Failure, not gear, is the most underused training tool in photography. A new sensor only records what the eye has already missed, while systematic mistakes expose the blind spots that money cannot fix. Treating each bad frame as data turns the camera from a trophy into a diagnostic device for perception itself.
On this method rests a blunt claim: one week of shooting one hundred awful images a day, then dissecting them, beats months of casual snapping with a premium body. This is deliberate practice in its pure form, a close cousin of error-based learning in cognitive science, where performance improves faster when the brain receives dense, specific feedback instead of rare, flattering success.
The rule is simple, and uncomfortable. Go out and chase mistakes: blown highlights, tangled backgrounds, flat light, awkward crops, lifeless moments. Then sit down and run a quiet post‑mortem on each: where did the exposure histogram go off, how did depth of field betray the subject, where did your eye wander before the shutter clicked. Pattern recognition does the rest, stitching those small autopsies into instinctive corrections the next time the viewfinder rises to your face.
