Forecast high-demand capacity from last year's same-period actuals, not from your current energy — the outside view beats the inside view

Why This Is a Rule

When planning for a future high-demand period, your brain defaults to estimating from your current state: "I feel energized and focused right now, so I can probably handle 5 projects during the December crunch." But your current state (calm October afternoon, well-rested, no acute pressure) is nothing like your December state (holiday obligations, year-end deadlines, dark cold days, accumulated fatigue). The current-state estimate is the inside view applied across time — it systematically overestimates future capacity by projecting present conditions onto different circumstances.

Start project estimates with reference class forecasting from 3-5 comparable past projects — base rates beat inside-view planning established reference class forecasting as the antidote to inside-view estimation at the project level. This rule applies the same principle at the seasonal level: your best predictor of December capacity is last December's actual capacity, not your current feeling about what December might be like. Last December includes all the real-world friction that your current optimistic imagination excludes: the holiday parties that consumed evenings, the family obligations that consumed weekends, the year-end reporting that consumed work hours, and the general fatigue that accumulated by December 15th.

The reference class is the same period in prior years, not the average of all months. Seasonal variation means that December and July have fundamentally different capacity profiles, even though they're both "months." Using annual averages would undercount seasonal effects. Using the same month from prior years captures the seasonal-specific reality.

When This Fires

• When planning commitments for upcoming high-demand periods identified by Audit the past 12 months of calendar data — rate each month as low/baseline/high-demand/crisis to reveal recurring seasonal patterns
• When your inside-view estimate for a future period feels optimistic relative to past experience
• When seasonal patterns repeat but your planning doesn't account for them
• Complements Start project estimates with reference class forecasting from 3-5 comparable past projects — base rates beat inside-view planning (project-level reference class forecasting) with the seasonal application

Common Failure Mode

"This year will be different" — the seasonal version of the planning fallacy. You remember last December was overwhelming, but you believe this December will be better because [reason that sounds compelling but changes nothing structural]. The structural drivers of seasonal demand (holidays, fiscal year-end, weather, school schedules) don't change year to year. Your capacity during those periods won't either.

The Protocol

(1) For each upcoming high-demand period, pull your actual data from the same period last year: how many projects were active? How many completed? How many slipped? What was your stress level? (2) Use last year's actuals as your capacity ceiling: "Last December I completed 3 projects, not the 5 I planned. This December, plan for 3." (3) If you have multi-year data, use the average across years — this smooths out single-year anomalies. (4) Allow modest upward adjustment only for genuine structural changes: "I hired an assistant this year" justifies slightly higher capacity. "I feel motivated" does not. (5) When in doubt, plan for the lower end of historical capacity. Under-committing during high-demand periods is vastly less costly than over-committing.