Replace Mechanic Broken - PLEASE LOOK INTO


#1

Ok, so I’ve posted about this before and it seems I was wrong but also right. The replace mechanic seems to be completely knackered for me, but not in the way that I thought.

If anyone else is getting this issue, please comment here.

What seems to be happening a lot, far too much to have it happen by chance, is that whenever I replace a card, 80-90% of the time my NEXT DRAW is the card that I replaced.

Does this sound right? Am I just noticing it more when this happens? I don’t feel like I am because it seems to happen all of the damn time.


#2

Source? Did you actually keep track of the cards you’re replacing or are those just random numbers? Because without proper evidence I can’t bring myself to believe in such a crazy statistic.


#3

This very thread of course. :slight_smile:

Can’t say I never had the feeling that it happens more often than it should - especially while playing Gauntlet decks where it’s common to only have a single copy of certain cards. Definately not 90% though. The topic is a recurring one on reddit. Most common consensus is that it’s only confirmation bias, and the OP should watch all of his match-history to provide accurate numbers. People are lazy to do so.


#4

We’ve had a bunch of threads from people being convinced that the whole rng system is broken, but we’ve never had any hard numbers to go with it.


#5

It may very well be confirmation bias, but guys, I can assurre you it holds a part of truth. Maybe not 90/100 but definitely more often than it should.


#6

Since the last thread you posted I have been paying it more attention, and I can confirm that more often than not, you draw the card you replaced


#7
  1. Play matches in Practice Mode (no consequences)
  2. Replace cards and gather data (relevant volume)
  3. Post results on forums
  4. ???
  5. Profit

#8

the 4. is too lazy


#9

in 1 match i replaced sister kelano 3 times and 3 times get her from replace lul)) 1sister in deck 3 times goes to me from replace, ez loose


#10

Oh this again… I guess there’s nothing for it than to really crunch the numbers. I was on the cusp of doing it a while now, but it takes so much time.

Fine, I guess I’ll try making a highlander deck and getting it done scientifically. I won’t complete this today though.


#11

Asking for exact numbers is kinda stressful in this discussion and just another way to say: "Shut up!"
Source are hundreds of games, replays and watched streams. Anybody who plays the game regulary should know that replace isn’t completely random for shure.
As I stated in an older thread about the topic: The question is if it’s a problem the way it is:
Supposed the cards in a deck have a good reason to be in the deck and the fact if I replace one card now (not needing it now) it’s fair to assume that I need the card later in the game. So I don’t feel the games replace mechanic is a problem right know (nor “broken”).
But I don’t doubt that it’s not 100% random. And I don’t need numbers for that. I just have to play the game to see it.


#12

Studies have shown that people are completely shit at estimating what randomness means (you can’t even begin to judge the randomness of a pattern before you reach a significant sample size). Studies have also shown that people see what they are expecting to see (as already said, that’s called confirmations bias).

Gut feeling and “But it’s obvious, when you look at it” are no sources for any sort of conclusion. Besides, implementing a random number generator with one restricting clause isn’t particularly difficult programming-wise.


#13

There are some bugs that as soon as you see them, you know how they occurred. Others, you can’t tell unless you know the code. But some just make no sense at all even without seeing the code, because there’s a simple way of doing things and the bug would require doing things in a completely different way.

The problem with the theory here is that it makes no sense from a programming perspective, even as a bug. The simple, logical algorithm is to put the replaced card back in the set and then randomly draw from it as normal. They would need to have code that would assign a higher priority for randomly picking replaced cards over other cards in the deck.

I’m not saying it’s impossible, I just find it highly dubious. And on the other hand, confirmation bias is a staple of human nature.


#14

What I wanted to say: I believe in what I see and I don’t need numbers for that. (I know that some people always need numbers to feel good - most probably natural scientists for example.)

I agree whith the rest of your post. But it doesn’t disprove me.

I could ask if you have numbers to verify complete randomness.
I could ask if anybody from CP stated anytime that it is completely random.
(Edit: And I asked in my previous post - between the lines - if it should be completely ramdom.)

But I don’t do because I don’t care. For me the way it is (which way ever this is) is fine as it is.


#15

So David Copperfield can really make trains disappear?

Seeing is believing, but that doesn’t mean it’s true.


#16

The hole thread is about believing because nobody here knows the truth.


#17

The entire bloody point of confirmation bias is that you DO need numbers, unless you prefer to take the approach of “don’t confuse me with facts, I’ve made up my mind”.

On several occasions in other games I’ve seen people make various claims related to probabilities, and I’ve taken the time to collect the data, and they’ve always been wrong.


#18

I agree, though I can imagine that if the RNG they’re using isn’t entirely random, it’s possible that certain cards could get preferenced over others depending where they get “put back”. I think a reasonable comparison might be the shuffle functions in some media players. They’re “random”, yes; but they often can also end up following semi-predictable patterns where groups of songs would be played back to back consistently. If they’re calling a stock rand value, it’s very possible that it’s not as random as it could be.

Though at the same time if you have a deck of (less than) 39 cards total with 13 unique cards and you’re replacing every turn it doesn’t seem too unreasonable to get back the same card you put in once or twice per game with a perfect RNG.


#19

@qeltar, @nickdigger:

I’m no stupid - I know it’s true from a scientific point of view. But from that point of view you have to prove the thesis “it’s 100% random” with numbers, too!
And @dreamgazer hit’s the nail as I think.


#20

You know that 100% randomness can’t even be achieved without specific hardware, right?