Jump to content

Fortune Box Research Part 2 (to 100 items) ... No significant change


Recommended Posts

BLUF: We ran the fortune box research out to 120 boxes from the previous 40 (see previous study) and found that chance of rare is 63.0%, common 5.3%, Legendary 3.7%, and Epic 28.0%.  The difference in calculated legendary rate increased 0.3%, meaning there was no significant (i.e. 5% being the cut-off) difference from the last study.  (Ignore the 60.5)

image.thumb.png.1f4f21a0c49e13f603b6ea3d163e68f6.png

LONG STORY: 

First, thanks to everyone who helped me out: @dirtycarot @epicspeedster@Fordekash Zbey Kse, @ZeRo_ @Sleckmo @diego brando  @GGoT

     So I ran this study out to 120 (was aiming for 100) because people didn't believe me when I said there was no significant difference between 40 and 100+.  So here you go: here's proof there's no significant difference.  If you really base your arena shop upgrades on whether the legendary rate is 3.3% vs 3.7% then you might need to seek professional help 😄 😛 

     I tried to make the colors in the chart match the game's color coding, but I was not able to edit them individually (gotta love Excel).  If you want to pay me to buy a student copy of IBM SPSS, so that the colors match, please let me know.

     The findings remain the same: for most people, it will not be advantageous to upgrade your fortune box rate in the arena shop if you want more legendaries: not until you upgrade the other important things like damage, capacities, hit points, resistance, backfire reduction, etc.

     However, I will also say that while the fortune box might as well be called the "Blue Box" because of rares being the majority (63%), I will say that this is at least reasonably generous of the game to give you legendaries.  And in this data, one of our members got a Bright Roar.  Another got an Energy Monkey torso.  I got an MPV from a fortune box once (but not while this study was going on).

     There are some limitations to this study, though, as always:

  • It appears the game bases your rates off of the code but using a RNG call to your host machine.  Therefore, if you happen to somehow be on a computer that has a broken RNG, you could get potentially higher rates or lower rates than these.
  • To help you understand the measure of significance: it's mostly agreed upon by Statisticians that 1% is the level of significance required for difference to be important, but only if you are doing things within the hard sciences.  For example, if we were manufacturing gun barrels and machine A's differences in finished product width was a 1% spread versus machine B was a 0.1% spread, this would be cause for concern because we might accidentally manufacture barrels that are out of specifications and thus cause accuracy problems on machine A, but machine B would be better for our needs.  For things that use probability and normal curves (human behavior experiments, things based in chance), the measure is 5%.  So if there had been a 5% difference between study A and study B, we would conclude that 40 is not enough data points.  However, what we see is 0.3%, which is less than both a strict hard-science 1% margin and a more soft-science psychology and probability 5%.  Therefore our data is very reliable.
  • More data points will only make the rates more accurate, but should not (as we have demonstrated) change your conclusions.
  • There appears to be no difference in who was providing the screenshots in terms of player rank.  There also seemed to be no difference in whether the fortune box was from the shop or from farming the campaigns.  Thus I believe it is safe to conclude that these stats apply to all fortune boxes.  But if you have a good reason to believe shop versus campaign is different, please let me know.
  • Keep in mind some times when you complete parts of the campaign, the fortune box animation is given, even though normally the resulting rewards are fixed.  It is unclear yet if there are times when completing the campaign will result in a higher drop rate for special circumstances.  However, we were not calculating fortune boxes for special circumstances (because that would take way too long).
  • The rates are PER ITEM IN THE BOX, not the entire box.  This is because the items in the box are not always identical.  You can sometimes get a box result (as we have shown in the pictures, which is available upon request) where you get one epic and one common.  It just happens.  Because fortune boxes seem to be "per item" chances, but seem to light up the light on the animation based on the highest rarity of item you receive (i.e. lights "legendary" when you get only one legendary, etc.), I thought this would be a more logical and simple way to run the numbers.  Your mileage may vary.

By the way, as soon as I finished, my alt got this in a fortune box

image.png.1f1ecf4719ab200c6836c75cca3122bd.png

Edited by SawzAll (see edit history)

"Play stupid games, win stupid prizes."

http://www.puresimplicity.net/~oneeyedcat/misc/supermechs.html

Link to comment
Share on other sites

wow ur awesome. Also take note of this that daily fortuneboxes have a different droprate pattern than regular ones. i usually get legendaries from them every 3 free fortune box in the shop. So 1/3 procent chance in getting a legendary from them. for me at least.

Link to comment
Share on other sites

Beautiful work and presentation. However, I think one of the most important questions to ask is:
'What is the drop-rate chance of fortune boxes from mission, in particular boss missions?'

Link to comment
Share on other sites

3 hours ago, Andypandy said:

wow ur awesome. Also take note of this that daily fortuneboxes have a different droprate pattern than regular ones

Doesn't seem to match the data my clan pulled.

2 hours ago, AftoKrator said:

i honestly can't remember the last time i got a legendary from any free fortune box, so it's probably a much lower chance for most other people

actually, how it works is that sometimes we get unlucky.  It's the chance as a whole that this measures. 

3 minutes ago, Marethryu said:

Beautiful work and presentation. However, I think one of the most important questions to ask is:
'What is the drop-rate chance of fortune boxes from mission, in particular boss missions?'

I don't think there is a difference.  But the problem here is that to fully test this would require 40 FB results from every campaign and mission that gives fortune boxes.  So that's like at least 12 x 40 = 480 data points.  I really don't feel like testing that.

"Play stupid games, win stupid prizes."

http://www.puresimplicity.net/~oneeyedcat/misc/supermechs.html

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...