In virtually all theoretical, experimental, and practical uses of the Borda rule, inputs are orders (complete and transitive rankings). This note asks whether elicitation should be matched to the rule, and tests Borda with two inputs: (i) Orders and (ii) Tournaments (all pairwise comparisons). In a preregistered online experiment (N=366, forming 61 societies) with full counterbalancing, each participant completed both formats on disjoint medium-size stimulus sets; societies outcomes were obtained by Borda aggregation. Contrary to the field's convention, Tournament cuts individual error rates by about 29.8% and lowers the mean weighted Kemeny distance between society rankings and members' true orders by about 8.2%. We conclude that, pairwise elicitation is superior to order elicitation as input to the Borda rule, at least for medium-sized option sets. Methodologically, we introduce a simple and generalizable procedure for testing elicitation-rule alignment, enabling analogous evaluations for other social choice rules.