Confidence Interval Coverage

Win rates are not fixed numbers — they are ranges. Your 25% win rate might have been 22% last quarter, 28% the quarter before, and 24% the quarter before that. A coverage model based on the average win rate assumes average performance. A confidence interval model accounts for the variance and tells you the coverage needed to hit target even if your win rate is on the low end of the range.

Confidence Interval Coverage Model

Historical Win Rates (8 quarters): 22%, 28%, 24%, 26%, 21%, 27%, 25%, 23%
Mean Win Rate: 24.5%
Standard Deviation: 2.5%

Coverage Requirements by Confidence Level:
  50% Confidence (hit target half the time):
    Coverage = 1 / 0.245 = 4.1x

  75% Confidence (hit target 3 out of 4 quarters):
    Coverage = 1 / (0.245 - 0.675 × 0.025) = 1 / 0.228 = 4.4x

  90% Confidence (hit target 9 out of 10 quarters):
    Coverage = 1 / (0.245 - 1.28 × 0.025) = 1 / 0.213 = 4.7x

  95% Confidence (near-certainty):
    Coverage = 1 / (0.245 - 1.645 × 0.025) = 1 / 0.204 = 4.9x

Recommendation: Plan to 75% confidence (4.4x) for operational targets.
Use 90% confidence (4.7x) for board-level commitments.