Estimates run long โ that's Parkinson's lawParkinson's law: work expands to fill the time allotted to it. Slack in an estimate tends to get used up, so tasks rarely come in early even when they could., not bad luck. Instead of guessing a padding factor, feed in how past tasks estimated versus how long they really took, and read off a buffer sized to your own track record.
| Task | Estimate | Actual | Slip | |
|---|---|---|---|---|
| 1.6ร | ||||
| 1.1ร | ||||
| 2.33ร | ||||
| 1.13ร | ||||
| 1ร | ||||
| 1.67ร |
Your typical task slips to 1.36ร of its estimate; at 80% confidence you'd budget 1.67ร. Push the confidence higher to cover more of the bad days โ at the cost of more padding you may not use.