Example SCI-106: Zero bias large error — Positive and negative period errors cancel.
What goes wrong: A correct total does not prove accurate timing. How we fix it: Evaluate a proposed revision; do not treat it as a future guarantee.
Interactive calculator loads automatically in modern browsers. All plain-English explanations, units, and tables are shown directly on this page.
Visual Before-and-After Chart (SCI-106)
Key Question to Ask: Inspect absolute error alongside signed total error.
How to Fix It: Evaluate a proposed revision; do not treat it as a future guarantee.
Simple Summary of This Example
- Example Number:
SCI-106(seasonal accessories) - Plain-English Topic: Why Over-Guessing One Week and Under-Guessing the Next Still Hurts 100%
- Main Goal: separate accuracy from bias
- Important Limit: Four toy periods; no claim of causal forecasting improvement.
Starting Numbers & Units for SCI-106
| Number Name | Unit of Measure | Starting Value | Allowed Range |
|---|---|---|---|
actual 1 (actual_1) |
pieces | 0 | 0 to 1000000 (integer) |
actual 2 (actual_2) |
pieces | 100 | 0 to 1000000 (integer) |
actual 3 (actual_3) |
pieces | 200 | 0 to 1000000 (integer) |
actual 4 (actual_4) |
pieces | 100 | 0 to 1000000 (integer) |
forecast 1 (forecast_1) |
pieces | 0 | 0 to 1000000 (integer) |
forecast 2 (forecast_2) |
pieces | 100 | 0 to 1000000 (integer) |
forecast 3 (forecast_3) |
pieces | 200 | 0 to 1000000 (integer) |
forecast 4 (forecast_4) |
pieces | 100 | 0 to 1000000 (integer) |
stress 1 (stress_1) |
pieces | 100 | 0 to 1000000 (integer) |
stress 2 (stress_2) |
pieces | 0 | 0 to 1000000 (integer) |
stress 3 (stress_3) |
pieces | 100 | 0 to 1000000 (integer) |
stress 4 (stress_4) |
pieces | 200 | 0 to 1000000 (integer) |
revised 1 (revised_1) |
pieces | 0 | 0 to 1000000 (integer) |
revised 2 (revised_2) |
pieces | 90 | 0 to 1000000 (integer) |
revised 3 (revised_3) |
pieces | 190 | 0 to 1000000 (integer) |
revised 4 (revised_4) |
pieces | 120 | 0 to 1000000 (integer) |
Compare Normal Day vs. Problem Day vs. Fixed Plan (SCI-106)
| Result Name | Unit | Normal Day (Baseline) | Problem Day (Stress) | Fixed Plan (Repair) |
|---|---|---|---|---|
total actual (total_actual) |
pieces | 400 | 400 | 400 |
total forecast (total_forecast) |
pieces | 400 | 400 | 400 |
signed error (signed_error) |
pieces | 0 | 0 | 0 |
absolute error (absolute_error) |
pieces | 0 | 400 | 40 |
wape percent (wape_percent) |
percent from zero to one hundred | 0 | 100 | 10 |
zero actual periods (zero_actual_periods) |
planning periods | 1 | 1 | 1 |
Other Related Examples on "Why Over-Guessing One Week and Under-Guessing the Next Still Hurts 100%"
- SCI-106: Zero bias large error (This Page)
- SCI-107: Zero demand period — One period has zero observed demand.
- SCI-108: Forecast shifted one period — The forecast pattern occurs in the wrong periods.
- SCI-109: Revision evaluated fairly — A revised forecast uses the same evaluation window.
- SCI-110: Outlier dominates aggregate — Large periods dominate a volume-weighted score.
- SCI-111: Sign convention disagreement — Teams use opposite definitions for bias signs.
- SCI-112: Accuracy is not a service plan — Forecast scoring is separate from inventory decisions.