Atma farm and why it will be nerfed.

[Deathscythe343]

You don't know that.

He could. All it would take is an idea of how long it took to get any atmas he currently has. Divide that number into 15min increments, one for each day, and count how many days it would be.

[Clavaat]

He could. All it would take is an idea of how long it took to get any atmas he currently has. Divide that number into 15min increments, one for each day, and count how many days it would be.

But that isn't how drop rates work.

[Aegis]

You don't know that.

I don't know that, but I know it's possible. From the people I've spoken to, my drop rate is far from unusual.

It could very well take longer. It's taken me roundabout 45 hours of play to get 10 Atma. That's 4.5 hours each. If I had farmed it in 15 minute sessions with the same drop rate then it would take 18 days per Atma.

There are people in my linkshells/FC with far worse luck than me also.

At 6 mo per weapon, it would take 4.5 years to get all my relics past the Atma stage.

But that isn't how drop rates work.

I'm assuming my drop rate is relatively normal. Certainly, conversations with ls/fc members and people on the forums gives me no reason to belive it's atypical.

Do you have a reason to believe only farming FATEs in short stints gives a better drop rate than marathon sessions?

[Deathscythe343]

But that isn't how drop rates work.

I didn't say that it was. I was merely saying that given your suggestion of farming for 15mins a day, and his experience with the system so far, that he could, mathematically say, that to reach the point he is currently at, would have taken 6 months.

Back to your reply, one could, after 6 months of farming, still need be any where closer to collecting all 12 atmas due to how the drop rate with this system works.

[Clavaat]

-snip-

http://en.wikipedia.org/wiki/Gambler's_fallacy

Basically, the idea is, it is not based on how many you do, but each on individually. You cannot base it off of time, because that would be a fallacy. 2.5% chance on one FATE is still a 2.5% chance on another. In other words, going by your time will not make sense based on other peoples' time. You each have a 2.5% chance per FATE. To make it simpler, let's say its a 50% drop rate. This does not mean in 2 FATEs, you should or are more likely to get the drop. It is calculated each and every time you enter.

Therefore, farming for 15 minutes (another hyperbole on my part, I apologize), has the same odds as doing it for 5 hours. Your odds don't change based on the amount you do.

[Aegis]

http://en.wikipedia.org/wiki/Gambler's_fallacy

Basically, the idea is, it is not based on how many you do, but each on individually. You cannot base it off of time, because that would be a fallacy. 2.5% chance on one FATE is still a 2.5% chance on another. In other words, going by your time will not make sense based on other peoples' time. You each have a 2.5% chance per FATE. To make it simpler, let's say its a 50% drop rate. This does not mean in 2 FATEs, you should or are more likely to get the drop. It is calculated each and every time you enter.

Therefore, farming for 15 minutes (another hyperbole on my part, I apologize), has the same odds as doing it for 5 hours. Your odds don't change based on the amount you do.

It's not an example of Gambler's Fallacy. If I were saying I've done x amounts of FATEs therefore my chance of getting the Atma on the next one is higher than if I'd only done one, that would be gambler's fallacy. I am extrapolating a likely timeframe from a sample set. That's how statistics work.

Flipping a coin once has a 50/50 chance of not getting a Heads.

Flipping a coin 100 times has a 1.268 to the 28th power of not getting a Heads

The odds of the 100th flip being a heads is still 50/50 but the chance of none of the previous flips returning a Heads is Vanishingly Small. Which is a statistical term for techinically possible but in practice close enough to impossible to not be considered.

You can assume rough timeframes from a known occurrence frequency. The precise occurrence frequency is not known but we do have a rough idea of it from testimonials from players. I know of no-one who got them all in less than 12 hours of play, and I know scores of people doing them.

[LineageRazor]

Therefore, farming for 15 minutes (another hyperbole on my part, I apologize), has the same odds as doing it for 5 hours. Your odds don't change based on the amount you do.

The difference in odds isn't because your odds improve, it's because you're doing multiple tries. The Gambler's Fallacy stems from the incorrect belief that if you've lost a bunch, then you're "due" for success. If you've flipped a coin nine times in a row and gotten tails, someone who falls for the Gambler's Fallacy believes that the next toss is more likely to be heads, because the odds of flipping ten tails in a row is only 1 in 1024. In reality, the odds of heads on the next toss is still just 1 in 2, just like any flip.

That doesn't apply here. Here, we are projecting odds of success. A person who plays for 15 minutes will have time to do maybe one or two FATEs. Someone who plays for five hours will have time to do more than a hundred. Someone who does two FATEs is less likely to get a drop than someone that does more than a hundred. The Gambler's Fallacy only applies on the individual attempt. If someone has done more than a hundred FATEs with no drop, and someone else has done one FATE with no drop, they both have an equal chance of getting a drop on the next FATE.

The reason the Gambler's Fallacy fails is because suckers who fall for it mistake past results as having the same odds as projected results. If you've flipped tails nine times, what are the odds that one of those flips is going to be tails? 1 in 1, obviously, because you've already flipped them. Multiply 1 in 1 together nine times, and you get... 1 in 1. Then multiply by 1 in 2, and you get 1 in 2. If you haven't started flipping, and you want to guess the odds of getting ten tails in a row, you multiply 1 in 2 by 1 in 2 by 1 in 2 etc, and wind up with a result of 1 in 1024.

tl;dr:

Believing someone who farms FATEs for 15 minutes in a single day is less likely to wind up with an atma than a person who farms for 5 hours in a single day is simple statistical mathematics. The Gambler's Fallacy has nothing to do with this.

Believing someone who has already farmed FATEs for 15 minutes is less likely to get an atma on the next FATE than someone who has already farmed for 5 hours, this is the Gambler's Fallacy.

[Clavaat]

Fair enough, I stand corrected.

[Aegis]

I would like to point out I am ok with the current drop rates. I spent 4 days in central thanalan and 4 days in outer la noscea, but I was pausing my farming for DF Expert Roulettes and such so it wasn't too boring.

[Synapse]

If I had farmed 15 minutes a day, It would take more than 6 months to get where I am now. And that's still not 12/12.

You could get a drop in each FATE and be over with 12/12 in 30 mins. Or you can spend 2-3 months collecting. RNG is ok when used properly, but RNG added with a unchallenged grind is worse.