Bonus Hero Drops

OneLungDave

I've seen a statistically lower drop rate for bonus heroes in the last three weeks. I only recently started tracking the actual data; which equates to 1 bonus 3* in 36 possible pulls, 0 bonus 4* in 14 possible pulls, and 0 bonus 5* in my last 10 legendary tokens that actually bestowed a 5*.

Is there a correlation to the amount of bonus heroes you have chosen? Am I just improbably unlucky? Has there been a change to the bonus structure that I might have missed? 

Any news or info would be appreciated. 

Jaedenkaal

Actually, your luck isn't even that improbable. 5% is simply not that large of a chance. You'd have to pull thousands or tens of thousands of tokens before your running average would be reasonably expected to approximate 5%

Also keep in mind that the chance of getting a bonus hero from a standard token is less than 5%

OneLungDave

The law of averages isn't useful in a set percentage of data, though. 5%=5% wether you open 20 tokens or 20000 tokens. I understand that 5% is a very small probability, and that one bonus in 60 possible chances is ALMOST 5%, and I understand that this is a miniscule complaint. I was just wondering if anyone else noticed a significant decline in their bonus rewards. Or if there is a correlation to the number of favorites you have chosen to the number of bonuses one receives. 

It's almost as tho they removed/altered the bonuses. 

And thus concludes my whining. 

Jaedenkaal

well I got a 4* bh followed immediately by a 3* bh just yesterday. Should I be asking if they raised the odds?

or do both of us simply have a sample size too small to draw any meaningful conclusions about BH rate?

OneLungDave

Considering you pulled two bonus heroes in two draws, I think it'd be safe to conclude that they have INDEED raised the odds. 2/2=100%. And, let it be known, that I'm immensely jealous of your luck. 

Don't mistake this as patronizing sass. I'm looking for information. I'm now thru 80 potential token draws with only 1 bonus. If there's anything I can do to increase the odds, or if they've done something different to the odds, I'd like to know. 

TPF Alexis

Assuming I'm using this calculator right, with a 5% chance per pull, on 80 trials, there's about a 7% chance of only getting one Bonus Hero. In other words, if we had 100 Players, and they each pulled 80 tokens, we could expect about seven of them to end up with just a single Bonus Hero each.

That's actually a pretty major simplification, because probabilities this small need (IIRC) something like thousands to tens of thousands of data points before they really start to look like the theoretical distribution. But it makes the point, which is that you're actually slightly more likely to go 80 pulls with only a single Bonus Hero than you are to get a Bonus Hero off any given pull.

Very short version: Statistics and probability are weird, nothing is broken, don't worry about it.

Starfury

Statistics are not weird, people just forget that "the most likely outcome" is not the same thing as "a likely outcome". The combined likelyhood of all other possible outcomes usually far outweighs the "expected" outcome.

To illustrate, I just ran a simulation with 100000 users pulling 80 times each:
Only 20030 users ended up with the expected 4 bonus heroes.
Almost as many (19614) users pulled 3 bonus heroes, and 23041 users pulled 2 or less users.
On the other end of the spectrum were 4609 users who pulled 8 or more bonus heroes. (i.e. twice the expected number)
Overall, there were 400236 bonus hero pulls, so large numbers won (as always)

Takeaway: 4 bonus heroes is indeed the single most likely outcome. But you still have an 80% chance of getting something else.

TPF Alexis

By "statistics is weird", I just mean that randomness often leads to results that are counterintuitive to people who don't have a lot of experience with it. Like, yes, the fact that the single most likely outcome often still has a relatively small chance of coming up if there are enough total possible outcomes.

Dayv

And, of course, when you're unhappy about a set of results, you're going to dwell on it, and start proving to yourself how unlikely those results were.  Unlikely or not, they happened, and when you have a hot streak you'll usually be too busy enjoying the rewards to do the math and prove it "shouldn't" have happened either.