Blowout Predictors
A heuristic model flagging projects likely to drift further based on historical patterns.
What it shows
For each active project: - A risk score (0–100) based on measurable factors. - The contributing signals (e.g. "TEI revised 3 times in 2 years", "completion date slipped by >18 months", "delivery entity has history of drift on similar projects"). - A peer comparison against projects of similar sector and scale.
How to read it
This is a heuristic, not a forecast. A high risk score means the project shares observable traits with historically-drifting projects. It does not mean the project will blow out.
Use the risk score to decide what to watch — not what to conclude.
Data sources
- Every TEI snapshot in the database.
- Completion date changes.
- Delivery entity historical performance.
- Scope change logs.
Caveats
- The model sees public data only. Internal project health is not visible.
- Pattern matching on historical BP4s is biased by whatever our dataset contains.
- High-risk doesn't imply misconduct — it implies complexity, ambition, or poor planning.
Related pages
- Cost Drift — the historical record the model learns from.
- Project Detail — see the risk signals for a specific project.
Last reviewed: 11 Apr 2026