This is a discrete-event simulation. Each tick = 1 work day. Both boards receive the same tickets at the same rate from an identical backlog. The only variable is how the team works — handoffs vs. collaboration. All parameters are drawn from peer-reviewed research and industry benchmarks cited below.
Deterministic & reproducible. Uses a seeded pseudo-random number generator (Lehmer LCG, seed 42). Every run produces the same sequence. Reset always replays identically.
Right-to-left processing. Each day, columns are processed from Released ← Deployed ← Review ← Development ← Backlog. This mimics a pull system — downstream must have capacity before upstream pushes. The handoff board ignores this discipline; the flow board enforces it via WIP limits.
Same inputs, different rules. Both boards share the same ticket pool, arrival rate, and team size (4 people). The only differences are: WIP limits (none vs. enforced), review model (separate QA vs. team-owned), deployment model (batch vs. continuous), and collaboration model (solo vs. pairing).
This simulation is simplified to illustrate core dynamics. It does not model: team morale/burnout, learning curves, ticket complexity variance, external dependencies, meetings, PTO, incident interrupts, or the human cost of constant context switching. In practice, these factors make the handoff model even worse than shown here.