Signal-to-Noise in Performance Reviews
The Physics: In radio astronomy, receiving a faint, ancient signal from deep space requires aggressively filtering out the overwhelming "thermal noise" of our own solar system.
The Scar: Watching a brilliant, foundational engineer almost get managed out because their quiet, stabilizing work was drowned out by the noise of a wildly chaotic peer.
The Lesson: A good engineering manager is a radio telescope. You must actively tune out the loud, localized static of office politics to find the deep, pulsar-like consistency of true talent.
The Problem of Local Amplification
If you point a radio telescope at the sky without any filters, you won't hear the elegant, rhythmic pulsing of a distant neutron star. You will only hear the deafening static of the sun, the interference of human satellites, and the thermal noise of the Earth's atmosphere. The loudest things are always the closest things.
Performance review cycles suffer from the exact same physics.
Local amplification naturally warps management perception. The engineer who writes sloppy code, breaks production, and then stays up until 3 AM dramatically "firefighting" the outage they caused generates an enormous amount of noise. Their name is in every Slack channel. Executives see them as a hero.
Meanwhile, the deeply experienced engineer who writes perfectly tested code that never breaks in the first place generates zero noise. Their work is completely silent. Because they never start fires, they never get points for putting them out.
Identifying the Thermal Background
To be an effective leader, you have to build high-fidelity filters. You cannot let the volume of the broadcast dictate your evaluation of the talent.
You must learn to identify the "thermal background noise" of your organization. This includes performative, late-night Slack messages. It includes being the loudest voice in an architecture meeting without actually writing any foundational code. It includes the endless, high-visibility bikeshedding over trivial features.
Noise is highly visible, but fundamentally chaotic and low-value. You have to tune it out.
Isolating the Pulsar
A true 10x developer rarely operates like a supernova. They operate like a pulsar: a reliable, deep, recurring signal that never wavers.
You have to actively scan the spectrum to find them. Look for the invisible mentorship occurring in the quiet 1:1 Slack threads. Look for the engineer whose pull request reviews are consistently the most thorough and educational. Look for the person who volunteers to refactor the brittle CI/CD pipeline so the rest of the team can ship faster.
This type of work stabilizes the entire organization, but it rarely generates a loud, self-promoting broadcast.
Without an active noise-cancellation strategy at review time, organizations organically promote the noisy and starve the signal. Build your filters, tune your instruments, and reward the quiet consistency that actually keeps the systems running.