Let's put some R in the RNG.

Mainloop25

babar3355 said:

It's funny that there is an argument in this thread suggesting that giving better drop rates to player that spend some money on the game will keep them engaged and lead to even more spending.

Wouldn't the logical conclusion then be that giving better drop rates to ALL players would encourage more spending as a community?

You won't get an argument from me there. I was merely speculating based on how many marketers think about customers, not how they should be acting. I also believe generosity gets you further than austerity. 

Corn_Noodles

Adjunct said:

so lemme get this straight. you admit to exploiting and you haven't been banned. what the hell.

10 points for Slytherin.

Corn_Noodles

Adjunct said:

lemme get this more straight. you admit to scamming and you aren't banned. seriously?

-10 points for Slytherin for trying too hard.

Steeme

Corn Noodles said:

Adjunct said:

lemme get this more straight. you admit to scamming and you aren't banned. seriously?

-10 points for Slytherin for trying too hard.

These forums are like Mortal Kombat.  FINISH HIM!

shteev

Adjunct isn't me by the way.

Volrak

Sirchombli said:

Considering the fact that cards in paper form have different frequencies amongst the same rarity (or at least used to. C1, R2 etc.) It wouldn't surprise me to find out the drop rates in this game are the same.

If mtgpq had to wrangle print sheet layouts there would be a more meaningful parallel here.

Sirchombli said:

I really feel that if there aren't different drop rates within the same rarity, the rng is broken, somehow. It seems to favor specific cards.

You wouldn't be the first!  Yet a distribution of actual cards with some degree of unbalance to it is exactly what you can expect if they were pulled with uniform random chance.

Sirchombli said:

I pulled 2 copies of glorious end from the same pack. Not the same bundle, but the same 5 card pack I got for progression in TOZ. The odds of that happening are absolutely astronomical.

One in about 20,000 5-card AKH packs would have two of the same mythic.  Impressive.  (But believe me, numbers used in astronomy are much, much, larger. :P)

Sirchombli said:

I'm thinking about starting a blog to document all of the tinykitty I encounter. Just for entertainment. Sharing these weird experiences makes them fun rather than frustrating 

I like this idea, because I like data.  Data collected carefully and representatively can actually give an answer to some of these questions, whereas anecdotes of possible things happening are just that.

Formulator

Volrak said:

You wouldn't be the first!  Yet a distribution of actual cards with some degree of unbalance to it is exactly what you can expect if they were pulled with uniform random chance.

Funny, it's also what you can expect if the odds were skewed. ¯\_(ツ)_/¯
15 copies of 1 rare while missing 13. How many copies does that have to reach before it's a red flag? I guess it would also have to happen to a lot of people too.

Edit:  I don't mean to bother you, Volrak, but I got to thinking. Is it possible for you to test the "rarity tier conspiracy theory" with the data you already have? Like,  if everybody had that kind of issue with 15 copies of one card while missing 13 others  of the same rarity, would that kinda pop out statistically?

Volrak

Formulator said:

Funny, it's also what you can expect if the odds were skewed. ¯\_(ツ)_/¯

That's exactly true.  The point is that the simple presence of skew in actual drops isn't discriminatory between two theories of random drops.

Formulator said:

15 copies of 1 rare while missing 13. How many copies does that have to reach before it's a red flag?

If people who experience this post screenshots of their rare collection I can make some comments.

Formulator said:

Is it possible for you to test the "rarity tier conspiracy theory" with the data you already have? Like,  if everybody had that kind of issue with 15 copies of one card while missing 13 others  of the same rarity, would that kinda pop out statistically?

This is a great question.  It's possible, but what's needed is data.  For example, Marvaddin's card logging thread is a great example of someone carefully recording useful dupe data.  I think there will eventually be enough data just from that thread to do some useful analysis (at least for commons!).

A similar but slightly different analysis was done in April using anonymous collection data that had been logged at mtgpq.info.  Rather than consider dupe rates directly (which weren't available), it looked at how skewed a whole population of player collections could be expected to be under uniform random and two "tiered" random models.  The tiered models assume that each card is placed into fixed tiers within a rarity, with those tiers being consistent for all players.  The best fit for the actual data out of those models was uniform random.