If anyone has any actual data bearing on this point, how about sharing it.

That is what I posted, forgot to add the % sign after the numbers. Corrected it should say 31.666%, left box 34.333% center box, 34.0009% right box. It was three weeks at 10 to 15 boxes per day. I think it was 275 sets.

We perceive patterns because that is what humans do and odd-seeming runs of certain results happen because every set of boxes is selected at random at the time, and has zero bearing on what the next selection might be.

No pattern, just a record of the outcomes. Does your experience include recording any statistically significant number of outcomes or are you working off the perceived appearance of a normal distribution?

Given the number of bugs in this game, i'm amazed this group dismisses the possibility so easily. No one else has mentioned generating any hard data...

edit: I see spyder has commented while I was typing and has data.

Computers can not generate random numbers but i suspect EA developers use a simpler method. My wife acccount is one maggie short of my account. I have noticed that if my 3 donuts box is her 3 donut box.
I suspect that there is a very large table or an array with entries for each selection. This can also give the apperance of randomness but still in a controlable manner.
If it actually indexed by selecting a new entry for each player it would really be random as the players would be the random variable
For example, say we have 321,231,123 then the first person would get 321, the next 231 etc. If my observation is correct then the same person is getting 321 the first time 231 the second etc

I enjoy reading all the theories, but I maintain that if the assignment of the donuts to the boxes isn't done randomly (or pseudo-randomly as with a computer random number generator), that pattern would have been figured out a long, long, time ago and we'd all be getting three donuts every time on our first choice instead of still discussing this years after this element to the game was introduced.

It's totally random. Like someone else mentioned I never settle for 1 donut. I'll always open a 2nd box if I get 1 donut. Don't open any more when I get 2 or more. Works out pretty well for me.

If anyone has any actual data bearing on this point, how about sharing it.

That is what I posted, forgot to add the % sign after the numbers. Corrected it should say 31.666%, left box 34.333% center box, 34.0009% right box. It was three weeks at 10 to 15 boxes per day. I think it was 275 sets.

You don’t really need do more than eyeball this data set to see that the resulting distribution is pretty darned uniform. But just to be rigorous, if you run through the usual goodness-of-fit test, you wind up with a chi-square test statistic (df = 2) of 0.36, which is associated with a p-value of .84. In other words, you’re not remotely in ball park as far as rejecting the null hypothesis.

Unfortunately, this discussion will almost undoubtedly not settle the matter. There will continue to be users who will be utterly certain that they can detect some kind of “pattern” in the generation of donuts.

PS I seem to always get 1 Bonut from Find Maggie and hate spending $150K to upgrade to 2 Bonuts. (I never get 3 from her)

Me too! Occasionally I get 2 but never 3. Ever. I thought I was remembering it wrong or just plain crazy till I started keeping track and writing it down. Plus you just reaffirmed what I though.

I do not accept unsolicited friend requests! PM me and say hi first.

But just to be rigorous, if you run through the usual goodness-of-fit test, you wind up with a chi-square test statistic (df = 2) of 0.36, which is associated with a p-value of .84. In other words, you’re not remotely in ball park as far as rejecting the null hypothesis.

Thank you for sharing this, Its really interesting to see practical math to apply against this question. Reading up a little on the process, i'll see if I can apply the process to numbers I have, tomorrow when I get to work. Post back the data and my attempt at the fit test.

I reached level 939 a week ago and have been KEM farming the whole time since then. Being short of game cash, I avoided spending extra $ on a second or third choice from the prize boxes to get extra bonuts. Then I tried just clicking on the center box ONLY and guess what? You will almost always get 2 or 3 bonuts and only rarely get 1. If I happen to get 1 only, I then spend the $ to buy another box with a guarantee of 2 or 3 from that choice. It seems to pay off far better than randomly choosing the left or right prize boxes first. I'm sure this no great revelation to long-time 939ers but I thought it might help those new to the lofty heights of TSTO's maximum level.

I've experienced this too but sometimes I click the left and right usually left gives me 1 bonut and right varies but when I get 3 on right or left I keep clicking them and they keep on giving me 3 for about 5 times straight if I did the rat trap delivery truck (I don't KEM farm for some reason)

There is no pattern but one thing i learned to do is to keep betting on any streak you see developing. So if it's been box #3 a couple times in a row and you're tempted to pick another box just because "it can't be box 3 again," and then it is box 3 again, that's when you really feel like a dip for betting against it. So i just keep picking that box until the streak has exhausted itself. Sometimes it can produce 3 or 4 winners in a row that way, "free of charge."

## Replies

That is what I posted, forgot to add the % sign after the numbers. Corrected it should say 31.666%, left box 34.333% center box, 34.0009% right box. It was three weeks at 10 to 15 boxes per day. I think it was 275 sets.

No pattern, just a record of the outcomes. Does your experience include recording any statistically significant number of outcomes or are you working off the perceived appearance of a normal distribution?

Given the number of bugs in this game, i'm amazed this group dismisses the possibility so easily. No one else has mentioned generating any hard data...

edit: I see spyder has commented while I was typing and has data.

I suspect that there is a very large table or an array with entries for each selection. This can also give the apperance of randomness but still in a controlable manner.

If it actually indexed by selecting a new entry for each player it would really be random as the players would be the random variable

For example, say we have 321,231,123 then the first person would get 321, the next 231 etc. If my observation is correct then the same person is getting 321 the first time 231 the second etc

You don’t really need do more than eyeball this data set to see that the resulting distribution is pretty darned uniform. But just to be rigorous, if you run through the usual goodness-of-fit test, you wind up with a chi-square test statistic (df = 2) of 0.36, which is associated with a p-value of .84. In other words, you’re not remotely in ball park as far as rejecting the null hypothesis.

Unfortunately, this discussion will almost undoubtedly

settle the matter. There will continue to be users who will be utterly certain that they can detect some kind of “pattern” in the generation of donuts.notMe too! Occasionally I get 2 but never 3. Ever. I thought I was remembering it wrong or just plain crazy till I started keeping track and writing it down. Plus you just reaffirmed what I though.

Thank you for sharing this, Its really interesting to see practical math to apply against this question. Reading up a little on the process, i'll see if I can apply the process to numbers I have, tomorrow when I get to work. Post back the data and my attempt at the fit test.

I've experienced this too but sometimes I click the left and right usually left gives me 1 bonut and right varies but when I get 3 on right or left I keep clicking them and they keep on giving me 3 for about 5 times straight if I did the rat trap delivery truck (I don't KEM farm for some reason)

THIS IS WITCHCRAFT! I REBUKE IT!