HQ % is misleading

[Ksenia]

In my experience over gathering sessions, 500 NQ means 100 HQ. That's from hitting the INS key 600 times, no GP use, just 100% Gather rate, 15% HQ on standard nodes.

What I have noticed is that even if I force HQ where GP allows, it will work out roughly the same. If you have 700GP you can force most nodes before ShB to 100% HQ and get all 4 hits to 4X gather for 16 HQ. All that does though is bring down the 'random' HQ amounts, or so I find. In the end I still end up with an average of 100 HQ to 500 NQ.

I'm at the point and assume most others are as well, Actual gathering is done with collectors glove on and farming is done by retainers so that's all retrospect.

[Iscah]

Not sure I understand your point. If I cannot use Unearth III i do not hit the node so it is indeed every node I'm taking about at 67%. But it seems like you might know something about statistics. If my % calculation is wrong (I got it off the internet and am not a math major) please correct me.

The calculation is right but your conclusion is wrong. Missing a "20% chance" four times in a row doesn't guarantee it on the fifth - and you're looking at the wrong statistic anyway because (assuming it is truly random) it's all about the individual chances per gathering attempt.

The simplest way of understanding it is that there is no "running tally" for random chance. You might have your 67% HQ chance and miss three times on a particular node - and yet it does not increase your chances of getting something on the fourth. You still have the same 33% chance of failure that you had on the first attempt.

Your 67% chance, if consistent, means that you are likely to have something in the realm of 67 HQ items and 33 NQ items after a hundred attempts. It's even more likely to reach those proportions over a thousand, or million, collected results.

But short-term you'll see variations. Runs of good or bad luck. None of them affect what you'll get on your next swing.

The breaking up of your string of HQ/NQ results into "chunks" of four chances doesn't help either. Maybe you regularly have a string of six HQ in a row but split over two nodes, and so consider neither of them to fulfil your "20% chance of four in a row".


It also just seems unlikely to me that nobody has noticed your proposed "reduced chances" until now.

[Shadygrove]

Kenyon, your math is off.

Also search "gambler's fallacy."

[Kenyon]

The calculation is right but your conclusion is wrong. Missing a "20% chance" four times in a row doesn't guarantee it on the fifth - and you're looking at the wrong statistic anyway because (assuming it is truly random) it's all about the individual chances per gathering attempt.

The simplest way of understanding it is that there is no "running tally" for random chance. You might have your 67% HQ chance and miss three times on a particular node - and yet it does not increase your chances of getting something on the fourth. You still have the same 33% chance of failure that you had on the first attempt.

Glad to hear the calculation is right. I have never suggested a guarantee at all. In fact we agree on this: its all about the individual chance per gathering attempt. Can you answer this: if something has a chance of happening 20% of the time and has never happened in 1000 attempts is it really 20%?

[Kenyon]

Kenyon, your math is off.

Also search "gambler's fallacy."

Love to see the correct math equation if you please.

[Iscah]

Now, not only do I not see every 5th node yield HQ on every hit...

I have never suggested a guarantee at all.

Suggesting you should expect to see a 4xHQ node at every fifth node is suggesting a guarantee that your one-in-five chance will happen within five attempts.

Can you answer this: if something has a chance of happening 20% of the time and has never happened in 1000 attempts is it really 20%?

Because the maths is more complex than that simple calculation.

That statistic indicates that at any time, there is a 20% chance that your next four attempts will yield HQ. That string of four, if you get it, will not necessarily line up with the intervals of four attempts at one node. (I assume that would make it 20% x 25% = only a 5% chance that a run of four would line up with being attempt 1 to attempt 4 at a single node.)

So if you are judging your stats by picking out one small possibility and judging your results on that, rather than how many individual times you get HQ/NQ, that's going to be an inaccurate measure of your actual luck.


Also, look at it this way: on any one node, you have four chances at the one-in-three outcome of an NQ item. The short-term odds of it disrupting your string of luck are pretty good, even if overall you got three HQ and one NQ - a 75% success rate, above your predicted average chance.

[Kenyon]

Suggesting you should expect to see a 4xHQ node at every fifth node is suggesting a guarantee that your one-in-five chance will happen within five attempts.

Reporting an observation that implies a statistical probability is no guarantee of anything.

But I will grant you one concession: I never considered that I could have 3 HQ hits to end one node and the beginning of the next node give HQ which results in four consecutive HQ strikes. Hence, you are right, it could have happened but I did not see it as I was not tracking cross nodes.

[Mhaeric]

Random is random and so streaks are expected. If you're not seeing streaks then it is not sufficiently random. Ergo, people posting streaks as evidence that it is not random are getting it completely backwards.

Also, there is no adjustment of the HQ percent based on previous hits. Each hit is independent, aside from possibly being modified by actions which may be influenced by previous hits. The number listed on a node is the actual chance for that hit.

[Shadygrove]

Take the example of 67% HQ chance. With 4 hits the chance that all 4 hits will be successful is, statistically .67*.67*.67*.67 = 20%. Now, not only do I not see every 5th node yield HQ on every hit I have NEVER seen it happen at all! This means that SE reduces your next hit chance after a success.

For example, if your HQ chance is 41% and you got HQ on the first of 4 hits, your chance of getting HQ is now effectively reduced to 16%. Get a second HQ success and your chance is now reduced to near 0% for the last two. I say near zero because I have actually seen my botanist get 4 HQ hits (a couple times). But never on my miner.

It is the assumptions in your math that are wrong.

Each try is its own independent universe with no tie in with any preceding or following attempt. If your HQ chance is 41% and you got HQ on the first of 4 hits you have a 41% chance on every subsequent attempt.

Also search "gambler's fallacy."

[Canadane]

RNG is RNG bro.
There was a dev response about this translated for us to read: http://forum.square-enix.com/ffxiv/t...Random-Comment