It isn't true RNG. Notice how Kabam words the descriptions in the hero crystals: 100% chance to pull 5*, 6*, etc. It's amazing how many times you can pull original 6* pool even when new stuff is added only a monthly basis.

Tiering does exist, even in the context of the game. Just look at Champion's description. He's set to gain more strength against the best champs in the game (at the time of his release). I'm not naive enough to believe this doesn't figure into the RNG algorithm.

It's speculative and anecdotal, I know, but if you haven't noticed that you generally don't get what you need when it comes to resources, then you've either not been paying attention or you don't play that much.

I'm sure you can back what you said up with some hard evidence right?

Science dictates that any one spin in a casino game, and it is a gamble. However as sample size increases, odds are supposed to resemble the true odds. In this case, 90 pulls should have some resemblance of the true odds the game provides, which clearly aren't even 1:6 ratios.

That's incorrect. Because random sequences are not supposed to be predictable, you can never be certain how any one random distribution will appear. If you could, that distribution would by definition fail a test for randomness. Instead, we can only determine the odds of a particular kind of distribution occurring. If we're trying to test to see if something is random or not, we try to see if distributions show up more often than they should. At some point, the odds against seeing what you're seeing become so high, we assume that it isn't coincidence or random chance, and we're seeing a real thing.

The odds of opening 90 T4CC crystals and getting exactly five of a targeted class are about one in a thousand (it is one in 951). The odds are actually slightly better at getting five or less: about one in 694 (there are more ways to get five or fewer than there are to get exactly five).

Statistically speaking, then, seeing five or fewer of the class you're targeting out of 90 crystals is unlikely, but not incredibly so. Out of every thousand people attempting the same experiment, I would expect at least one or two people to have the same result on average.

This may seem pedantic, but this is also something people often get confused by. Statistics doesn't tell you the result you're going to get. It only tells you the odds of getting a result. If you think you're seeing something that "proves" the crystals are not random, you first have to ask what the odds of seeing what you're seeing are, and then ask out of all the players playing the game how many are likely to be seeing the same thing. If those odds are incredibly low and are so low it is unlikely anyone has ever, or will ever see that thing, that's worth looking deeper at. But if you're seeing something that one out of 700 people are seeing, and there's probably hundreds of other players, if not thousands, that would have seen the same thing if the crystals were entirely random, then that's unlikely to be convincing evidence of anything.

You said it yourself: "one random distribution". This is 90.

Further, you said "odds are better getting five or less". So you are saying odds get BETTER the FEWER you get? I'm guessing you made a typo here.

I do agree with your number of 1:694 as I found 5 in 90 pulls is 0.143%. However I disagree that it doesn't resemble a sample that would resemble a mean distribution. You said it yourself: this is a "one out of every thousand" spin. For that spin to ALSO match the class I needed, the ONLY class I needed is a significant data point. To ignore that is only half the story.

90 is not really a large sample size. Statistically, it's not even the bare minimum sample size for a 95% confidence interval of a six unit wide measurement. The bare minimum is 96.04. That happens to be your measurement btw. 6 units. bare minimimum is 96.04. So, open another 6 and then we'll talk about your bare minimum statistics.

In poker, we don't consider a sample size of less than 10k hands to be meaningful in any way.

## Comments

22,017★★★★★3,746★★★★★1,997Hello @will-o-wisp. Thank you for checking and replying to bring some light on this.

1,962★★★★★I opened 100 t4c crystals and only got 9 cosmic. This is not RNGMaybe play with this for few minutes and see what comes out.

https://djindevstuff.github.io/crystals/tier4cc.htm

2,007★★★306★★★16★In poker, we don't consider a sample size of less than 10k hands to be meaningful in any way.