Fair and unbiased coins have a bias to land on the same side they started

[Pmalek]

hmm if he did the exact test again and got 51-49 again it is a lot more likely there is a bias.

It's worth pointing out that this test had 48 participants who participated in the coin tossing. It wasn't only one person. They all had varying degrees of bias for the same side, and about a dozen of them showed a bias to the opposite site. The 51% chance that the coin will land on the same side is an average taken from all participants and all tosses.

[1714794361]

1714794361

[1714794361]

1714794361

[PrivacyG]

So couldn't it be just a coincidence even though the flips where that many?

I also wonder the same.  In theory the more you flip the closer to 50 percent you will get.  This is very reminding of Gambling where you think 50 percent chance of win means out of 10 rolls you will win 5 and lose 5.  But it is very possible to win 10 in a row just as much as none.

What if after all these 300+ thousand flips they try 10 thousand times more and 9 out of 10 thousand flips the opposite side lands only.  It changes the outcome.  Now the 10 thousand flips had 90 percent bias.

Interesting nonetheless.  Statistics are EXTREMELY annoying to me.  I would say, why not ask the 48 people at the end whether they thought about manipulating the results.  But then what if some of them lie.  What if all of them lie!  Then why not blindfold them, put a coin inside a cone, hand them the cone and let them throw.  But how random is the movement of their hands really?  What if the way some body throws a coin marks the actual landing.  UGH!

Could these findings lower the security of seeds generated from coin tosses, considering that the starting positions of the coins aren't known to a third party?

I have no idea!  How do you measure security of random?  We are humans after all and even if we TRY flipping a coin 100 times the same way we will not do it the EXACT same every time.  So there is some sort of random in that.  Even if you try to manipulate.

A manipulator could fail at manipulating the toss.  What if the manipulators actually landed more randomly than the rest.  How do you measure?!

My mind says DON'T TRUST HUMANS AT RANDOM GENERATION but how random is random really.  I had a few happenings where I generated a random number of 5 to 10 digits and it generated me a number like 0011110.  It was maybe random.  But putting this into Security, how secure is the PIN 0011110 compared to the PIN 3749201?  How random was the PIN I just 'randomly' pressed on my keyboard though?  Does it make my PIN less secure than the 0011110 randomly generated PIN?

The more I think of this the more my brain goes, WHAT THE FUCK!  The more you get to a perfect statistic the better you realize there is no such thing.  Closest you can get is a 'more perfect' statistic.  But is it really?

[Pmalek]

I would say, why not ask the 48 people at the end whether they thought about manipulating the results.  But then what if some of them lie.  What if all of them lie!

Perhaps the subjects shouldn't have known what the purpose of the test was. But they began the experiment knowing what they are trying to prove. The writers of the research paper mentioned that concern as well, and I agree with them.  

My mind says DON'T TRUST HUMANS AT RANDOM GENERATION but how random is random really.  I had a few happenings where I generated a random number of 5 to 10 digits and it generated me a number like 0011110.  It was maybe random.  But putting this into Security, how secure is the PIN 0011110 compared to the PIN 3749201?  How random was the PIN I just 'randomly' pressed on my keyboard though?  Does it make my PIN less secure than the 0011110 randomly generated PIN?

I think the 3749201 PIN is more secure than 0011110 simply because it uses a greater variety of digits. In the first PIN, you only have zero's and one's. In the second one, you have seven different numbers. My layman's conclusion is that the first PIN would be easier to bruteforce than the second one.