# Essay on Cases: Decision Theory and Value

Submitted By swathinr
Words: 1013
Pages: 5

ORION CONTROLS

RECOMMENDATIONS AND REASONS
I would recommend Armstrong in response to Avion’s proposal, to go with the option of starting new project development.
Reason:
The initial constructed with different phases like software development, valve redesign with two cases of dramatic performance and modest performance, manufacturing and probabilities of success and failure and costs involved and revenues generated are shown in( Exhibit 1.)The
Reason for this recommendation is, if they start new project the expected value is 105.5 .There is Gain if this is selected.
Assumptions involved with the analysis are
 In the Initial tree ,successful working of software short cut option would lead to take up subsequent phases
 The probabilities and revenues mentioned are accurate
 The costs involved in the test process includes all operational and other expenditure costs
 There are no hidden costs involved in the process.
 All the time durations mentioned are appropriate.
Monetary Calculations:
Costs involved
 There is an investment of -120 kmillion dollars if shortcut option is successful and -240 k million if it is not successful.
 There is an investment of -80 kmillion dollars if the valve redesign is successful and -80k million dollars if not successful.
 There is -525 k million dollars investment if valve redesign is drastically performed and also performance is modest.
Revenue generated
 For the old system (sv44a) , there is revenue of 100k million

 For the new system, and drastic performance the revenue generated is 1000kmillion dollars and for modest performance the revenue generated is 600k million dollars.
Sensitivity Analysis
Sensitivity analysis is performed on initial case.
Tornado graph (Exhibit 2)
 The Tornado graph indicates branch value ‘dramatic’ node of improvements has higher sensitivity on expected value
 The Tornado graph indicates ‘branch probability of ‘works shortcut option has second highest sensitivity on expected value
 The Tornado graph indicates ‘branch probability of dramatic of node improvements has next high sensitivity on expected value.
Sensitivity graph analysis for three most sensitive values
 From the sensitive analysis graph for branch value ‘dramatic’ node of improvements the insights derived are, If the value of revenue-costs are less than 475, continuing with old option is profitable but after hitting 475 value the new option increases and expected value is higher, so it is profitable.(exhibit 3)
 From the sensitive analysis graph for branch probability of ‘works shortcut, the insights derived are, if the probability is below 75%,old option is profitable and after it touches
75%,the new system is profitable and the expected value is higher.(exhibit 4)
 From the sensitive analysis graph for branch probability of dramatic of node improvements, the insights derived are, if the probability is below 80%, old option is profitable and after it touches 80 %, the new system is profitable. Expected value is higher
(exhibit 5)

1) Options included (exhibit 6)
 First option included is the decision node for deciding to continue further if the software short cut option does not works. We chose to stop. This option changed the expected value at that particular node from -74.5 to -20.
 Second option included is to decide between continue or stop if the valve redesign performance is modest. We chose to stop the new system and continue with old system which would give us 100k million profit.

 Third Option included is to decide between continue or stop if the valve redesign performance is modest. We chose to stop the new system and continue with old system which would change the expected value at that point from -45 to -25.
All These Options raised the expected value from 105.5 to 122.5

2) Eliminating Risk of failure for this activity (exhibit 7)
Risk of failure can be at the Valve redesign phase. To eliminate that risk