Worst streak pulling 5s - LTs NEED to change

cyineedsn

Pulled 10 more with some LTs I had saved up and championing, so I'm now at around 30-35 without one. Oof.

amusingfoo1

amusingfoo1 wrote:

I broke my streak of 40-straight 4* pulls yesterday, finally. I'm a 4* transitioner, so it could be worse, but that was really frustrating.

Just an update. I've since pulled about 15 more legendaries (and have a few more, unopened), and have almost gotten up to 10%..

Tenaciousdecaf

Just ended a streak of 56 without a 5*. I now have 6 total covers out of approximately 150 tokens opened. I feel so far behind the curve.

Polares

Tenaciousdecaf wrote:

Just ended a streak of 56 without a 5*. I now have 6 total covers out of approximately 150 tokens opened. I feel so far behind the curve.

56 ? THIS IS RIDICULOUS! Devs NEED to address this problem NOW, ASAP!!!

How is it possible that someone who has won 150 tokens has just 6 covers!

DEVS THE SYSTEM IS BROKEN!!!! 32% of the people that has voted has had an streak of more than 40 tokens without pulling a 5!!! One third!!!

I dont know how they can be so ****...

Polares

I know there are a lot of issues open with Champions, etc. But I don't want to let this thread go, I want answers from devs, because this is clearly not working.

I made some math and the chances of getting 40 non 5s in a row is 1.4%, and 30% people from the poll has had this streak. So, clearly there is something wrong going here.


PS: Please dont tell me that for every token the chances are 10%, I KNOW THAT, but the chances at the start of the streak of opening 40 tokens and not pulling any 5 is that. I KNOW that after 40 tokens my chances for the next one are still 10%, but if BEFORE opening even one wanted to know the chances of this happening, this are the odds, 1.4%.

Eddiemon

Polares wrote:

I know there are a lot of issues open with Champions, etc. But I don't want to let this thread go, I want answers from devs, because this is clearly not working.

I made some math and the chances of getting 40 non 5s in a row is 1.4%, and 30% people from the poll has had this streak. So, clearly there is something wrong going here.


PS: Please dont tell me that for every token the chances are 10%, I KNOW THAT, but the chances at the start of the streak of opening 40 tokens and not pulling any 5 is that. I KNOW that after 40 tokens my chances for the next one are still 10%, but if BEFORE opening even one wanted to know the chances of this happening, this are the odds, 1.4%.

Except they aren't the odds. It helps if you understand math when you rant about it.

The odds of getting 40 non 5s in a row is 1.4% (Well seeing as it is 1.478% you should really be using 1.5 if you are going for two significant digits.) Now if you only open 40 tokens ever your odds of experiencing the event is 1.4%. But once you open that 41st token your odds rise to 2.7%. By the time you have opened 60 tokens it has jumped to 25%, and at 80 tokens it has jumped to almost 44%.

So if the average respondent had opened around 70 tokens then 30% of them experiencing that failure rate seems correct.

Polares

Eddiemon wrote:

The odds of getting 40 non 5s in a row is 1.4% (Well seeing as it is 1.478% you should really be using 1.5 if you are going for two significant digits.) Now if you only open 40 tokens ever your odds of experiencing the event is 1.4%. But once you open that 41st token your odds rise to 2.7%. By the time you have opened 60 tokens it has jumped to 25%, and at 80 tokens it has jumped to almost 44%.

So if the average respondent had opened around 70 tokens then 30% of them experiencing that failure rate seems correct.

What do you mean with that ? The odds of not pulling a 5 with 41 tokens is 1.3302% and with 80 tokens is 0.0218%, they get even smaller not bigger!!! You can't consider the odds for 40, and then for the 41 element consider it is 10% and multiply them (I don't really know what are you doing there), you consider all the streak as a whole, and then EVERY particular pull at 10%, because of course, the system doesn't have memory.

Please explain your math because I don't really know what it is you are calculating there...

Spencer75

Polares wrote:

I know there are a lot of issues open with Champions, etc. But I don't want to let this thread go, I want answers from devs, because this is clearly not working.

I made some math and the chances of getting 40 non 5s in a row is 1.4%, and 30% people from the poll has had this streak. So, clearly there is something wrong going here.


PS: Please dont tell me that for every token the chances are 10%, I KNOW THAT, but the chances at the start of the streak of opening 40 tokens and not pulling any 5 is that. I KNOW that after 40 tokens my chances for the next one are still 10%, but if BEFORE opening even one wanted to know the chances of this happening, this are the odds, 1.4%.

Huge problems with the presentation of your math.

First off if everyone pulls enough tokens every single one of us will have a steak of more than 40 without pulling a 5 star.

Second there's a huge selection bias in the people taking the poll.

Who is going into this thread and then answering that poll? This isn't a fair sample of users, it's a lot of people mad about their pulls.

Lemminkäinen

Polares wrote:

Please explain your math because I don't really know what it is you are calculating there...

Imagine opening a million tokens. How likely do you think it is to get a streak of 40 without pulling a 5* somewhere along the line? It's very close to 100%. Eddiemon was calculating the chance of getting a streak of 40 when opening more than 40 tokens.

Eddiemon

Polares wrote:

Eddiemon wrote:

The odds of getting 40 non 5s in a row is 1.4% (Well seeing as it is 1.478% you should really be using 1.5 if you are going for two significant digits.) Now if you only open 40 tokens ever your odds of experiencing the event is 1.4%. But once you open that 41st token your odds rise to 2.7%. By the time you have opened 60 tokens it has jumped to 25%, and at 80 tokens it has jumped to almost 44%.

So if the average respondent had opened around 70 tokens then 30% of them experiencing that failure rate seems correct.

What do you mean with that ? The odds of not pulling a 5 with 41 tokens is 1.3302% and with 80 tokens is 0.0218%, they get even smaller not bigger!!! You can't consider the odds for 40, and then for the 41 element consider it is 10% and multiply them (I don't really know what are you doing there), you consider all the streak as a whole, and then EVERY particular pull at 10%, because of course, the system doesn't have memory.

Please explain your math because I don't really know what it is you are calculating there...

I'm calculating the odds of getting 40 non 5s in a row when opening more than 40 tokens. Which is the specific event you were going on about because that's the specific question in the poll.

I'm not calculating the odds of 41 non 5s in a row, or 80 non 5s in a row as they are irrelevant to the poll or your complaint.

To simplify it, what are the chances of flipping heads twice in a row if you flip a coin twice. 25%. What about if you flipped it 10 times (still only looking for 2 heads in a row)? Way way better than 25% right? When you flip it ten times there are 9 sets of two coin flips in a row (First and Second, Second and Third, ... , Ninth and Tenth). Each of those sets of two has a chance of being a pair of heads.

So when you open tokens, when you open exactly 40 you have one set of 40 to check. But when you open 41 you now have two sets of 40, the set 1-40 and the set 2-41. When you open 80 tokens you have 41 sets, the sets 1-40, 2-41, ... , 41-80. So you need to look at the odds that any of those sets doesn't contain a 5*. And as you are running the test 41 times, the probability increases significantly over running the test just once on a specific 40 tokens.

woopie

Finally broke my 50+ streak and pulled a red SS, black SS, and black OML with all the champion LTs. Let's hope this streak does not resume...

Phaserhawk

I would say the pull rate is much closer to 5% than 10% that it's suppose to be. Perhaps in the long run it would go to 10% but the problem is they will probably change the system before you could ever start hitting your stride.

Fact, plain and simple.

4* and 5* progression has stalled due to randomness. The worst thing, I mean the worst thing they ever did to this game was get rid of the 1300 4* cover for an LT token.

You went from 100% chance of getting a cover to a 1.36% of getting that cover.

At this point all championing has done is make it less hurtful when you pull a cover you need

Polares

Eddiemon wrote:

Polares wrote:

Eddiemon wrote:

The odds of getting 40 non 5s in a row is 1.4% (Well seeing as it is 1.478% you should really be using 1.5 if you are going for two significant digits.) Now if you only open 40 tokens ever your odds of experiencing the event is 1.4%. But once you open that 41st token your odds rise to 2.7%. By the time you have opened 60 tokens it has jumped to 25%, and at 80 tokens it has jumped to almost 44%.

So if the average respondent had opened around 70 tokens then 30% of them experiencing that failure rate seems correct.

What do you mean with that ? The odds of not pulling a 5 with 41 tokens is 1.3302% and with 80 tokens is 0.0218%, they get even smaller not bigger!!! You can't consider the odds for 40, and then for the 41 element consider it is 10% and multiply them (I don't really know what are you doing there), you consider all the streak as a whole, and then EVERY particular pull at 10%, because of course, the system doesn't have memory.

Please explain your math because I don't really know what it is you are calculating there...

I'm calculating the odds of getting 40 non 5s in a row when opening more than 40 tokens. Which is the specific event you were going on about because that's the specific question in the poll.

I'm not calculating the odds of 41 non 5s in a row, or 80 non 5s in a row as they are irrelevant to the poll or your complaint.

To simplify it, what are the chances of flipping heads twice in a row if you flip a coin twice. 25%. What about if you flipped it 10 times (still only looking for 2 heads in a row)? Way way better than 25% right? When you flip it ten times there are 9 sets of two coin flips in a row (First and Second, Second and Third, ... , Ninth and Tenth). Each of those sets of two has a chance of being a pair of heads.

So when you open tokens, when you open exactly 40 you have one set of 40 to check. But when you open 41 you now have two sets of 40, the set 1-40 and the set 2-41. When you open 80 tokens you have 41 sets, the sets 1-40, 2-41, ... , 41-80. So you need to look at the odds that any of those sets doesn't contain a 5*. And as you are running the test 41 times, the probability increases significantly over running the test just once on a specific 40 tokens.

Ok! Now I know what you meant

Then, I have to check (my statistics are a little bit rusty) but I don't think you can just multiply the chances by 41 when considering 80 tokens (there are 41 occurrences so I multiply by 40) because the samples you are considering overlap with each other, so I think the maths to calculate these series of events are more complex. It is going to be more than what I said, but I doubt it is going to be 44% for 80 tokens.

If you were correct, with 150 tokens (so more than 100 ocurrences) the chances would be 100% of having a 40 tokens streak.

Edit: Yeah I just checked and your math is incorrect, the correct result is

S(N,K) = p^k + Sum j=(1,K) of p^j-1(1-p) * S(N-j,K) (Very hard to write formulas in a forum )


I can't calculate it here but it is not just multiplying the number of samples by the base chance, as you can see is far more complex.

Lemminkäinen wrote:

Polares wrote:

Please explain your math because I don't really know what it is you are calculating there...

Imagine opening a million tokens. How likely do you think it is to get a streak of 40 without pulling a 5* somewhere along the line? It's very close to 100%. Eddiemon was calculating the chance of getting a streak of 40 when opening more than 40 tokens.

Yeah of course, with a sample that big it will probably be close to one, but we have just opened around 80-150 tokens, not a million!

Polares

I needed some time to code a function to calculate it but here it is:

For 80 tokens, the chance of having a streak of 40 no 5s is:

7.3% !!!

Which you can see is far less than what you thought, and far less than the results we are getting.

With 150 tokens the chance raises to 17.05% which it is still much less than the results of the poll.

simonsez

Polares wrote:

I needed some time to code a function to calculate it but here it is:

For 80 tokens, the chance of having a streak of 40 no 5s is:

7.3% !!!

Which you can see is far less than what you thought, and far less than the results we are getting.

With 150 tokens the chance raises to 17.05% which it is still much less than the results of the poll.

Well, that leaves two possible reasons: people are lying/mistaken; or, pull rates fluctuate, so that if you cash in a lot during an ebb period, your chances of having a long runout are greater than if the odds were fixed at 10%.

cyineedsn

I don't even know what a solution would be anymore, I only know that I haven't gotten one since OML was released back in october(?). It's slowly getting more depressing with each pull (pulled another 10 yesterday, no dice)

GrumpySmurf1002

simonsez wrote:

Polares wrote:

I needed some time to code a function to calculate it but here it is:

For 80 tokens, the chance of having a streak of 40 no 5s is:

7.3% !!!

Which you can see is far less than what you thought, and far less than the results we are getting.

With 150 tokens the chance raises to 17.05% which it is still much less than the results of the poll.

Well, that leaves two possible reasons: people are lying/mistaken; or, pull rates fluctuate, so that if you cash in a lot during an ebb period, your chances of having a long runout are greater than if the odds were fixed at 10%.

Third reason: The population sample size is skewed towards people who are part of the forums and upset with the system.

Tatercat

Me reading this thread:


Dev's reading my post:

"Working as intended."

The Viceroy Returns

Yeah, I just opened about 15 from all the Champion rewards, and not a single 5 star.
So my dead streak is now into the 30's I think.
I guess I'm pulling back to the average of 10% now though (I've dropped down to about 13% I think, so still technically above average)?
Who knows...

SnowcaTT

I opened my sixteen saved LT's the day before championing began. I hadn't seen a 5* in at least ten, likely 20+ pulls.

Those sixteen had zero 5*'s and seven covers I had to toss (along with three more that I championed the next day, would have tossed before). Ten of sixteen basically worthless, and no 5*'s. It was...disheartening to say the least.

I keep seeing teams with something like level 150 as best maxed 4*, and twice as many 5*'s as I have (when I have maybe TEN TIMES as many 4*'s). I keep seeing teams that had lucky 5* pulls after 500 points in PVP, the only ones that can float anymore.

I said when 5*'s launched it would likely make me quit the game. It's not -quite- there yet, but it is getting there quickly.