Stones' colour distribution is NOT random - MASTER Board Conspiracy

If you were to take 5 red heroes and receive as many red tiles as usual then sometimes you would win easily (because sometimes you receive a lot of red tiles), sometimes you would win with difficulty (because sometimes you receive only a moderate proportion of red tiles) and sometimes you would lose (because sometimes you receive few red tiles).

Do you find that you experience something different? Please post your data so we can all share it.

2 Likes

No, but I experience a lower average dropping rate of a color, if multiple heroes of this color are in the team.

It’s more felt than proven, but it is a possibility to balance chances.

You can believe that if it makes you happy to, but there’s actually quite good evidence that you get the same drop rate of each color of tile no matter how many heroes of that color you have in your team.

1 Like

:popcorn::popcorn::popcorn:

( checks message count )

Also taking side bets on first mention of nazis ( linky, linky )

3 Likes

I think we’re done here.

7 Likes

My latest favorite :smile:

I did end up winning but that was a first.

3 Likes

Nice! No tiles of a color is pretty rare. Should only happen once in every 500 board or so.

1 Like

Very strange defense set up your opponent is running there.

Maybe they think Wu stacks? Would not be my first choice for a D…

:grin::grinning::grinning::grinning::grinning:. … Alright we just take whatever we are told

It is carefully random to exclude the stacked color…not new…mostly at the starting boards.

So @Oliz, you somehow believe that the overwhelming statistical evidence from the hundreds of boards of data that have been systematically collected in this thread and over in the Color Stacking Fairness Project thread are just wrong? Or that somehow the programmers know when you’re collecting data and only give fair boards then? Or what?

3 Likes

Not sure if anyone has brought this up and I would love to see what the devs say…

My theory, and it’s just a theory:
The algorithm has a percentage attached to the randomness. Each color, for example has a 20% chance to spawn. Now prior to that spawn, when the board is created the 20% goes up or down based on the presence of a particular color.
For example, say you run a full house of 3 blues and 2 reds. That means that purple, yellow, and green go up in percentage while the blue and red go down. I don’t know, say its 1% for each.
This puts the rng at:
Blue - 18%
Red - 19%
yellow, green, and purple - 21%
By doing this, you can still get a decent board through rng, and if your heroes are strong enough, you can outlast your opponent through attrition, but an even match you would be outgunned by sheer statistics in that case.

1 Like

I will go further and thats what i mean all the time. The boards are very much equal over a 100 games that doesnt mean it is random. Random means the is no algorythm behind. Here it is clearly.

We all need to change our team in raids and the game recognize it and give us the color we dont need. in the next game we have another team with other color and the game give us the tiles we dont need but all colors are now equl in the destribution again.

The 100 boards with the same color are equal because the game make it equal at some point… But who want to stick with the same color?

How this theory explains challenge events when top players (not only top actually) go with mono color teams to minimize time to pass a level? According to your statement color of a team must be always less available then others. Which is obviously not true.

2 Likes

In a computer game you always deal with an algorithm. Even a Random Number Generator (actually a Pseudo Random Numbers generator) is an algorithm.

That’s a conspiracy. Especially taking in account that it equally gives us boards that we do need or we completely don’t need while 80% of boards are so-so.

2 Likes

This theory is testable.

Suppose you first ran 100 raids with a rainbow team, counting the tiles on the opening boards. Per your theory, we would see 20% of each color, within sampling error.

Then set up an unbalanced team and do another 100 raids, counting the tiles on the opening boards. Per your theory, we should see a bias away from the stacked color.

That’s exactly the tests run upthread and in the Color Fairness thread linked above. No bias was found, disproving your theory.

3 Likes

So you mean it cant be random distribution?? interesting…

Strictly speaking, that’s true. The Pseudo-Random Number Generators that are commonly used aren’t genuinely random. Good ones, though, will pass all the standard tests for ramdomness. Moreover, since there are tens of thousands of players drawing from the PRNG in sequence and genuinely randomly, even a mediocre PRNG is going to produce results that are effectively random for any given player.

To suppose that SGG has spent the programming resources to craft a devious algorithm to subtly distort the RNG results simply defies any business sense. Why would a company spend extra time/money to make customers unhappy?

This thread just keeps looping back on itself.

  1. New player observes some bad boards, draws sweeping inference
  2. Said player writes a post here, usually presenting three or four screenshots as “proof”
  3. Old hands point out that the statistical evidence contradicts the “fix is on” hypothesis
  4. Said player denies relevance of proof. “My situation is different”
  5. Old hands point out basics of psychology, confirmation bias, selective recall
6 Likes

We have no evidence to suggest anything like that is happening, there is no incentive for developers to build any such thing into the code (though there are incentives for them not to do so), it’s needlessly elaborate, and the developers have already stated unambiguously that boards are random.

So yeah, I’m not seeing anything of value in that theory.

  1. No, we don’t, but your comment tells us that you are colour stacking, and so are making yourself vulnerable to losing when you get disadvantageous boards. This might be making you annoyed and a little paranoid.
  2. Some players have kept a stacked team unchanged and collected a large sample of boards, avoiding the problem you imagine. They have found no evidence of board bias.
  3. Other players have changed their stacked colour, but logged occurrences of the stacked colour compared to unstacked (and omitted) colours. This also avoids the problem you imagine, and they also have found no evidence of board bias.

:arrow_up::arrow_up::arrow_up::arrow_up::arrow_up::arrow_up: This, this, this. :arrow_up::arrow_up::arrow_up::arrow_up::arrow_up::arrow_up::arrow_up:

3 Likes