourself to have a +XX gear not spending too many AP and not having a hard time achieving it. and you're thinking in the far future how much will it cost based only on that information and experienced? and you dont want that kind of stuff. IF that's so.. go play a private server.. donate yourself with a full +XX set. then satisfy yourself.

Just because you spend AP doesn't mean you're entitle to be strong or entitle to have a higher chances on whatever things. ITS NOT A GUARANTEE THAT WHEN I SPEND A MILLION DOLLARS TO MY ACCOUNT THAT I WILL HAVE +XX SET (although you could if you spend it wisely but thats not the point) Just because they profit in that way doesn't mean its unfair to you.

If you can't take the hardship don't play it.
Display posts from previous:   Sort by: