Pandora's Box 2018 Feedback Thread

DeletedUser4945

Guest
So either I was taking too small of a sample to make a theory about how sending random troops = WIN or a flaw was recognized and corrected, but my recent attempts in sending the wrong troops hasn't been as effective in my latest attempts. I give up. There's no secret to making this less frustrating.

*Except as suggested above by not playing at all.
 

DeletedUser34939

Guest
I just send the minimum troops allowed in the best ratio based on what I have available. I finished box 27 earlier today and I think the max troops that could be sent were like 498 but I was sending less than 200 per attempt and using proper ratios was getting like a 40% chance of success where as sending max troops would have only increased it to like 46% so I'm getting 3x as many attempts with my troops for only a slightly lower chance of success.

I'm not getting anything near the 2 in 5 success rate I should have at 40% but at least I'm not running out of troops before closing a single box. So the only thing really frustrating now is the time it takes for an attempt to finish. If it's taking 25 mins per attempt and I'm only successful about 20% of the time it's taking over 2 hrs for each successful capture assuming I'm available to click every time one round finishes.
 

DeletedUser31931

Guest
There is a very simple fix for this event in my mind. Double the length of missions, and double the chance of their success. And cut in half the number of units given out, to keep the same ratio as current event.
 

DeletedUser34939

Guest
I think that just increasing the chance of success would be enough of a fix.

The duration would be totally reasonable if you didn’t fail so much and the rewards are perfect - enough to make it worth playing the event but not big enough to abuse.

I’m a pretty active player and I’ve spent gold on troops and I’m struggling to close a box a day for the last few days.

There is still a lot of time in the event so we will see how it works out in the end I guess.
 

DeletedUser36697

Guest
some points wrt to this event
World Hyperborea speed 1
spent no gold
got all 10 troops increases each day
Delay after completing a reward..5hrs.. totals delays after completing reward/level 20... 3.958333 days ..19x5=95hrs = 3.958333 days
gets better...39 x 5 = 195hrs = 8.125 days if you get to level 40

That means I would have approx 11 days of playing TOTAL time


Now factor in the increasing time to complete the missions as well as the ever increasing failure rates of missions..seems alot to take on

I would have to .. guessing here... complete 4/5 REWARDS A DAY FOR THE First 5 days to get nears 35 much less 40

so stopping at 20 and I have 1000's of troops left

shes (Pandora) not worth the frustration or aggravation
 

DeletedUser31852

Guest
Mojosirin thanks for the advice! Works well for me too to use less troops

I mean what I love about this event the most is that it’s all got us struggling, theorizing and working together on how to crack these boxes open in reasonable limits. And events aren’t meant to be finished by everyone or even anyone in my opinion ^_^ it’s a high bar that’s maybe unattainable. The event definitely brought a great twist to daily troop building and farming I find, which most easy events don’t - they just mean a few more robotized clicks to do. Well this one is different, you have to sculpt your activity to fit it.

And honestly I sincerely doubt there is anything hidden behind the success rates percentages. From a programmer perspective it’s very easy to put a given success rate and the computer will do it - of course on few trials (less than 100 for example) it may seem like some patterns emerge like what was suggested here that base value is better or what I talked about with others in my world that cutting time in half increases chance of success, but I think those are all just delusional fictions and hopes concocted by our human brain and it’s just programmed correctly: success rate we see if actual success rate. Why would someone bother programming it more complicated
 

DeletedUser34939

Guest
Yeah I don't see why they'd put anything behind the programming to screw with the success rate. Sorry if my previous posts made it seem like that's what i was alluding to. Pretty sure even if inno wanted to do that their lawyers would stop them.

Using less troops per attempt definitely isn't going to improve your odds. It's just gonna give you more opportunities before you run out of troops and I'll take 3 tried at 35% over 1 try at 40% any day.
 

DeletedUser31852

Guest
And if an event is judged by posts in feedback thread this one will be up top ^_^

Innovation is always good, even if it was the wrong direction, you always learn things
 

DeletedUser21774

Guest
I keep getting the same mission over and over. All I know is I better not meet Pandora on the street or we are going to have a chat. She's like a 2 year old "Don't open that &@M# box!"
 

DeletedUser31931

Guest
And honestly I sincerely doubt there is anything hidden behind the success rates percentages. From a programmer perspective it’s very easy to put a given success rate and the computer will do it - of course on few trials (less than 100 for example) it may seem like some patterns emerge like what was suggested here that base value is better or what I talked about with others in my world that cutting time in half increases chance of success, but I think those are all just delusional fictions and hopes concocted by our human brain and it’s just programmed correctly: success rate we see if actual success rate. Why would someone bother programming it more complicated

While I agree from a logic standpoint, I now have no doubt that there is something hidden behind the success rates percentages. You can see my data here: https://en.forum.grepolis.com/index.php?threads/proof-that-pandoras-box-is-a-lie.61859/ and let me know your thoughts.
 

DeletedUser34939

Guest
I only looked over your data briefly so I could be wrong here but it seems like you just took your average chance of success and then your actual average success rate and that doesn't really work.

If I have 5 chances to do something and the first I'm guaranteed 100% success rate and the next 4 I only have a 25% success rate the simple math says that 200% total/5 attempts = 40% average success rate. That makes it look like I should be successful 40% of the time when I'm spending most of my attempts well below that threshold at 25%.

I find the failures frustrating for sure and I think they should tweak that a bit so that it doesn't effect player morale but I highly doubt inno would have anything going on behind the scenes to skew our success rates. I'm no lawyer but it seems to me that since they collect money for these events they'd be in some legal hot water if they misrepresented what your money is buying you. They're a sizable company these days. Doubt they would make that mistake.
 

DeletedUser31931

Guest
I only looked over your data briefly so I could be wrong here but it seems like you just took your average chance of success and then your actual average success rate and that doesn't really work.

If I have 5 chances to do something and the first I'm guaranteed 100% success rate and the next 4 I only have a 25% success rate the simple math says that 200% total/5 attempts = 40% average success rate. That makes it look like I should be successful 40% of the time when I'm spending most of my attempts well below that threshold at 25%.

That's simply not how the math works. I will even use your numbers:
unknown.png

The 3rd column is basically if 1 trillion people did these same 5 trials, the percentage of them that would succeed on that one attempt. It is, of course, possible that only 23% succeed, or that 26% do so. However, if the data sample is large enough it is very unlikely for the observed percentage to greatly differ from the expected.

The average chance of success and the percentage of times you can expect to succeed are exactly the same; I hope this makes sense. It is basic probability. :)

I have also wanted to believe that Inno was not lying to us, but unfortunately the data is too clear to ignore by this point.
 

DeletedUser34939

Guest
It’s not that simple.

If you have a 1/100 chance of success and 100 attempts you are less likely to be successful than if you have a 50/100 chance of success and only try once.

But if you have a 10/100 chance of being successful and 100 attempts you are more likely to be successful than if you only have 1 shot at 50/100 odds.

A simple average is not a realistic way of rationalizing your odds of being successful over a series of attempts when the ratio never changes even before you try rationalizing your actual success rate.

The attempts where you have a lower chance of success have a greater effect on your overall rate of success than simply averaging them together with the greater chances of success.

You’re trying to use basic probability to explain my example and expanding that over the number of attempts made (million, trillion...) which would be correct. No matter how many people make an attempt at 25% we should expect to see 25% roughly succeed.

But multiple attempts at the same ratio is not the same thing as overall chance of success across many attempts with fluid ratios.
 

DeletedUser31931

Guest
If you have a 1/100 chance of success and 100 attempts you are less likely to be successful than if you have a 50/100 chance of success and only try once.
I am sorry, but this math is simply atrocious.

The average result from 100 attempts of a 1% chance is 1 success and 99 failures. (100 * 0.01)
The average result from 1 attempt of a 50% chance is just half a success, and half a faliure. (1 * 0.5)
1 > 0.5

Not only is your statement incorrect, it's off by a rather absurd amount. I am not going to debate simple math with you here. You could just as easily google "middle school probability lessons" and get a better teaching than anything I can provide.
 

DeletedUser34939

Guest
I can't tell if something went over your head or if you're reading too deeply but either way you're not getting the point.

Just because the basic math says you will fail 99 times and succeed once doesn't mean that is the case. Realistically you are going in to each of those attempts with a 1/100 chance of success and most likely you will not be successful ever. your 100th attempt is not taking into account that you already failed 99 times.

You've got a better chance of success with 1 at 50/100 than 100 attempts at 1/100. If you don't understand this I have a poker game I'd like to invite you to.
 

DeletedUser31931

Guest
I can't tell if something went over your head or if you're reading too deeply but either way you're not getting the point.

Just because the basic math says you will fail 99 times and succeed once doesn't mean that is the case. Realistically you are going in to each of those attempts with a 1/100 chance of success and most likely you will not be successful ever. your 100th attempt is not taking into account that you already failed 99 times.

You've got a better chance of success with 1 at 50/100 than 100 attempts at 1/100. If you don't understand this I have a poker game I'd like to invite you to.
I wasn't joking when I said to google middle school math principles; they are likely things you have forgotten over the years. I tried to dumb down my explanation so much that it would even get into your head, with examples and simple math. But if you want to pretend that you actually have any clue what you're talking about, here is the more advanced explanation:

b(x; n, P) = { n! / [ x! (n - x)! ] } * Px * (1 - P)n - x
The is the unaltered binomial probability formula.
"
Binomial Experiment
A binomial experiment is a statistical experiment that has the following properties:

  • The experiment consists of n repeated trials.
  • Each trial can result in just two possible outcomes. We call one of these outcomes a success and the other, a failure.
  • The probability of success, denoted by P, is the same on every trial.
  • The trials are independent; that is, the outcome on one trial does not affect the outcome on other trials.
Consider the following statistical experiment. You flip a coin 2 times and count the number of times the coin lands on heads. This is a binomial experiment because:

  • The experiment consists of repeated trials. We flip a coin 2 times.
  • Each trial can result in just two possible outcomes - heads or tails.
  • The probability of success is constant - 0.5 on every trial.
  • The trials are independent; that is, getting heads on one trial does not affect whether we get heads on other trials.
Given x, n, and P, we can compute the binomial probability based on the binomial formula:

The binomial probability refers to the probability that a binomial experiment results in exactly x successes."
https://stattrek.com/probability-distributions/binomial.aspx?tutorial=ap

b(0; 100, .01) = { 100! / [ 0! (100 - 0)! ] } * .01*100 * (1 - .01)100 - 1
Here I have just substituted in the values we're trying to solve for.
x = The number of successes we're testing for. I am starting with 0
n = The number of trials. (You said 100)
P = The probability of success on any given trial (You said 1%, or .01)

b(0; 100, .01) = 0.36603
This is simple arithmetic, which you are more than welcome to double check yourself. I did it first by hand, then with a calculator.

I even found an online calculator which does the same thing:

As you can see, the chance of not succeeding in any of the 100 tries is ~36.6%. That means that the chance of succeeding once or more is ~63.4%. I certainly hope you don't need me to explain this concept to you as well. In case you were having some troubles with this part too, 63% is bigger than 37%. ;)

And that, my friend, is why you are simply incorrect:
If you have a 1/100 chance of success and 100 attempts you are less likely to be successful than if you have a 50/100 chance of success and only try once.

Now how do I go about joining this poker game you speak of??
 
Last edited by a moderator:

G-Fight

Chiliarch
middle school math principles
Middle school? So thats about for 12 years olds amirite?

Binomials aren't "basic" statistics tho, you only see them in high school if you follow a harder math-oriented course or when you go to uni lol
 

DeletedUser31931

Guest
Middle school? So thats about for 12 years olds amirite?

Binomials aren't "basic" statistics tho, you only see them in high school if you follow a harder math-oriented course or when you go to uni lol
Haha that's why I first tried to explain it to him with expected value, in an even more dulled-down format.

"How many cookies does the cookie monster get, if half the time he drops them, and he makes ten cookies." hahaha

But alas, even that didn't get through to him, so I went with the more mathematical explanation. Yeah I did these in high school, but they never really stuck till uni. I certainly wasn't suggesting that little Joey starts learning binomial distribution theory in 5th grade. xD

I just can't stand when people with absolutely no idea what they're talking about get all snotty and rude. Like what the heck, we all have different fortes, so why try to pretend he is familiar with an area he clearly isn't...ya know?
 

DeletedUser31931

Guest
Let's not start arguing and attacking each other. I'll bring this up with Inno in the morning and get some clarity on the matter.
Was not trying to argue here at all, just instructing someone on the basic principles of math. ;) My other thread should be much more helpful for inno to look at...though I think they know exactly what they are doing...
 
Top