EP-02

Estimation Precision Distribution — Fat Tail

Estimation Precision

Default severity: medium
EstimationQualityBusiness ValueEffort

What it detects

The distribution of (actual_time / estimated_time) ratios has a fat right tail — a significant proportion of issues are severely overrunning their estimates. Even if the median is acceptable, tail risk dominates actual delivery. Measured across all issues with both estimate and actuals. ratios = \[time_spent / original_estimate

Detection formula

FOR each issue WHERE both fields set AND time_spent > 0\] p90 = PERCENTILE(ratios, 90) p95 = PERCENTILE(ratios, 95) IF p90 > config.ep.p90_ratio_threshold // default: 3.0 (90th pct = 3x estimate) THEN FLAG fat_tail IF p95 > config.ep.p95_ratio_threshold // default: 5.0 THEN CRITICAL Report: distribution histogram of ratios in buckets \[\<1, 1-1.5, 1.5-2, 2-3, 3-5, >5\]

Examples in practice

  • A team shows the distribution of (actual_time / estimated_time) ratios has a fat right tail while p90 ratio threshold is set to 3.0 (90th pct = 3x estimate).
  • Example signal: The distribution of (actual_time / estimated_time) ratios has a fat right tail — a significant proportion of issues are severely overrunning their estimates.

Suggested response

Recalibrate estimation precision—check size classes, drift, and BV vs effort trade-offs.

Coaching playbook

Symptom

The distribution of (actual_time / estimated_time) ratios has a fat right tail — a significant proportion of issues are severely overrunning their estimates. Even if the median is acceptable, tail risk dominates actual delivery. Measured across all issues with both estimate and actuals. ratios = \[time_spent / original_estimate

Why it matters

When "Estimation Precision Distribution — Fat Tail" keeps appearing, the team is signalling a repeatable process gap. Left unexamined, the pattern hides where work really stalls and makes improvement metrics harder to trust.

What you can achieve

Recalibrate estimation precision—check size classes, drift, and BV vs effort trade-offs.

Facilitation questions

  • What system change would stop "Estimation Precision Distribution — Fat Tail" from firing again?
  • What do the cited issues have in common — same root cause or same workaround?
  • Who owns the two-week experiment and how will we verify on the next import?

Run this rule against your own tracker data with Flow Analyzer.

Product tourSign in
EP-02: Estimation Precision Distribution — Fat Tail — FlowAnalyzer