One more road.
One worse commute?
A city gets a free shortcut. Every driver chooses the fastest route. Predict the result, then change the traffic to find where your intuition works.
The original city
Two routesThe city with a shortcut
One added roadMake a prediction. Then meet the other timeline.
Both worlds have identical demand and road rules. The only difference is the new connector. The roads are a mathematical thought experiment; the glowing dots illustrate route use.
Why this happens · inspect the model
This is a simplified illustration of Braess’s paradox and network routing games. It shows a possible incentive problem, not a forecast for a real city.
Home → A and B → Work each cost x / 100 minutes, where x is that road’s traffic. Home → B and A → Work each cost 45 minutes. The added connector goes A → B, at zero cost. Drivers independently choose minimum-time routes at equilibrium; traffic is treated as divisible. There are no traffic lights, physical queues, changing departure times, or adaptive road capacities.
Without the connector, demand N splits evenly and trip time is N / 200 + 45. With the connector: for N ≤ 4,500, everyone uses it and time is N / 50; for 4,500 < N < 9,000, time is 90; for N ≥ 9,000 the connector carries no traffic and time is N / 200 + 45. These formulas are evaluated locally. At N = 4,000 the times are 65 and 80 minutes.
The numerical result is deterministic. Particle timing is decorative and is not a microscopic traffic simulation. The tour changes demand, so each scene is a separate paired comparison. No live model calls, account, analytics, audio, or network requests are needed to play.