Daniels is more believable both in general (she has made fewer demonstrably false public statements than Trump) and in particular (her actions and statements in this particular case are less suspicious than Trump’s actions and statements in this particular case). As a general rule: anyone who is neither generally nor particularly untrustworthy is more believable than Trump (who is both untrustworthy in general and is untrustworthy with regard to just about all particulars that he chooses to discuss publicly). To be clear, that’s just a baseline. Even Trump deserves case-by-case evaluation. But if the available information requires a framing tool for proper assessment, the general rule is perfectly serviceable for generating a prior probability distribution. For cases where Trump is in dispute with someone who is either generally or particularly untrustworthy, no general rule is available and extra special attention must be paid to the specifics. It must be admitted, however, that Trump tends to lose those matchups as well.