Old Events a Statistical Impossibility?

[PawnXIIX]

*WARNING* - Entering no moaning zone.

I wrote a program to simulate the time it would take to open boxes using what I thought were the correct opening chances for each of the boxes:

Free - 1/3
Cash - 1/2
Gold - 1/1

The way I would have personally done the events would be if the box was an open, award a random prize from 1 - 7. One prize is the event item, the other 6 are the other miscellaneous awards.

Here is what was found over 1 trial:

Free open:


Attempts Made: 252
Boxes opened: 80
Gold used: 0

Attempts Made: 176
Boxes opened: 48
Gold used: 0

Attempts Made: 195
Boxes opened: 67
Gold used: 0

Attempts Made: 195
Boxes opened: 66
Gold used: 0

Attempts Made: 162
Boxes opened: 51
Gold used: 0

Averaging out to 196 openings which requires 8.1666 days on average to complete. Sorry free players...should have just spend the in game money.

Cash open:


Attempts Made: 197
Boxes opened: 93
Gold used: 0

Attempts Made: 124
Boxes opened: 63
Gold used: 0

Attempts Made: 195
Boxes opened: 86
Gold used: 0

Attempts Made: 144
Boxes opened: 78
Gold used: 0

Attempts Made: 185
Boxes opened: 92
Gold used: 0

Giving an average of 169 opening required to complete the 10th item. That would cost a person 7.04 days. Quite a shocking thought when you look at it.

Gold opens:


Attempts Made: 43
Boxes opened: 43
Gold used: 645

Attempts Made: 73
Boxes opened: 73
Gold used: 1095

Attempts Made: 77
Boxes opened: 77
Gold used: 1155

Attempts Made: 65
Boxes opened: 65
Gold used: 975

Attempts Made: 70
Boxes opened: 70
Gold used: 1050


Averaging out to just over 65 opening to receive the 10th item. Average gold spent it in the realm of 984, and the total time to complete said task would be 2.75 days.

This is a completely random program, and it was not the exact model of how Funzio did the rewards of prizes but this is how I would have made such an event. Cash opening would have taken 7 days on average even with perfect openings.

This is not a moaning or complaining thread. This is a statistical look at how each of the options plays out in a perfect world.

Anyone want to discuss? anyone...?

[Eyelusion]

I find your research very interesting.

This would change the theory of "pure luck" to more of "not enough time."

[PawnXIIX]

This is assuming that what they mean by "low chance" and "fair chance" is accurate to the fractions. If the open rate is lower than .5 for the "fair chance" than it's even less D:

5 runs is never enough to justify a result. Once is a coincidence, twice is a pattern, but 5 times is never enough to match the law of averages. This also takes into account that on the very first open you can get any reward, because we never really established when the first random items started coming into play. I think the earliest I ever saw was somebody got a bike at 3 items.

[Euchred]

So then in this scenario you have a 1 in 7 chance of getting the 10th tire or whatever? after you have the 9th of course.

[LeDude]

Well given enough time you can easily make enough attempts to eventually get the end prize in the original gacha event. That's why they limit the amount of time, and seeing these results pretty much explains why they never really ran the events for a full 7 days. Given the possibility of good luck and random chance working in a player's favour, even the cash option would reward the end prize to a lot of free players probably more often than the corporation would have intended.

[Eyelusion]

What's impossible to simulate is the Actual "Prize" versus the other "Miscellaneous Item"s per "# of Actual Prizes"? Hehe

Obviously the odds of getting your first few prizes are very high, but then the "Prize Rate" drops off significantly and would highly skew your research into more time needed. What would the 10th prize odds be on opens after all, 1 in 40, 1 in 50, 1 in 70? It's definitely not 1 in 7 so the research is much more extensive than this effort.

[Euchred]

That being said I still think it comes down to time when your playing free. You have to be very lucky to get and event item period nevermind in just 5 days.

[Eyelusion]

That being said I still think it comes down to time when your playing free. You have to be very lucky to get and event item period nevermind in just 5 days.
Quoted for truth and near impossibility.

[jobadass]

What's impossible to simulate is the Actual "Prize" versus the other "Miscellaneous Item"s per "# of Actual Prizes"? Hehe

Obviously the odds of getting your first few prizes are very high, but then the "Prize Rate" drops off significantly and would highly skew your research into more time needed. What would the 10th prize odds be on opens after all, 1 in 40, 1 in 50, 1 in 70? It's definitely not 1 in 7 so the research is much more extensive than this effort.Exactly. Using gold guarantees an opening, but not a "chip". It's hard for me to justify spending RM on an unknown.

[procsyzarc]

Although nice work doesn't really prove anything since we all know the chance of getting the first item is more like 100% and the 10th is 1-2%, this example assues the 10th item is just as easy to get as the first which obviously isn't the case. If the last item really was 1/7 I think we would all have every item and I would have no issue using gold for it.

[i need muney]

You can make more accurate stats out of "log your progress" threads.

[PawnXIIX]

Although nice work doesn't really prove anything since we all know the chance of getting the first item is more like 100% and the 10th is 1-2%, this example assues the 10th item is just as easy to get as the first which obviously isn't the case. If the last item really was 1/7 I think we would all have every item and I would have no issue using gold for it.

Yeah, well this is showing that even with a perfect and ideal situation where every single item is as likely as the last, it will still take over 7 days on average with perfect cash openings. For any astute viewers people would notice that the odds of getting the items are even less than advertised, so it would take even longer.

In the end free openings just come down to pure luck and not the number of openings you have. You can open in theory 10 and win...but then you can open 170 and still not get the 10th item.


You can make more accurate stats out of "log your progress" threads.

I have a feeling that the events were mutually exclusive odds but i'll try :P

For anyone who cares:



public class oldEvent
{
public static void main(String [] args)
{
int totalOpens = 0, totalAttempts = 0, goldUsed = 0;
final double [] FAIL_RATE = {.667, .5, 0};
final int OPEN_TYPE = 2; //0 = free, 1 = cash, 2 = gold
int [] rewards = new int[7]; //item, bike, $25k, $125k, baton, vest, sledgehammer
final String [] REWARDS = {"Item", "Enhanced Bike", "$25,000", "$125,000", "EU Security Baton", "Colnel's Armored Vest", "Carnival Sledgehammer"};
for(int i=0; i<rewards.length; i++)
rewards[i] = 0;

while(rewards[0] < 10)
{
totalAttempts++;
if(Math.random() <= FAIL_RATE[OPEN_TYPE])
System.out.println("FAIL");
else
{
int tempRandom = (int)(Math.random() * rewards.length);
System.out.println(REWARDS[tempRandom]);
rewards[tempRandom]++;
goldUsed += ((OPEN_TYPE == 2) ? 15 : 0);
totalOpens++;
}
}
System.out.println("----------------");
System.out.println("Event completed!");
System.out.println();
System.out.println("Attempts Made: " + totalAttempts);
System.out.println("Boxes opened: " + totalOpens);
System.out.println("Gold used: " + goldUsed);

}
}

[Eyelusion]

Did you sub "Item" for "500x RP"? Nice clean code (error in spelling "Colonel's...", but I'm super picky on code, lol).

Your point is well taken from this code; with the same likelihood of all items, 1 thru 10, the time needed to win would still be much greater than the time allowed by the event.

Given additional restrictions based on the real event odds, it certainly is an event of luck.

For what it's worth, I thought your exercise was awesome! I hope you continue to do more.

Well done!

[PawnXIIX]

Maybe I forgot it? :O

ohh God...Don't tell me i'm that stupid ._. Wow I forgot it. And yeah I'm the biggest stickler on conventions you can imagine...everything needs to be spaced the same, indented the same, lined up...

Changed:
- Added 500xRP
- Made it so that until you get to 4 if you open -> guaranteed item.



Attempts Made: 90
Boxes opened: 41
Gold used: 0

Attempts Made: 71
Boxes opened: 32
Gold used: 0

Attempts Made: 104
Boxes opened: 49
Gold used: 0

Attempts Made: 130
Boxes opened: 67
Gold used: 0

Attempts Made: 71
Boxes opened: 41
Gold used: 0


This is different results now. Average attempts = 93.2 which would yield you 10 after 3.88 days.

This proves something completely different. This proves that there is a system behind the rewards. There is definitely something at work that prevents us from getting the items this way and...well it shouldn't matter any more but this is the first somewhat proof that the events were not truly random.

[Max Power]

That being said I still think it comes down to time when your playing free. You have to be very lucky to get and event item period nevermind in just 5 days.

Of course it does. If there wasn't an end time on the events, everyone would get the grand prize eventually.

The events where you are racing against other players if a different matter, like the current one.

[PawnXIIX]

The events where you are racing against other players if a different matter, like the current one.

This one has us racing other players and the clock :S But at least they gave us a good amount of time, 6+ days works for me. Right now it's not so much a race against the other players, i'm 1 outside the top 500 and i'm wanting to stay close until the end and hopefully I can make a move :S