Watch Path of Exile live and win a Helmet Skin with Twitch Drops!

"

ScootaLewis wrote:

...and that red and gold frisbee hat looks dank.

That pretty much sums up the Dragon hat I just got. Thank you GGG for the free Frisbee hat. ^_^

"

havz0r wrote:

"Every five minutes of view time you'll have a 1% chance to receive one of the helmet skins which means it would take an average of 8.33 hours of view time to receive a Twitch Drop reward."

How did they do this "math"? 8.33 hours would mean 100 5-minute periods each with a 1% chance, but the probabilities are not cumulative. Even if they were, it would mean that you'd get a 100% drop after 8.33 hours? If you "play" once and fail, future plays will still have a 1% probability, not a bigger one. Classic Gambler's Fallacy, which is concerning when coming from game developers.

You'd get a 50% probability of winning a hat after 69 plays (0.99^69 is ~50%), which is 5 hrs. and 40 mins.

The bad part is that, if you watch streams for 8 hours a day for 7 days, you still have a 0.11% chance of not getting a drop. So 1 in 1000 players can expect that to happen to them, sadly.

Average. If we plot how long each hat took to drop, the average viewing time would be 8.33 hours.

"

SabreWolfy wrote:

Average. If we plot how long each hat took to drop, the average viewing time would be 8.33 hours.

How so?

48+ hours so far... nothing, another one of the unlucky people that helps the average "8 hours" out when people getting a helm earlier than 8 hours... #feelsbad

"

SabreWolfy wrote:

"

havz0r wrote:

"Every five minutes of view time you'll have a 1% chance to receive one of the helmet skins which means it would take an average of 8.33 hours of view time to receive a Twitch Drop reward."

How did they do this "math"? 8.33 hours would mean 100 5-minute periods each with a 1% chance, but the probabilities are not cumulative. Even if they were, it would mean that you'd get a 100% drop after 8.33 hours? If you "play" once and fail, future plays will still have a 1% probability, not a bigger one. Classic Gambler's Fallacy, which is concerning when coming from game developers.

You'd get a 50% probability of winning a hat after 69 plays (0.99^69 is ~50%), which is 5 hrs. and 40 mins.

The bad part is that, if you watch streams for 8 hours a day for 7 days, you still have a 0.11% chance of not getting a drop. So 1 in 1000 players can expect that to happen to them, sadly.

Average. If we plot how long each hat took to drop, the average viewing time would be 8.33 hours.

I'm at 81 hours 19 minutes without anything. Why is it not cumulative or at least have one of the common tier a guaranteed drop after 8 hours...

You still have 15 days to get it. Don't worry guys :) I got my Dragon Helmet this morning.

3 days and still don't have hat)

This "1% every 5 min" is so unstable system...

Chances should increase, after every failed attempt:

2% at 10 min, 3% at 15 min and so on)

Otherwise will be some people who newer see free hat... Just like Exited orb, which newer drop)

EDIT: Also, I can't know, if I get bad roll or even don't get roll at all, there is no any sign or message like:

"Yesterday you watch streams for 20h, but don't get hat. Good luck next day!"

I watched for whole week 60+ hours mathil, raiz and others, didn't get it. This Morning I woke up watched DCLara1 for 10 mins and got the hat!


You found an item from Twitch Drops
From: System on Dec 3, 2017 9:00:12 AM
To: lordon

You have been awarded 1x Infernal Helmet

Good luck to others that didn't get it!

"

havz0r wrote:

"

SabreWolfy wrote:

Average. If we plot how long each hat took to drop, the average viewing time would be 8.33 hours.

How so?

I'm not sure a plot is the best way or what I really meant. Instead, I've run a simulation with 1000 "people". The simulation repeats 1000 times, once for each "person" in the promotion.

It selects a random number between 1 and 100 (inclusive) and sets a counter to 1 (for one roll of the 100-sided die).

If the random number is 1 ("win a hat"; we could choose any number as the winning number), then that person's rolls are over and they win in 1 roll.

If not, it repeats, selecting another random number and increasing the counter until the random number is a win. It notes down the number of rolls it took to get a win. It also adds this to a grand total.

It then repeats the whole process again for the next player, and so on, until 1000 people have played.

The final output is the average of the grand total divided by 1000 players. This always comes out very close to 100, as expected. Sometimes it is 98, sometimes it is 103. This means 100 rolls on average are required to win. Each roll takes 5 minutes, so 100 * 5 minutes = 8.33 hours.

Increasing the simulation to 1000 people results in an average of exactly 100, as expected.

This average means that, by the end of the promotion, the average time each person spent would be 8.33 hours. Some would have spent much less (5 or 10 minutes) and some would have spent much more.

Run the simulation yourself if you know how to use Python. The code is here:

https://pastebin.com/ExKEfvaY

"

Guilty1 wrote:

You still have 15 days to get it. Don't worry guys :) I got my Dragon Helmet this morning.

I already started to think that it will be another 15 days with no hat..