Incomplete induction generalizes from examined cases while leaving unexamined cases. The traditional text nevertheless lists certain forms of established experience among the sources of demonstrative certainty.
In that framework, repetition is not treated merely as counting events; the intellect also recognizes a dependable connection under the relevant conditions.
In evaluating an actual claim, limited observations, varying conditions, or counterexamples weaken any assertion of necessity. Repeated occurrence should not be equated mechanically with warranted universal certainty.