If the carrier suspension on the Asia–EU lane holds six weeks, retail lead times breach SLA in 62% of runs
Median delay +9.4 days (CI 7.1–12.8) across 10,000 simulations. The largest driver of the outcome is tier-2 supplier rerouting capacity — not port congestion.
Outcome distribution — lead-time delay, weeks 0–8
SLA breach probability
62%
CI 57–66%
Median delay
+9.4 days
CI 7.1–12.8
Most sensitive input
Reroute capacity
±20% moves breach rate 49–71%
Assumptions — visible in full, editable by anyone
- suspension_duration
- 6 weeks (the scenario under test)
- reroute_capacity
- 35% of suspended volume absorbed by alternates
- tier2_lead_time
- Lognormal, mean 21d — fitted to 24 months of shipment history
- demand
- Negative binomial, seasonal profile, no spike assumed
- port_congestion
- Baseline; no compounding disruption assumed
- simulations
- 10,000 · seed 84127 · engine v2.3
Disagree with an assumption?
Good — that's the point of showing them. Fork this run, change any input, and the engine re-runs the full batch. Your version and this one sit side by side, and the disagreement becomes a comparison instead of an argument. Forking is free and needs no account.
Fork this runIn the live product this button opens the instrument with this run's assumptions pre-loaded.
Verification
What the mark certifies
This finding carries the Decision Observatory mark. It certifies that the computation ran on the calibrated engine at the stated scale, from exactly the assumptions listed above, and that the result is reproducible from the attached seed — same seed, same result, byte for byte.
It does not certify the assumptions. Assumptions are the analyst's judgment, shown in full so you can exercise your own. The engine guarantees honest arithmetic; the reasoning is on display; the track record shows how well-calibrated the arithmetic has proven to be.
Analyst attribution, firm styling, and the permanent citation line appear here on live findings.