Computational catalysts screening is an increasingly popular technique, in which the mechanism from a known good catalyst is commonly adopted, parameterized from linear scaling relationships, and then used in a microkinetic model to identify other metal alloys with incrementally improved activity. This strategy, however, fails to identify truly novel catalysts that operate under nontraditional reaction conditions and exhibit alternative dominant reaction pathways. Using methane oxidation and reforming we investigated a series of O* and OH*‐assisted C‐H scission and C‐O bond formation pathways. Notably, for methane oxidation we discovered a second local optimum for O*‐assisted C‐H bond activation near Ag, which is inactive if only the direct C‐H scission route is considered. In light of the significant qualitative difference in the predicted catalytic trends when parallel mechanisms are allowed, we propose a minimum barrier assumption to rapidly screen for potentially important alternative pathways without the need for costly density functional theory simulations. © 2016 American Institute of Chemical Engineers AIChE J, 63: 66–77, 2017