I feel like every one of these should have a "reverse" effect to correct the bias, for the reasons people believe them are sometimes closely related and if a few things were modified the belief would turn out to be right.
For example the gambler's fallacy is very closely related to bayesian inference and I suspect that's why a lot of people confuse the two. It'd be very helpful if that was pointed out next to each box.
It'd be also interesting to know which biases are fixable and which ones are not. I suspect placebo effect is not easily fixable whereas stereotyping can be avoided after some amount of learning.