Gathering and probability....

[Nenin]

TL;DR = Math is fun! And also very misleading...

I'm pretty well versed in math but I don't really get what you're driving at here.



If you want to use strict definitions the numbers generated by a RNG aren't actually random anyway. They are pseudo-random because they need a seed number which is usually some combination of internal states like time/temp.

This doesn't mean that modern RNGs are crap though. Many are able to generate an almost perfectly uniform distribution. The level of 'randomness' is robust enough for massive physics computations in fields like magnetohydrodynamics.


I can't explain a result like 14 consecutive failures with a ~93% chance to succeed though. They undoubtedly use a generator like the Mersenne Twister which is more than good enough for our purposes. I guess maybe if they used a terrible RNG, and they used a client side seed, and your computer was a potato with a broken clock it could happen.

[Rbstr]

All of the anecdotal QQ in the world won't get me to believe it's actually messed up unless you document every hit and probability. Instead of just recalling a particularly disappointing set. That's just your bias hanging out.

[LloydShade]

Statistic! If for you that doesn't sound evil or nightmarish, then:
A) It's your job.
B) You don't know enough about it.

What if...
-SquareEnix controls the successes server-wide. Then, it doesn't matter if ONE player takes on all the failures, 99% will still hold. The scale might not be limited to one player.
-SquareEnix decided to create rarity on an item and decided to make it harder to get it server-wide for a certain time. In the long run 99% will still holds. The scale might not be instantaneous.
-99% does hold and you are just very unlucky.

So, over an amount of players, over an amount of time, "luck", take your pick!

On a side note, nothing is random in a computer, NOTHING. Everything is mathematically exact and precise. A so called random number on the computer (dice roll for instance) only looks random. In reality, it's just a really complex mathematical formula based on processor speed and time and the number range. In the end, SquareEnix controls everything.

[AccountBanned]

I was at 80% and missed 6 attempts in a row. I was like "wtf is that?!" right in front of my wife. Is that what you want Sony? You want me to look dumb in front of my wife? Because that's what you're doing to me.

It really is annoying.

[JayCommon]

I'm pretty well versed in math but I don't really get what you're driving at here.

I guess I could have concluded my post a little better than I did.

Basically, all I'm getting at is that many of the people complaining about the 'percentages' in this thread are looking a few instances and deciding to base their entire experience off of a bad run or two, but like I stated in my instance of getting 5/5 HQs with a 31% chance of landing, the extremes can land on either side of the fence.

It seems like people are taking their 84% (just an example of a %) straight to heart when I don't think it's meant to be that way. Even if you have an 84% to hit, you may miss 9/50 which is far less than 84%, but there is no way for you to know how many you missed out of 1000 unless you sat there and counted for yourself. And no one here who is complaining is going to remember the times they had 84% and got 50 in a row to hit (I know it happens, I almost have my Luminary pickaxe, I have mined enough to see that these percentages DO equal out over time, or at least get real close).

Basically, we don't get a counter in the game or a stat screen to prove how many hits/misses we get at every % point. People see that they miss 3/4 on a 84% and are like "QQ!!!" but don't want to mention the 40 in a row they landed before missing those 4.

[kukurumei]

Well there is a difference in gathering since you do not have a sample size restraint. It's pretty easy to get a string of "oddness" because you essentially have unlimited tries to produce it.

Now if you try to predict the amount of times you get 4/5 fails in that exact combo, is much much harder.

[TetsuyaHikari]

Since I've been leveling Miner, I can definitely see that the probability in this game certainly isn't right. If I have an 80%+ probability to mine something, don't have me obtain nothing 3-5 times in a row. On the opposite hand, don't have me succeeding on every strike while sitting at 30%+ probability.

It just doesn't work that way. I swear, it's like every time I'm low on probability and bump it up to 70 or 80%+, I actually manage to miss more than I was before I even used it, at a lower probability. Oh and if you're going to give me a 93%, don't miss three times consecutively.

That's just a kick in the dick. I understand the 7% chance is there, but come on... Really?

EDIT: At the end of the day, I'd just be comfortable with them sticking us with a '?' like we get when we find an unknown when we're mining. Allow the skills and such to increase our chances, but don't give us a percentage. It's obvious the number there is just a placeholder.

[Millerna]

Computers are often not great at generating random numbers actually...

I've had a fail once at 100% chance, but I'm pretty sure that was a bug. Haven't been able to reproduce it either.

[Nictuswolf]

98% chance to HQ item... Get normal..

Not once, not twice, not 3 times.

I have gotten this NUMEROUS times.

[Pseudopsia]

Computers are often not great at generating random numbers actually...

I've had a fail once at 100% chance, but I'm pretty sure that was a bug. Haven't been able to reproduce it either.

It's call pseudorandom for a reason. ^^
pseudo == false.

I would consider very minor bug in what is shown to the UI. If the UI is rounding when formatting the number 0.996 then it will show as 100%.

With that percentage it will be hard to reproduce.