I would like to test how different odds influence betting behavior. After the market closes, I will call @FairlyRandom to generate a number from 1 to 100. If the number is between 1 and 52, this market resolves to YES; otherwise, it resolves to NO.
I will also record betting behavior for this and related markets with varying odds and analyze the data statistically to look for patterns in participants’ decisions.
1,000People are also trading
I hypothesise that the systematic bias for these will be in the direction of 50% (e.g. towards NO here because 52%>50%) because irrational bettors will likely prefer a lower prob of winning bigger (exciting gamble) over a bigger prob of winning smaller (boring hedge), even though they both have the same expected value
@Velaris It's partly exploratory, but one thing I want to measure is how people react to slightly uneven odds. By running many markets with controlled probabilities and tracking betting patterns, I can estimate how much participants overbet or underbet relative to the true odds. One goal is to quantify systematic biases.
@4fa You won't see the same systematic biases in cases where the true odds are so obviously known though -- it has been done many times before and people will not bet away from the true price by more than a tiny amount unless they are trolling (which will be quickly corrected). See the daily coin flip markets for example.
@A Several people already bet away from the true price in the 51% market by more than a tiny amount. Not sure I would call that trolling rather than simply calling it bad bets.