RunTheSim
Every driver has the same reflex: look for the wreck. You crawl for two minutes, you pass nothing at all, and then you are back at 120 as if the road had changed its mind. That reflex is expensive, because it points at the wrong repair. If jams come from obstacles, you clear obstacles. If they come from density, there is nothing to clear.
A ring road with no obstacle still jams
In 2008 a Japanese research group put 22 cars on a circular track, 230 metres around, and asked the drivers for one thing: hold about 30 km/h and keep a safe gap (Sugiyama et al., New Journal of Physics).
For a while the ring held. Then the gaps began to breathe, and within minutes five cars stood still. The knot kept its size and drifted backward at roughly 20 km/h, the same speed measured on real motorways. Cars entered it from behind, stopped, and accelerated out the front, while the knot stayed where physics put it. Nothing on that track had gone wrong: no crash, no roadworks, no slow lorry. Density alone was enough. The paper puts the point in one line:
A bottleneck is only a trigger and not the essential origin of a traffic jam.
That is the reframe. A jam is not an object sitting on the road, it is a pattern moving through the cars, which is emergence in its plainest form.
Switzerland counts the same effect in hours
The ring track is a laboratory. The Swiss motorway network is not, and it keeps a ledger. The Federal Roads Office recorded 55'569 hours of congestion on national roads in 2024, 13.9 percent more than the year before, and it books about 87 percent of them under plain overload rather than crashes or roadworks (ASTRA, traffic flow report 2024).
of Swiss congestion hours come from overload
Crashes and roadworks account for the rest. Overload is an administrative word for too many cars per kilometre.
Read that as a threshold report. On the days the network holds, the same braking gets absorbed a few cars back. On the days it carries more cars per kilometre than it can smooth out, the same braking survives, spreads, and shows up in the statistics as hours.
Sometimes you really did pass the cause
The honest objection: you have passed the roadworks, you have seen the flashing lights, so the jam clearly had a reason. Often it did. What the trigger explains is where a jam begins, not why it keeps standing four kilometres upstream long after the lane reopened.
| What you notice | Jam with a trigger | Jam out of nowhere |
|---|---|---|
| What starts it | A crash, a lane closure, a steep on-ramp | One driver braking a little too hard |
| What holds it together | Density above the critical point | Density above the critical point |
| What you pass at the front | The cause, once | Nothing at all |
The middle row carries the argument. The trigger changes where the wave is born; the load on the road decides whether it dies out or keeps travelling backwards.
Below the threshold none of this happens
Under the critical density the same braking is swallowed within a few cars, which is why an empty motorway at three in the morning shrugs off a slow lorry. The ring experiment is also a single lane with identical drivers, and real roads add exits, lane changes, weather and tired people, all of which shift the threshold in both directions. Every model on RunTheSim trades that mess for one visible rule, which is the whole point and the whole limit.
So the next time the brake lights come on, watch the wave instead of hunting for the cause. Hold the gap you have, brake early and softly, and hand the car behind you a smaller shock than the one you got. One driver cannot clear a jam. One driver can stop feeding it.
You can read about the threshold, or you can add cars with a slider and watch the knot appear.
Run the traffic jam sim