Netflix Algorithms Reward Velocity Over Coherence Since 2021

Anya Taylor-Joy sprints across Nevada in a silk blouson, pursued by federal agents who are themselves pursued by CIA operatives, who are themselves being deceived by a tech billionaire who may or may not exist.

None of this makes sense—and Lucky, which premiered on Hulu in June, knows it. The review cycle knows it, everyone involved knows it, and the consensus remains untroubled. It's summer entertainment, it's preposterous, and it's irresistible.

The unstated move is subtle but total. By naming something bunkum, the reviewer grants it immunity from the standards used to evaluate everything else. The acknowledgment becomes the excuse—"It's absurd but entertaining" functions as a closed loop, as if naming the problem exempts you from living inside its consequences.

Twenty years ago, asking a viewer to accept seventeen contradictions in a ten-episode arc meant the show had failed. The machinery broke and you noticed. Now the contradiction isn't a flaw to overcome—it's something to celebrate, as long as the viewer clicks Next Episode. The shift isn't in taste, it's in incentive.

A show that holds you for eight hours generates the same algorithmic signal as a show that satisfies you. The industrial optimization is invisible, but the reviewer experiences it as freedom. So when someone sits down with Lucky and consumes it without friction, they're not discovering that plot holes don't matter—they're participating in a system that profits from their remaining in a state of gentle incoherence.

The show works. The platform works. The viewer clicks through.

The question isn't whether Taylor-Joy is charismatic or the explosions are competent. The question is whether you can still tell the difference between a show you genuinely want to finish and one designed so you can't stop watching it. That difference used to be something you could feel.