System inventory
Each point is the total inventory held at the end of a completed round, using the left axis.
Simulating Eliyahu Goldrattβs Chapter 14 campfire experiment: watch why statistical fluctuations and dependent events cause a "balanced plant" to collapse into missed deliveries and overflowing inventory.
Ready. Plate 1 is active.
Each row is one round: seven turns, with all six plates and throughput acting once. The tables log every stage transfer; deviation compares each with the 3.5 midpoint.
Each point is the total inventory held at the end of a completed round, using the left axis.
Each completed round shapes the lines. Hover any line to reveal a value; plate lines use the left axis and the system total uses the right.
Actual transfer made by each stage during the round.
| Round | P1 | P2 | P3 | P4 | P5 | P6 | Throughput |
|---|---|---|---|---|---|---|---|
| Roll a turn to start the first round. | |||||||
Matches pulled from the previous stage minus the 3.5 midpoint.
| Round | P1 | P2 | P3 | P4 | P5 | P6 | Throughput |
|---|---|---|---|---|---|---|---|
| Roll a turn to start the first round. | |||||||
Around the campfire with his son Davey and four scouts, Alex invents a game using a box of matches, bowls, and a die. Each bowl represents a work center; each roll (1β6, average capacity 3.5) represents variable daily output. He expects ~20 finished matches after several rounds.
The boys experience reality: rolling a 6 is useless if the previous bowl only has 2 matches (starvation). Rolling a 1 leaves matches trapped (inventory accumulation). Positive fluctuations cannot carry over, while negative fluctuations permanently delay the system.
Alex stays up contemplating the disaster. He realizes that a factory balanced to match exact market demand is mathematically impossible. In a dependency chain, variance accumulates strictly downward. The only way to survive is to manage bottlenecks rather than balance capacity.