Streetlight Effect
Looking where the light is good rather than where the answer is.
The streetlight effect is the tendency to search where observation is easy rather than where the thing you are looking for actually is. The name comes from the parable of a drunk man searching under a streetlight for keys he dropped elsewhere, because, as he explains, "this is where the light is." The joke works because the logic is locally sound and globally useless.
In practice the effect is rarely a personal failing so much as a property of Instrumentation: a system logs what someone thought to log, a metric measures what was cheap to measure, and both attract disproportionate attention afterward simply because they are the data that exists. A dashboard full of request latency says nothing about the request that never got dispatched. Investigation converges on the visible layer, and the write-up mistakes thoroughness of the search for coverage of the problem.
Because the cause is instrumentation rather than laziness, the fix is rarely "look harder" — a wider search under the same streetlight still only covers the same patch of ground. The fix is Observability: extending what can be seen at all, so that looking becomes possible in places it previously was not. This is also why the effect compounds with Survivorship Bias and Selection Bias — cases that never reach the lit area are excluded from the sample before any bias in interpreting the sample even gets a chance to operate, which is a Root Cause Analysis built entirely on the wrong half of the evidence.
See also4
Hand-picked in the note itself — the neighbours worth reading next.
Ground Truth
The measurement that arbitrates, as opposed to the artifact that merely reports.
Method92 connections
Silent Failure
A failure that produces no signal, so the absence of an error reads as success.
Method62 connections
Truncation Bias
Reading a truncated result as if it were the whole result, so every counterexample is invisible.
Method26 connections
Fermi Estimation
Reaching a defensible order-of-magnitude answer by decomposing a question into estimable factors.
Method21 connections
Linked from3
Notes elsewhere in the wiki that reach for this one.