Card Distribution

Some may find this interesting. Yesterday I ran several dungeons and noted what cards were drawn (I've got two monitors, it was easy)

Here's the Data(link)

Its worth noting that over 200 draws, there were numerous 'dry runs' where cards simply did not appear - and I think these runs are statistically significant when they run over 10+ draws.The average was around 19 draws without seeing a given card, and the longest dry spell was a whopping 31 draws in a row! I've annotated the data to note when these runs occur.

There's also a propensity to see certain cards draw more or less back to back, followed by a few 'random' draws before sticking on another card.

Bole and Arrow far and away had the highest draw chance at 19%, while Balance and Spear were the lowest at 15%.

They were all around 17%. AKA working as intended.

This statement ignores the dry spells of cards - over 200 card draws they are around 17% sure, but in order you're seeing a lack of certain cards for on average around 19 draws.
This could just be a bole, arrow, or balance you just won't see for two or three encounters. At all.

Consider the fact that your chance of not drawing a given card is 5 in 6. To not draw a card in say, 3 draws is (5/6)^3 - which is about a 57% chance.
(5/6)^19 is ~3%.

These dry spells happened 14 times during the test - which is pretty dang crazy for something that only has around 3% chance of happening. Yesterday I apparently should have been playing the lottery.

Quote:

---

Originally Posted by Phenidate

This statement ignores the dry spells of cards - over 200 card draws they are around 17% sure, but in order you're seeing a lack of certain cards for on average around 19 draws.
This could just be a bole, arrow, or balance you just won't see for two or three encounters. At all.

Consider the fact that your chance of not drawing a given card is 5 in 6. To not draw a card in say, 3 draws is (5/6)^3 - which is about a 57% chance.
(5/6)^19 is ~3%.

These dry spells happened 14 times during the test - which is pretty dang crazy for something that only has around 3% chance of happening. Yesterday I apparently should have been playing the lottery.

---

The game's RNG has always been seen as fairly swingy, as any serious Crafter or Gatherer will attest to. It seems to get "stuck" on certain number ranges at times. It probably has something to do with the way it's seeded.

Then that's an issue with how they're drawing random numbers then. Was there something wrong with the well tested Randoms that comes with nearly every programming language?

Dungeon this morning yielded a crap ton of The Balance--seemed like half my cards drawn. It started before the queue popped and continued to the end of the instance and was kinda beautiful.

SE RNG though, I probably won't see one for a week.

Quote:

---

Originally Posted by GeekMatt

Dungeon this morning yielded a crap ton of The Balance--seemed like half my cards drawn. It started before the queue popped and continued to the end of the instance and was kinda beautiful.

SE RNG though, I probably won't see one for a week.

---

In one Ravana EX run, I had enough Balances to last 2 minutes straight. I think it might happen again before 2016, but we'll see.

Quote:

---

Originally Posted by Phenidate

Its worth noting that over 200 draws, there were numerous 'dry runs' where cards simply did not appear - and I think these runs are statistically significant when they run over 10+ draws.The average was around 19 draws without seeing a given card, and the longest dry spell was a whopping 31 draws in a row! I've annotated the data to note when these runs occur.

---

I have highlighted the problem with your method. Come back when you have data for at least 200.000 draws.

I am continually amazed how little most people understand statistics and randomness.

Quote:

---

Originally Posted by Phenidate

Then that's an issue with how they're drawing random numbers then. Was there something wrong with the well tested Randoms that comes with nearly every programming language?

---

I'm willing to bet there's nothing wrong with how they generate random numbers...

The problem you are having is that it is random.

What people always seem to miss is that drawing 12 Bole cards in a row has exactly the same probability as drawing 1 of each card in order twice in a row.

Drawing a certain type of card has absolutely no impact on what the next card will be. Each draw is a mutually exclusive event.

We as humans will always subconsciously try to find patterns where none exist.

Reminds me of the constantly common occurrence of missing an item on my botanist multiple times in a row when it has a 98% chance.