We distinguish empirical, personal, and institutional warrantsWarrant. A body of observed instances backs empirical warrantsWarrant. Personal acquaintance backs personal warrantsWarrant, especially concerning personal dispositions. The rules of an institution back institutional warrantsWarrant. An exception to an empirical warrantWarrant instances a relevant variableRelevant variable. We can determine the strength of an empirical warrantWarrant through observation which discloses the relevant variablesRelevant variable and the warrant’sWarrant resistance to them. The method of relevant variablesRelevant variable to assess argument strength is radically different from Bayesian inference. HahnHahn, Ulrike and Oaksford adopt a subjectivist interpretation. PremisesPremise and conclusions reflect subjective degrees of belief. We can measure the strength of an argument by measuring the strength of belief of a recipient before and after being exposed to the argument. But does degree of epistmic justification, determined by such factors as simplicity, consistency with previous results, or the comparative reliabilityReliability of the sources vouching for a belief correspond to strength of belief? SalmonSalmon, Wesley asks us to consider a class of hypotheses ℋ concerning some subject matter. Where \( \mathcal{A} \) ⊆ ℋ, \( \mathcal{A} \) is the subclass where we can say something about the relative frequency of true hypotheses, Pr( \( \mathcal{T}/\mathcal{A} \) ). SalmonSalmon, Wesley understands Pr( \( \mathcal{T}/\mathcal{A} \) ) as the plausibilityPlausibility of ℋ. But why should plausibilityPlausibility be connected to PascalianProbabilityPascalian probability?