Examples of Statistical Forecast Calculations
This example demonstrates how statistical forecast calculations are performed in Cobra.
For the following examples of the different statistical forecast calculations, assume that forecasts are being generated following the closing of period 8 for a work package with the following budgeted, earned value, and actual costs:
Period 1 | Period 2 | Period 3 | Period 4 | Period 5 | Period 6 | Period 7 | Period 8 | Period 9 | Period 10 | Total | |
---|---|---|---|---|---|---|---|---|---|---|---|
Budget | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 1000 |
Earned Value | 100 | 75 | 75 | 75 | 100 | 75 | 100 | 100 | 700 | ||
Actual Costs | 100 | 100 | 100 | 100 | 125 | 100 | 125 | 150 | 900 |
PF = 1
This method always calculates ETC by subtracting Earned Value from BAC, which implies that performance is following the budgeted schedule (that is, assumes a performance factor of 1):
ETC = 1000 – 700 = 300
EAC = 900 + 300 = 1200
PF = 1/CPI Cum to date
Uses the cumulative-to-date CPI, which trends ETC based on the total past performance of work completed.
PF = 900/700 = 1.2857
ETC = 1.2857 * 300 = 386
EAC = 900 + 386 = 1286
PF = 1/CPI current period
This method calculates a performance factor using a CPI based on the most recent fiscal period (in the following example, period 8):
PF = 150/100 = 1.5
ETC = 1.5 * (1000-700) = 450
EAC = 900 + 450 = 1350
PF = 1/CPI last 3 periods
This method calculates a performance factor using a CPI based on the three most recent fiscal periods (in the following example, periods 6, 7, 8):
PF = 375/275 = 1.3636
ETC = 1.3636 * 300 = 409
EAC = 900 + 409 = 1309
PF = 1/CPI last 6 periods
This method calculates a performance factor using a CPI based on the six most recent fiscal periods (in the following example, periods 3, 4, 5, 6, 7, 8):
PF = 700/525 = 1.3333
ETC = 1.3333 * 300 = 400
EAC = 900 + 400 = 1300
PF = User input
This method allows the user to input a PF value. In this example, assume that the user entered a value of 1.1:
PF = 1.1
ETC = 1.1 * 300 = 330
EAC = 900 + 330 = 1230
PF = 1/(a*CPI) + (b*SPI)
For this method, the calculation of the performance factor is based on cumulative-to-date SPI and CPI, allowing the user to define what proportion of each should be used, where a and b are the proportional factors and a + b = 1. In this case, assume the user has set both a and b to 0.5:
PF = 1/((.5 * 700/900) + (.5 * 700/800)) = 1.21
ETC = 1.21 * 300 = 363
EAC = 900 + 363 = 1263
Defining a as zero and b as 1 results in the following calculations:
PF = 1/(0 + (700/800)) = 1.143
ETC = 1.143 * 300 = 343 (342.86)
EAC = 900 + 343 = 1243
PF = 1/(CPI * SPI)
This method provides a worst-case scenario by multiplying the cumulative-to-date performance indices together, thereby heightening the effect of cost and schedule variance.
PF = 1/((700/900) * (700/900)) = 1.653
ETC = 1.653 * 300 = 496
EAC = 900 + 496 = 1396
Percent Complete Ranges
This method allows for changing the calculation of the performance automatically depending upon the percent complete of the level at which the performance is being calculated. Cobra determines this percentage using the following formula:
Percent complete = 100 * (Earned Value/BAC)
For this example, assume the following ranges and methods have been defined:
Range | Method |
---|---|
0 – 25 | PF=1 |
26 – 80 | PF=1/CPI |
81 – 100 | Method A (Retain EAC, a manual forecasting method) |
The current percent complete of the work package is calculated as follows:
Percent complete = 100 * (700/1000) = 70
As a result, Cobra uses method 2 and calculates the EAC as 1286.