Choke Point Analysis: Finding Where a Route Can Fail

A choke point is any feature that compresses movement and removes options — a place where a vehicle or a walking formation can neither maneuver nor turn around. Choke point analysis is the discipline of finding those places before the route does.

This is a component of route work, not a substitute for it. If you haven't built the route survey itself yet, start with how to run a route survey that holds up. This post assumes you have a route and now need to know where it can fail.

Why choke points matter

Two things decide it: predictability and compression. A choke point takes away what keeps a moving target hard to act against — speed, spacing, and the ability to leave. Inside one, you are slow, canalized, and committed. Everything an adversary needs to solve their targeting problem, the terrain has already solved for them.

The part that should bother you more: an adversary doesn't need to know your route. If a venue sits at the end of a peninsula, behind a river, or inside a downtown grid with three viable approaches, then every route to that venue passes through a small set of predictable points. They don't have to surveil your movement. They have to surveil the map.

Historically, attacks on protected movements cluster where the target is predictable and the options are few. That is a description of a choke point.

You don't need incident citations to act on that logic. Accept that the terrain around your venue is doing analysis for the other side whether or not anyone is currently looking, and do your own analysis first.

Finding choke points on a vehicle route

Start with the features that reliably produce them. On a vehicle route, look for:

• Bridges and overpasses: no lateral options, and elevated observation over you
• Tunnels: no options at all, plus degraded comms
• On-ramps and off-ramps: single lane, committed geometry, often walled
• Single-lane construction zones: a choke point somebody else created last week and your map doesn't show
• Controlled intersections with long signal cycles: a stationary vehicle on a schedule
• Garage entrances and ramps: slow, tight, blind corners, and predictable because the venue dictates them
• Gated communities and checkpoints: deliberate stops at a known location
• Dead-end approaches to venues: one way in means one way out
• One-way streets: the map removes your turn-around before anyone else has to

Then run the method: map first, then drive. The map pass comes first because it's cheap and it scales. Satellite plus street-level imagery on every segment where lanes drop, medians appear, walls close in, or turn options disappear. Mark every candidate. You will over-collect on the map pass; that's fine. The map's job is to make sure the drive misses nothing.

Then drive it — and drive it at the time of day the movement actually runs. A choke point at 0800 may be open road at 1400. A garage ramp with a queue at shift change is a different feature at midnight. Traffic is part of the terrain, and terrain assessed at the wrong hour is a different route wearing the same name. The drive confirms or kills each candidate from the map pass and catches what imagery can't: the delivery trucks that always double-park on that block, the school pickup line, the construction that started Tuesday.

Three dispositions: avoid, mitigate, or accept and watch

A list of choke points is collection; analysis is deciding what to do with each one. Every choke point on the route gets one of three dispositions:

• Avoid. Reroute around it. This is the cleanest answer and the one you should reach for first, but it isn't free. The reroute has its own choke points, and sometimes the avoidance route is worse than the feature you dodged. Compare, don't assume.
• Mitigate. You're going through, so change the conditions. Adjust timing so you hit it off-cycle. Adjust spacing and speed so time-in-the-tube drops. Put advance eyes on it before the package arrives. Mitigation is a set of specific decisions, not a mood — "we'll be careful" is not a mitigation.
• Accept and watch. Some choke points can't be avoided or meaningfully mitigated; the garage ramp at the venue is the garage ramp at the venue. Put it in the brief by name, and name who watches it. Assigned eyes, a trigger for what "wrong" looks like there, and a decision point short of it while options still exist.

A choke point you've named and assigned is a different animal from one you discover in the moment.

One boundary worth stating plainly: this is planning work. What happens if an attack actually initiates inside a choke point is a different mission set; that lives in SOPs and rehearsals, not in route analysis.

Choke points on foot

The vehicle drops the principal, and the same geometry problem walks inside with you. Venue interiors compress movement constantly: stairwells, elevators, single-door corridors, loading docks, press lines, the pinch between the ballroom exit and the service hallway. An elevator is a moving choke point with a schedule you don't control.

The analysis is the same — find the compression, assign a disposition, name a watcher. It just happens during the site survey instead of the route drive. The interior coverage list in site survey vs. advance survey covers where these hide inside a venue.

What the red cell does with your choke points

If you wargame the advance (and you should), watch what the red cell reaches for first. It's the choke points, every time. An adversary planning against your movement doesn't start with your strengths; they start with the places where your options are already gone. Which means your choke point list, built without kidding yourself, is a preview of the enemy's target folder. Running that adversarial pass deliberately is the subject of wargaming an advance; choke point analysis is the input that makes it worth doing.

Doing this with a small team

A full route package with alternates usually means a primary, an alternate, and maybe a third. Analyze the choke points on all of them — the map pass is cheap enough that there's no excuse not to. But spend the drive-time where the probability is: the primary gets driven at movement hour, the alternate gets driven if time allows, and the third may live on imagery alone. That's allocation, not sloppiness.

And ask a local driver. The person who drives that city daily knows which pinches are real and which look bad on satellite but flow fine at your hour, and knows about the construction zone that appeared last week. Ten minutes of that conversation can outrun an hour of imagery. The executing team owns the plan and the agents on the ground own the route — but neither should own it uninformed when local knowledge is a phone call away.

This is one way to run it, not the only way. But some version of it belongs in every advance, no matter how thin the team: find the compression, decide what to do about each point, and put names against the ones you accept.