This paper extends the Markov Riding Position Sampler (MaRiPoSa) algorithm to analyze fairness across multiple Canadian federal elections (2011–2025). While previous work demonstrated that alternative, allowable boundary configurations could produce different governments from the same vote distribution in a single election, this analysis examines whether such boundary-dependent outcomes represent systematic patterns or isolated anomalies. We evaluate electoral fairness through three complementary frameworks: (1) partisan fairness and reciprocity, testing whether parties would receive equivalent seat shares if vote levels were exchanged; (2) expected outcomes and typicality, comparing observed results against the distribution of possible outcomes; and (3) proportionality, measuring deviation from proportional representation. These results raise questions about democratic legitimacy when the same voters, casting the same votes, in the same locations, produce different governments depending solely on valid boundary configurations. Canada’s electoral integrity requires computational tools capable of handling multiparty systems, complex political geographies, and regional normative contexts to inform evidence-based discussions of electoral reform.