The expected utility formalism combines a probability distribution relating actions to their probable consequences, with a utility function over consequences of actions, in order to assign expected utilities to actions. A key issue is the form and derivation of the probability distribution relating actions to consequences, usually held to be a counterfactual probability distribution.
In the Mathjax bit of the Allais Paradox section, the sign on the second line shows as being backward (greater-than sign on both lines):
https://ibb.co/55xvtg5
If I look at the source (by proposing an edit) it looks as if it ought to be correct, but I’m not familiar with Mathjax so I’m not sure what the issue is.