Traditional project management literature often portrays heuristics as flawed shortcuts leading to errors, advocating for rational, debiasing strategies to prevent cost overruns and benefit shortfalls. This is problematic as heuristics can be effective. Building on Gigerenzer’s concept of fast-and-frugal heuristics, this study examines the use of such smart heuristics by senior managers in a large engineering consultancy firm during the early bid/no-bid decision-making phase of infrastructure projects. Employing a qualitative method from the naturalistic decision-making program, the research uncovers a decision strategy termed "thresholding." This strategy distills extensive experience and interpretation of ambiguous information into binary decisions, effectively de-selecting projects that could be potentially disastrous. The approach also gives credence to agency, as it only deselects disasters but keeps many potential alternatives in the portfolio to mature into potentially ‘good projects’. At the same time, it addresses Flyvbjerg’s call for some scrutiny at the front end of projects to avoid catastrophic projects that start on the wrong premises. Our chapter adds to the debate on the Hiding Hand by not being concerned with the “hidden”, but instead, with what can be known in the early fuzzy front-end of projects.