Talk:Showdown Update/@comment-36912561-20190701175437/@comment-34918040-20190704004124

@ SunnyTomato:

Sorry for being a bit late with my answer, I didn't have a look at this page for a while. First:

Facts and math don't care if you believe in them or not. They exist.

It's up to you: Either you can accept that the "always have some cards you need in your inventory" conjecture has been proven wrong and act accordingly, or you can continue playing your own way, with the risk of not getting what you want and being disappointed.

My statistics project is an invitation, nothing more. But I can say that I follow these rules, too, and they work. In a bit more than one year, I got over 200 cars and to Champion League, I can't complain.

Besides, as a consequence of my advancing quite fast, I'm nearly always short on resources. I can't even enumerate all the cases I got cards I shouldn't have obtained if the "golden rule" was true. I always get the cards I want, whether I have them in my inventory or not. But as I don't claim something on my own, possibly biased experience, I tested it systematically with the experiment behind the link above.

Same for all other busted myths. I think of an experiment that either verifies or refutes a claim and then I have a look at the real numbers from a matter-of-fact point of view. I don't care if my personal assumption is proven wrong, as I always profit from knowing the real mechanisms.

I've mostly asked my friend before publishing a mythbuster or introducing a new statistical feature (she's a psychologist and a badass statistician), but there can still be flaws. If someone finds mistakes, go ahead! It's a wiki, it lives from people contributing and making the system more precise.

So it's not a matter of being friends or something. Statistics is a valid method to reveal GL's dirty little tricks, and we've become quite fast at doing so. Feel free to take into account the statistics tables or decide to do it your way.

@ Stupidphd:
 * " So we should keep a healthy inventory of cards you need. I agree."

You can do so, but it won't help.


 * "The B drivetrain has been a crisis, 0% from B part boxes."

This is actually an example that your protests helped. Compare (the old version with no drivetrain) and  (the new version with a more balanced distribution after June 28).

@ SunnyTomato:


 * "[...] but the stats don’t paint the full picture."

I think a total sample size of 15,303 cards is quite representative. With the help of all data contributors, I might add. A big thanks to them!


 * "[...] what thousands of other players are stating!"

And they were right concerning the. That's what the statistics project is for: Reaveal unfair processes, but also prove claims wrong that don't stand a thorough investigation.


 * "as I was never good at maths"

That's why I try to avoid formulas in background information articles like drop rate or law of large numbers, but do provide them in other in-depth articles for players interested in the mathematical background.

@ Nianasto:


 * "If I have two slices of bread and you have none then we have on average a slice of bread each."

Right. That's why the results of one single opened box can deviate significantly from the average.


 * "Statistics is the way that politicians use to lie."

Right. That's why I try to be as transparent as possible and provide the exact way I proceed together with the data used for every experiment. Unfortunately, I'm not able to share the whole database with you at the moment. I currently have all this in a huge Excel file which does the statistical evaluation. Making the database accessible for everyone would mean to have an SPSS or R installation running on a server—this would cost money. As far as I know it's not possible to have an R instance running on a Fandom server, but I'm already thinking of ways to make the database accessible.


 * "Each one's 'small' frame of the whole picture IS his full picture."

Let's put it this way: The bigger the "small" frame gets, the more it will resemble the full picture. See Law of large numbers. The results of only a few opened boxes may deviate from official drop rates, but in the end the results will be the same as the official values.

---

Again, thanks to all of you for your feedback and data contributions.

By the way, I've started to roll out a new feature that makes deviations of box content more transparent. You'll notice a new "Card distribution per box" tab on many Pro Kit Box pages. More details in a separate post.