Targeting the Head of the Distribution
When harm follows a power law, policing the average case wastes effort on the long tail. Removing or deterring a few dominant offenders can erase disproportionate harm and reset expectations across the system.
You do not need to catch everybody. Shut down a small fraction of the worst actors, and a large share of the harm may disappear with them.
E1Enforcement has a distribution problem
The strategy works when harmful activity is concentrated rather than evenly spread. Under a Power Law, the leading offenders generate vastly more volume than the typical offender, so treating every case as equally valuable misallocates enforcement. Rank actors by harm, target the small head, and you get two effects: their own outsized volume vanishes, while the visible punishment changes what the remaining actors expect the system to tolerate.
E1Where it shows up
The top offenders
Catching and punishing a small fraction of the biggest actors removes much more harmful volume than pursuing the same number of average cases—and makes an example that reaches beyond those directly caught.
E1First prove there is a head
This is not a license to neglect ordinary cases. It depends on harm actually being concentrated and on the leading actors being identifiable. If damage is evenly distributed, or enforcement merely replaces one top actor with another, targeting a few names will not fix the system.
Rank harm before allocating effort
Tomorrow, list the actors or causes behind the problem, estimate the harm attributable to each, and sort them from largest to smallest. If the first few dominate the total, move enforcement time and deterrence toward that head instead of dividing attention evenly.
Episodes that teach this
-
Simple Maths Explains Why Rich Get Richer - Power Law - FutureIQ
· explained at 10:48
11,972 views
"It is not necessary to catch and punish everybody. You just take a small fraction of the top guys and you shut them down... it gets rid of a lot of the volume entirely and it serves as an example."