Strong Opinions Loosely Held

Two traits held at once

The phrase predates the founders who use it as a screen (it is usually attributed to the forecaster Paul Saffo and was popularized broadly in technology culture), but Brendan Foody offers one of the clearest operational definitions in describing what he looks for at Mercor. He names the combination as a top culture signal: "They have immense conviction in whatever idea they're telling to me, but when they're presented with an idea that is counter to that or data or evidence against it, they change their mind very quickly."¹ [11:51]

By Foody's account the signal is two distinct properties held simultaneously. The first is strong opinions: high conviction, a willingness to take a position and push for it under uncertainty rather than wait for consensus. The second is a loose grip: the conviction attaches to the best available model, not to being right, so a better argument or new data triggers a genuine change of mind rather than a grudging partial concession.

The failure modes on each axis

The reason the pairing works as a filter, in Foody's framing, is that each trait alone produces a recognizable failure mode. Strong opinions tightly held is overconfidence: a person who is decisive but wrong and cannot be corrected, which he treats as corrosive in a company whose environment changes faster than any fixed belief system. Weak opinions loosely held is the opposite failure, epistemic cowardice: someone who updates fast but never commits, is therefore never wrong, and produces no useful signal in a debate because they will not own a call.

The combination is what he wants because it yields people who are both useful in decision-making, since they hold a view and can quarterback an initiative, and epistemically healthy, since they do not tie their identity to being right and so let the organization's collective beliefs track reality.

Conviction needs an outlet

Foody attaches the signal to a note on decision rights: "Sometimes it's their call, sometimes it's mine, but always trying to have a very thoughtful way of arriving at that answer."¹ The dependency runs one direction. Strong opinions loosely held presupposes a culture where people are actually trusted to make calls, because without that trust conviction has no outlet and there is no live decision against which to test whether a belief was held loosely enough.

Foody positions the trait as complementary to other hiring signals rather than a replacement for them. Where a care-based screen measures the magnitude of a candidate's investment in the work, this one measures the quality of their reasoning about it, and he frames the ideal hire as someone who scores on both. He treats it as a behavioral marker for talent that accelerates a team rather than dragging it, which is how it connects to the broader A-Player Framework: strong performers tend to hold both traits, while weaker ones tend to collapse into either overconfidence or an unwillingness to commit.