Let’s say you flip a coin three thousand times. If it comes up heads two thousand times, I think it’s safe to say that people will assume this is a rigged coin in spite of the fact that each outcome is equally likely. Closer to the middle, the line gets less certain. Some people would consider 1550/1450 a genuine outcome whereas others wouldn’t. Is there research that has attempted to measure the crossover point here?
I’m looking for some study that has presented people with “random” data, such as coin flips or die rolls, and asked them if they think it was “rigged” or not.