| Spent | Slack | Available | |
|---|---|---|---|
| BA | 0 | 0 | 0 |
| DEV | 0 | 0 | 0 |
| QA | 0 | 0 | 0 |
| DEVOPS | 0 | 0 | 0 |
| Spent | Slack | Available | |
|---|---|---|---|
| BA | 0 | 0 | 0 |
| DEV | 0 | 0 | 0 |
| QA | 0 | 0 | 0 |
| DEVOPS | 0 | 0 | 0 |
| Spent | Slack | Available | |
|---|---|---|---|
| BA | 0 | 0 | 0 |
| DEV | 0 | 0 | 0 |
| QA | 0 | 0 | 0 |
| DEVOPS | 0 | 0 | 0 |
| Spent | Slack | Available | |
|---|---|---|---|
| BA | 0 | 0 | 0 |
| DEV | 0 | 0 | 0 |
| QA | 0 | 0 | 0 |
| DEVOPS | 0 | 0 | 0 |
FLOWSIM runs two software delivery simulations side-by-side so you can compare the outcomes of two different ways of managing work: Push (no WIP limits) and Pull (Kanban with WIP limits). Both boards receive identical team effort each day, so any difference in completion time is caused purely by the flow policies.
Every card moves through the same pipeline on both boards:
BACKLOG → ANALYSIS-WIP → ANALYSIS-DONE → DEV-WIP → DEV-DONE → TEST-WIP → TEST-DONE → DEPLOY-WIP → DEPLOYED → DONE
Each -WIP column is where active work happens. Each -DONE column is a buffer holding completed cards waiting to be pulled into the next stage.
| Policy | PUSH board | PULL board |
|---|---|---|
| WIP limits | None — unlimited cards in every column | Enforced per column (adjustable with the ± buttons in each column header) |
| Work item priority | Newest item first (LIFO) | Oldest item first (FIFO) |
| Backlog replenishment | Pushed in continuously | Pulled in only when there is room in ANALYSIS-WIP |
| Rework (defects) | Defective card moved back to DEV-WIP; dev effort consumed on rework, then remaining effort continues new work | Same — rework is prioritised first, then any remaining dev capacity is used for new work |
Each card is assigned a random effort estimate at creation time (converted to Fibonacci story points: 1, 2, 3, 5, 8, 13…). Each stage of the pipeline has its own estimate, so a card may take 2 points of BA work but 8 points of DEV work.
A player's dice roll determines how many effort points they contribute that day. Points are applied to work remaining on the card. When work remaining reaches zero the card advances. If a player has leftover effort after completing a card they do not carry it to another card — effort is lost at the end of each day.
Any card has a 30% chance of becoming blocked when it enters a WIP column. A blocked card is highlighted magenta and cannot be worked until the blocker clears (randomly after 1–7 days). Blockers represent external dependencies, waiting for decisions, or environment issues.
Any card has a 30% chance of being found defective when it reaches TEST-WIP. A defective card turns red and is sent back to DEV-WIP for rework. Once rework is complete it returns to TEST-WIP and must pass testing again before it can advance to DEPLOY.
Controls the total size of the backlog. More cards means a longer simulation and more data points in the charts, but the relative difference between Push and Pull will remain consistent.
Each player rolls a random number within this range each day to determine how many effort points they contribute. A narrower range (e.g. 3–4) produces more predictable flow; a wider range (e.g. 1–10) introduces more variability and makes WIP limits more impactful on the Pull board.
A blocker freezes a card in its current stage for a random number of days. Higher chance or longer durations increase overall cycle time on both boards — but the Pull board is more resilient because WIP limits prevent blockers from stacking up and starving downstream stages.
When a defect is discovered in testing, the card is sent back to Dev for rework. This consumes additional effort and inflates cycle time. On the Push board, many defective cards can re-enter Dev simultaneously with no limit, competing for dev capacity. On the Pull board, the Dev WIP limit throttles how many rework cards can be in flight at once. Higher defect rates widen the gap between Push and Pull outcomes.
Sets the upper bound on how much effort each stage requires per card (actual values are randomised and converted to Fibonacci story points). Higher values mean cards spend more time in each stage. Increasing Dev relative to Analysis and Test will make the Dev bottleneck more pronounced, amplifying the benefit of the Pull board's Dev WIP limit.
| # | Winner | Push days | Pull days | Days saved | Push cost | Pull cost | Money saved |
|---|
No sims recorded yet.