Avg Std Dev
Key Avg Std Dev in Simulation Report refers to the estimated standard deviation of the distribution that governs the theoretical mean ͞Q of a quantity Q of which the sample values generated through a simulation are qᵢ, i = 1, ..., N, where N is the number of scenarios in that simulation.
Look at Average for more details regarding ͞Q and the values qᵢ, i = 1, ..., N.
Note, the best estimate of ͞Q is the average <Q> of qᵢ, i = 1, ..., N reported in Average
The value here quantifies how good this estimate is in the following sense:
For a given calculated average <Q>, the true - but unknown - mean ͞Q should fall anywhere in the vicinity of <Q> in accordance with the following probabilities:
Prob( <Q> - σ ≤ ͞Q ≤ <Q> + σ ) = 68.27%
Prob( <Q> - 2σ ≤ ͞Q ≤ <Q> + 2σ ) = 95.45%
Prob( <Q> - 3σ ≤ ͞Q ≤ <Q> + 3σ ) = 99.73%
In the special case of a simulation consisting of randomized sub-simulations, the number N refers to the number of sub-simulations and each qᵢ to the average of the respective sub-simulation.
