Misty is definitely NOT 50/50
Awhile ago, I felt like the first flip of Misty was flipping tails way more than heads.
So I decided to start counting.
I am at 150 flips so far. 97 tails. 53 Heads
If I count the flips beyond the first, it's +28 more heads
Based on my small sample, the first flip seems to actually have 70/30 odds tails vs heads, while the subsequent flips have 40/60 odds.
This would explain the 4-7 heads streaks that seem to happen at an uncanny rate when the first flip DOES succeed.
128 flips isn't enough to not discount this as just bad luck. So is anyone else counting?
Comments
-
Im not counting, but its obvious. Everyone knows that Head is less often than Tails, not just Misty but in general.
1 -
I think most people had the same feeling, but we must also take into account how easy it is to forget the times in which you nuked a match with heads on Misty while remembering the times in which Misty was just a wasted deck slot and trainer.
Personally, I agree that Misty can be swapped for something more consistent in water decks, especially now that Manaphy is around. She'll most likely end the match most of the times in which you flip heads, but in the huge amount of times in which you'll flip tails she'll just be a wasted card.
0 -
That means nothing, that's not how probabilities work...
A 50% chance means an equal probability of something happening or not. Out of two possible outcomes, either can happen with equal likelihood so it's impossible to predict in favour of any of them.
So even if all players flip Mist coins 10k times, there is a chance that every single one of them get always tails, not likely but possible.
Your counting or anyone else counting means nothing.
0 -
We would need a larger sample size than just one player. This doesn't necessarily indicate a problem with the randomness by itself.
Also, humans are very bad at recognising actual randomness. True randomness can result in weird clusters of data points that do not appear random to the casual human observer. Ironically what would look random to a human would be an algorithm that resulted in a forced even distribution.
Tossing a coin 1000 times and getting 1000 tails does not feel random to a human, but it is just as statistically likely as any other specific result.
0 -
I'm getting tired of people who act like they understand probability better than they really do.
Let's say we test for how likely it is to flip three coins and get all three coins to flip heads.
All of the following have evenly distributed probability
Tails Tails Tails
Tails Tails Heads
Tails Heads Tails
Tails Heads Heads
Heads Tails Heads
Heads Tails Tails
Heads Heads Tails
Heads Heads Heads
Two fallacies are being committed by some here. First, not acknowledging that multiple instances of the same count can happen (Like Tails Tails Heads and Heads Tails Tails). Second is suggesting that one of these probabilities is just as likely as ALL of the others, instead of EACH of the others. In reality, each likelyhood is 1 in 8, NOT 1 in 2.
So when I flipped 150 coins, the odds of getting what I got is 1 in 151.
Now as I said, while very unlucky, my anecdote isn't truly indicative by itself.
HOWEVER, if the sample size is large enough, the indication can become undeniably uncanny. This plays into what is called "The Law of Averages".
For instance, if 10 other people counted 150 coin flips and ALL TEN experienced more tails than heads, that is a pretty strong indication, whether the difference is +1 or +150.
Suggesting that a sample size of 10,000 flips wouldn't indicate anything regardless of result is just... anyway, if 7,000 of those flips were tails and only 2% of them had more heads than tails, you'd have to consider which is more likely, that the odds are actually 70/30 or that 98 out of 100 people were just really unlucky.
1 -
This. Thank you. A man of culture.
I would be really interested to see how the game generates the coin flips. Even better, the game should track and show how many coin flips youve made, and the results of those flips. Super easy to track, even easier to display.
Equally interesting would be getting a phone emulator and save stating the game to see just how random these flips truly are. My guess is not at all.
I would speculate the game predetermines and pools the first ~50 coin flips at the start of the march and then pulls from that pool in order, and the graphic of the coin flipping has nothing to do with the result. I soeculate this because object pooling is, in general, basic and good game design.
Ive noticed many times on multi-flips coins beyond the first appear to be landing on tails or heads and then somehow invert to the other side. If you pay close attention its very easy to notice and then very hard not to notice every time after.
I havent personally kept track of my tails/heads, but i should start, and i would bet money i don't have that across all coin flips I have flipped more tails then heads and that the coin flip is not 50/502 -
I have to imagine that "for the sake of game balance" it's actually not even coded as a 50/50. That's why it only says if it lands on heads, it doesn't mention what the chance will be that it happens. (obviously in the real-world we assume it to be 50%, but that doesn't have to be true for a digital coin). I could absolutely see them doing something like that internally since energy can turn the tide of battle so easily. Though I'd much rather have a misty that has a true 50% chance but can only add like, 2 energy, as a better way to balance it…
It may also be that, due to the massive amount of random values the server has to be constantly calculating for pack openings and battles, they may be using a high-performance Random value algorithm that doesn't have a very uniform result distribution, but is "close enough". So it likely trends a lot more towards a 0 result.0
