UPI Study › Courses › Quantitative Analysis › Free lessons › Lesson 6.2 Free with a free account
Business Decision Strategies
Quantitative Analysis · Chapter 6: Bayesian Revision and UtilitySection 6.2 · lesson 27 of 5511 min · free with a free account
Free preview
Create a free account to keep watching
It takes a minute and costs nothing. Your free account unlocks all of Chapter 1 and 3 more lessons of Quantitative Analysis.
Lesson complete
Earn the 3 college credits for Quantitative Analysis
Enrolling adds a quiz after each lesson, the graded work and the official transcript. Recommended for credit by NCCRS.
Next lesson: Reasoning Backward From Evidence →← Revising the EstimateNext lesson: Reasoning Backward From Evidence →
In this lesson
Strategic logic beyond a single expected value - dominance, randomisation, flexibility, and staged commitment.
- Identify a dominant strategy and use it to shrink the problem
- Recognise situations that call for a mixed, randomised strategy
- Quantify strategic flexibility and the irreversibility premium
- Compare full-commit and pilot-first strategies on EMV and option value
Transcript
Everything said in the video, section by section (1,665 words). Read along, or search it.
Business Decision Strategies · Section 6.2
Playing It Smart
Most decision analysis ends with a single number. You compute expected monetary value for each option, and pick the largest. Real strategic decisions are rarely that tidy. They involve competing objectives, commitments you cannot undo, and rivals who react to you. They also involve something expected value ignores, the value of changing your mind later. This section works through five inter-related ideas. Dominant strategies, mixed strategies, strategic flexibility, the irreversibility premium, and sequential commitment. Then we apply all five to two firms facing one new technology. By the end, you should know what an irreversible commitment must earn before it is worth making.
Beyond Highest EMV
Five Ideas That EMV Alone Cannot Reach
Start with what expected monetary value does well. It collapses an uncertain future into one comparable number per alternative. That is genuinely useful, and it is where most courses stop. But it quietly assumes the decision is a one-shot bet. You commit, the world resolves, and nothing you learn in between matters. Real decisions almost never work that way. A dominant strategy is one that is never worse, whatever the world does. A mixed strategy is deliberate randomisation, used when no pure action is safe. Strategic flexibility is the economic value of keeping future choices open. The irreversibility premium is what a one-way door must pay, and sequential commitment stages the spending.
Dominant Strategy
The One Unconditional Decision Rule
A dominant strategy has a precise definition, and the precision matters. It pays at least as well as every other alternative, under every possible state of nature. And it pays strictly better under at least one of them. Put plainly, whatever the market or a competitor does, this option is never worse. Notice what the definition never mentions, probabilities. So if a dominant strategy exists, the analytical work is essentially finished. No likelihoods to estimate, no expected values to compute, no trade-offs to weigh. You simply take it. This is one of very few places in business analysis where the rule is unconditional. That is why a careful analyst scans any payoff matrix for dominance first. Removing a dominated option shrinks the problem, and the conclusion survives errors in your probabilities.
Worked Example
Finding the Dominated Alternative
Here is that check in practice. A firm can lease equipment, buy it, or outsource the work. Three states of the economy are in play, recession, stable, and boom. Leasing pays eighty thousand, one hundred and forty thousand, then two hundred thousand. Buying pays fifty thousand, one hundred and fifty thousand, then two hundred and fifty thousand. Outsourcing pays sixty thousand, one hundred and ten thousand, then one hundred and sixty thousand. Now compare leasing with outsourcing, column by column. Eighty beats sixty, one hundred and forty beats one hundred and ten. And two hundred beats one hundred and sixty, so leasing dominates outsourcing outright. Outsourcing can be struck out immediately, with no further analysis needed. Buying pays more when the economy is stable or booming, leasing pays more in a recession. Neither dominates, so that choice still needs probabilities and a view on risk.
Mixed Strategies
When Being Predictable Is the Problem
Sometimes nothing dominates, and the reason is a competitor. Your best action depends on what a rational rival will do. In that setting, the optimal behaviour can be to randomise. Game theory calls this a mixed strategy. If a competitor can predict your action, they will pre-empt it. Choose each action with a calibrated probability, and you become unpredictable. Unpredictable, while still maximising your long-run expected payoff. A retailer that always discounts at weekends is trivially easy to undercut. Tax authorities and quality inspectors randomise visits, which is what makes compliance the safest response. Negotiators mix firm and flexible postures, and bidders randomise their markup across repeated auctions. A mixed strategy is not indecision dressed up as analysis. It is a computed distribution over actions, the unique allocation that leaves your opponent unable to exploit you.
Strategic Flexibility
The Economic Value of Keeping Options Open
Classical expected value analysis treats every decision as final. But many alternatives leave the door open to future choices. Strategic flexibility is the value of preserving those options. It is the practical core of real options thinking, and it has three building blocks. The option to defer lets you wait for more information before acting. The option to expand or contract lets you scale up on good news, or pull back on bad. The option to abandon lets you exit at limited cost if conditions sour. Take a logistics firm choosing between buying a truck fleet and leasing one. Buying has an expected value of one point two million dollars, leasing one point zero five. On paper the purchase wins by one hundred and fifty thousand dollars. Now add a forty per cent chance of an electric vehicle mandate, forcing a switch within eighteen months. The lease preserves the right to abandon, renew, or upgrade, and that option value can easily exceed the gap.
Irreversibility
The Premium a One-Way Commitment Must Clear
Now the sharpest version of the same idea. Take two alternatives with identical expected values. One can be undone, the other cannot. The reversible one is strictly better, because it preserves the option to change course. So an irreversible commitment has to earn its place. Choose it only when its extra expected value covers the optionality you give up. That gap has a name, the irreversibility premium. It is expected value of the irreversible option, minus that of the reversible one, minus the options preserved. And here is the part managers routinely miss. The greater the uncertainty about the future, the larger the premium must be. Irreversibility hides in sunk capital, in long contracts, and in reputations you cannot quietly walk back. It also hides in headcount, in culture, and in regulatory licences that are costly to unwind.
Sequential Commitment
Staging the Decision Over Time
Sequential commitment turns one big decision into a staircase, staged over time. The classic product launch version has three stages. Stage one is a test market, low risk in a small geography. What it buys is real customer response data. Stage two is a regional launch, with medium risk in a single region. That yields scaling and logistics data a test market cannot give you. Stage three is the national launch, high risk across the full footprint. At every stage the firm can continue, modify, or abandon. Heavy commitments only happen after favourable information is already in hand. But the trade-offs are real, and you should say them out loud. Staging is slower, competitors may catch up, and per-unit costs stay higher until full scale. The strategic question is whether the risk reduction is worth those costs.
Case Study 6.2
Full Commit Versus Pilot First
Now put all of it together. Two competing firms face the same new manufacturing technology. Firm A commits the full investment immediately, Firm B runs a pilot first. Firm A puts forty million dollars in up front. Firm B spends six million on the pilot, then thirty-four million more only if it scales. Both start with the same prior, a fifty-five per cent chance of market success. Both earn a hundred and twenty million if the market succeeds. If it fails, Firm A loses thirty million in sunk capacity. Firm B loses only ten million, a limited write-off. The pilot is informative, but it is not perfect. A good signal appears eighty per cent of the time when the market will succeed. And twenty-five per cent of the time when it will fail, so it is evidence, not proof.
Strategy A
Firm A - One Chance Node, No Way Back
Start with Firm A, because its tree is trivial. There is one chance node, and nothing after it. The full forty million is committed before any market information arrives. So expected value is the probability weighted payoff, less the up-front cost. Fifty-five per cent of a hundred and twenty million is sixty-six million. Forty-five per cent of a thirty million loss is thirteen and a half million. Sixty-six, minus thirteen and a half, minus forty, leaves twelve and a half million. So Firm A's strategy is worth twelve and a half million dollars. That is positive, so a full commit is worth doing on its own terms. What it cannot do is limit the downside. Once the capacity is built it is sunk, and a bad market costs the whole thirty million.
Strategy B
Bayes' Rule, Then Roll the Tree Back
Firm B's tree has three layers, and that is the whole point. First the pilot outcome, then the scale or abandon decision, then the market outcome. The pilot accuracies feed Bayes' rule. The chance of a good signal is fifty-five per cent times eighty, plus forty-five times twenty-five. That comes to roughly fifty-five per cent. Given a good signal, the chance of success climbs to about eighty per cent. Those are rounded, the exact figures are fifty-five point two five, and seventy-nine point six four. Now roll the tree back, from right to left. After a good signal, scaling gives eighty per cent of a hundred and twenty million. Plus twenty per cent of a ten million loss, less the thirty-four million scale cost. Ninety-six, minus two, minus thirty-four, leaves sixty million. Abandoning is worth zero, so Firm B scales, and sixty million carries back. Fifty-five per cent of sixty million, less the six million pilot, is about twenty-seven million, an upper bound.
The Verdict
Why the Flexible Strategy Wins
Compare the two strategies side by side. Firm A is worth twelve and a half million, with a worst case of minus thirty. Firm B's upper bound is about twenty-seven million, and a conservative reading puts it near nineteen. Take the conservative figure, and Firm B still beats Firm A comfortably. Three forces drive that, and they are worth naming separately. First, the option to abandon caps Firm B's exposure at six million on a bad signal. Firm A has no exit, and absorbs the full thirty million downside. Second, a good pilot signal lifts the chance of success from fifty-five to roughly eighty per cent. So Firm B scales only when the evidence is strong. Third, most capacity is built after the pilot, so the loss profile is asymmetric. Small downside on bad news, full upside on good. The pilot costs more per unit and reaches scale more slowly, and still wins. That is the whole argument of this section, compressed into one comparison.
About Quantitative Analysis at UPI Study
Quantitative Analysis (Math 250) is a self-paced online course of 55 video lessons in 13 chapters, from decision analysis and simulation to inventory and queuing models, project scheduling, linear programming and transportation models. It carries a NCCRS credit recommendation for 3 semester credits.
Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Enroll now → Course details Download the syllabus All Math courses Every course with PRO →
Compare: Statistics vs Quantitative Analysis
Why take it here
Why students take Quantitative Analysis with UPI Study
Where the credit goes
Universities that work with NCCRS
Quantitative Analysis carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Thomas Edison State UniversitySigned partner
University of the PeopleSigned partner
Charter Oak State CollegeSigned partner
Strayer UniversitySigned partner
Columbia Southern UniversitySigned partner
Indiana TechSigned partner
Penn StateWorks with NCCRS
Southern New Hampshire UniversityWorks with NCCRS
SUNYWorks with NCCRS
+many more cooperating universitiesFind my college →Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.
FAQ
Questions students ask us
About Business Decision Strategies
What is a dominant strategy in business decision-making?
When should a business use a mixed strategy?
What is the irreversibility premium in strategic decisions?
Is it better to run a pilot first or commit fully to a new investment?
About the Quantitative Analysis course
How many college credits is Quantitative Analysis?
How much does Quantitative Analysis cost and what is included?
Which universities accept UPI Study credits?
How long does Quantitative Analysis take?
How are the exams done?
How do my credits reach my college?
Ready for the full course? Earn 3 credits in Quantitative Analysis.
55 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.
3 credits · NCCRSEnroll →