Regarding the Jumbo Cactpot

[Kacho_Nacho]

Hi! I just had another night of cashing in my weekly jumbo cactpot tickets. I actually got one 4th prize this week!

That got me thinking.

Let's assume I only wish to match the last number and the cactpot winning number is created by a randomly generate number program.

Are my chances of matching the last number truly three in ten or are certain numbers more likely to be generated by a RNG program?

[Hasrat]

Funny enough, I'm in the same boat. First time I got 4th in a while.

For a presumed 3 in 10 chance, I sure do fail to get anything better than consolation most weeks...

[Kacho_Nacho]

Funny enough, I'm in the same boat. First time I got 4th in a while.

For a presumed 3 in 10 chance, I sure do fail to get anything better than consolation most weeks...

Same! I would assume I would get a 4th prize way more often than I do but it's so rare than it's a surprise when I win. It's definitely rarer than three hits every two and a half months. Something feels off.

[JamieRose]

Hi! I just had another night of cashing in my weekly jumbo cactpot tickets. I actually got one 4th prize this week!

That got me thinking.

Let's assume I only wish to match the last number and the cactpot winning number is created by a randomly generate number program.

Are my chances of matching the last number truly three in ten or are certain numbers more likely to be generated by a RNG program?

It's known that no computer can truly draw a random number, however if you wanted to predict what an RNG program was going to draw, you would need access to its code and the higher math skills to draw the correlations.

I wouldn't worry too much about it. I've been playing the same three numbers for four years and I've never done better than 3rd.

[Sove92]

Yes, the RNG is explicitly rigged to draw a different last number than you specifically have chosen the vast majority of the time. This all despite the fact that the jumbo cactpot number is the same for the whole server.

All jokes aside, if you pick a different last number for all three tickets, your chances to win the 4th prize is 30%. Given you can only play once a week, a 30% rate isn't that great as even on average, you could only be expected to win the 4th once a month, which is pretty much in line with my experience over many years.

[Daeriion_Aeradiir]

While computer number generation cannot ever be 'true random', pretty much every software program these days that requires RNG uses formulas and generation techniques to make it as close to true random as it can possibly achieve. Unless you were to go through datasets of trillions of generations, it'd look like true random to the average person.

Keep in mind, it's only a 30% chance (assuming you play 3 different end numbers) that occurs once a week, meaning you're going to have negativity bias purely because the odds of winning are low vs how infrequently you can actually participate. Go back to any old pokemon game and spam 1-hit K.O moves and tell me how many times you manage to actually get them to land over hundreds of uses - that's exactly the odds you're working with here. Each time you fail to land it is another week in FF14 of the Jumbo cactpot.

[Eisi]

A 30% chance to win one and lose two versus a 70% chance to lose all three is in line with my personal experience. I roughly win one ticket every three weeks to a month.

The total number of won tickets should be 10%.

[UkcsAlias]

It's known that no computer can truly draw a random number, however if you wanted to predict what an RNG program was going to draw, you would need access to its code and the higher math skills to draw the correlations

Thats because random isnt actualy random, every single particle can, and will affect the result. The trick is making this too difficult to predict, and use a system to translate this to a number that cannot be influenced.

For example. To generate a random 8 bit digit, you want a system in which you can ensure that all 8 bits are perfectly randomized. And there are tricks for this (inefficient, but because of that reliable). But it does require you to know weaknesses in the system:
- For example, each of the 8 bits, must effectively be influenced by all other bits before and after. (this waterfall effect is also demanded in encryption for the very same reason)
- For any spike in activity (throwing off RNG normaly), the spike must also be balanced out.

A simple way to do this (and i probably miss taking some factors here):
Take the number of options, this is your base. Now you take any weakness numbers (for example if powers of 2 are weak, you take 2), this normaly means all primes below that number you selected. Now multiply all these together (for example if you want a random number between 1 and 4, you take 2x3x4=24 bits), and then take the modulo of the number of options from them. Using tricks you dont need a GB or more of data even when calculting numbers within millions (while it generates, you can already apply modulo at certain steps).

This method is already that reliable, that even if a burst of 1's gets generated by outside activity, the primes cause this to cascade as a waterfall upon the number. And because the speed of generating is faster than the backward calculation (even in quantum physics! if you are using a QP safe method of obtaining the bits), the final number is safe from external influences. Especialy since you must set a treshold of bits in which you can ensure no external manipulation is done (at least 40% of the bits have to be a 1, and 40% must be 0). And these are kept relatively open because RNG has bursts naturaly happen. Its just to avoid a burst attack from changing too much values. Because of all the prime factors, even if half of them are negated in value, cascading still happens reliably.

Its slow for generating a lot of numbers, but lotteries do depend on things like this, and probably have even more tests to detect potential external influence.

Most RNG systems demand to be faster, which causes the weaknesses to appear. But perfectly random info can be generated, as long as you can be sure that the external source is not influenced.

[SassyAssassin]

Thank you for reminding me this exists damn

[Gemina]

With the CPU RNG out of the way, I just want to point out that buying three tickets does not increase your chance to 3/10. Each ticket is a 1/10 chance to get 4th prize.