3 ms·
I wonder how much of the first experiment described is not due to causation, but is instead from certain kinds of people being both more likely to win and to ch
by ikeboy 11y ago
I wonder how much of the first experiment described is not due to causation, but is instead from certain kinds of people being both more likely to win and to cheat.
Some of the other ones do point to causation, but it would be interesting to see an experiment where people thought it was skill, but the actual winners were chosen randomly (e.g. you get points, and they tell you who wins at the end, but you don't see the actual calculation, so you won't be suspicious.)
- ScottBurson 11y agoThe first game was designed so that cheating was impossible.
- ikeboy 11y agoThat doesn't have anything to do with my point above, which is that the alternative hypothesis "people with some property X are better at game Y, and separately, people with property X are more likely to cheat" is not ruled out by experiment 1 as described. Edit: you could also differentiate between "winning->cheating" and the confounder possibility by having the cheatable game first.