I have published fifty-eight case studies, with twelve more in the pipeline, and I will keep writing them. The case method is a strong way to teach judgment about a decision at a moment in time. A good case puts students in the shoes of a manager facing a hard decision, gives them the facts that manager had, and asks what they would do. Through the discussion, students learn to form a point of view from incomplete evidence and to defend it in front of their peers.
A case teaches one decision, taken once, and argued from an outsider’s point of view. In real life, management is a sequence of decisions, each taken under the consequences of the ones before, and from the inside. The case method handles sequential and dynamic decisions poorly. Instructors write A, B, and C cases to build a multi-act drama, but teaching systemic and dynamic decision making calls for a different method. In this post, I discuss the limitations of case studies and describe the situations in which a simulation teaches more effectively, with examples from marketing, operations, supply chain management, and capital budgeting.
Where the case method falls short
Consequences stay on paper. The student recommends a price, a launch, or a plant closure, and the discussion ends. Nobody lives with the result. In practice, the price teaches customers to wait for the next promotion, the launch drains the cash the firm needed next quarter, and the closure surfaces as a quality problem eighteen months later. A B case can describe those consequences, but students remember consequences they experience far better than ones they read about.
The clock stops after one decision. The manager in a case decides once. Most management problems unfold over many periods, and the hard part is timing: when to raise capital, when to add capacity, when to stop a promotion. A sequence of cases resets the clock each time, so the student never carries a position forward into the next decision.
The student watches from outside. The student reads about the manager, with the full exhibit set in view and the hindsight the case writer built in. Managers do not get the exhibit set. They sit in a cockpit, facing a dashboard with conflicting signals, advisers who disagree, and a rival deciding at the same time. The experience is visceral, closer to bodystorming than to brainstorming.
The answer is set in advance. A case offers a limited set of choices, and the discussion moves toward the one the case writer prepared. The instructor knows the destination and tends to steer the discussion toward it, and the strongest students learn to argue for the prepared answer. Students do not discover unintended or unknown outcomes, because the outcomes were decided before the class began.
Everyone gets the same experience. Every student receives the same facts and the same result, so the class learns one route through one situation. The instructor cannot compare thirty results, show which decisions separated the strong performers from the weak, or replay a decisive choice the other way and watch what changes.
When to use simulations
The limitations of case studies point to the situations where simulations teach better and deeper lessons.
Decisions that pile up. A founder who raises capital in quarter two on weak terms owns less of the company in quarter eight, and every decision in between is made under that constraint. A retailer who runs promotions in three consecutive quarters has trained customers to wait, and the fourth quarter shows it. This is the normal condition of management. In a simulation, the student carries a position forward and meets it again. In DashKart, our founder’s-seat simulation, the sequence of raises decides what the founder keeps, and students read that lesson off their own cap table at the end of eight quarters.
Cause and effect separated in time. Many serious management errors come from acting on a signal that arrives late. Capacity ordered today arrives in a year. Quality problems from a delegation decision surface two to four periods after the decision. A case can tell students that lags exist; a simulation makes them order the capacity, wait, and watch the market move before it arrives. In Agency, our simulation of a law firm handing work to AI, only the errors that are caught enter the log, and the uncaught ones surface later, so the dashboard looks healthiest when the firm is most exposed. Students learn to distrust a clean dashboard, a lesson a case study struggles to teach.
Lessons about how it feels to decide. Some lessons concern what it feels like to decide: the pull of a familiar metric, the temptation to push a colleague who is not cooperating, the wish to wait for certainty while the storm approaches. A case discussion examines those pulls from a distance, and a simulation puts the student inside them. In YieldForge, students play the first two weeks on instinct and the utilization gauge. In week three, the game shows them what an hour of machine time was worth and reprices the choices they already made. Students experience the lesson about opportunity cost viscerally, and lessons learned that way tend to stick.
Systemic decisions. Many valuable lessons in management concern the whole rather than the part: how a decision on price moves cost, how a decision on staffing moves trust, how an efficient operation loses the slack it needs in a disruption. A case can describe a system; a simulation runs one, and the student watches the parts interact to produce the behavior of the system. Simulations have their clearest advantage here, so I go deeper in the next section.
Why simulations teach systems so well
Jay Forrester built system dynamics at MIT in the 1950s to explain why well-run companies behaved in ways their managers did not intend. John Sterman carried that work into the classroom with the People Express simulator and the beer game, which have taught two generations of students that the structure of a system produces its behavior. Most serious business simulations carry that inheritance. The phenomena system dynamics identified are the ones a case struggles with and a simulation handles directly.
Stocks and flows. Cash, capacity, reputation, and skill are stocks. They fill and drain slowly, and a decision that changes a flow today changes the stock much later. Watch a reputation stock decline for three quarters after a service cut, then try to rebuild it, and students understand brand equity in a way a definition does not give them. In MarketMoves, awareness decays about eight percent a quarter and brand spending rebuilds it, so students learn the arithmetic of a stock by managing one.
A case can describe a system; a simulation runs one.
Delays. Capacity takes a quarter to build. New hires take a period to arrive and longer to train. Managers overshoot in the gap between a decision and its effect, and a simulation lets a student overshoot without harming a real firm. In DashKart, new dark stores take a quarter to open, so the second Diwali tests whether the founder learned from the first.
Nonlinearities and capacity limits. Pushing utilization past ninety percent does not add output in proportion; it lengthens waiting times sharply. Every operations course teaches this, and it is hard to teach on a whiteboard. In YieldForge, the plant is full and the money is missing, and students discover that a full plant can lose money because the wrong orders filled it. In Turnaround, capacity growth above twelve percent in a period sets off a hiring wave and a training drag that raise wage costs for the rest of the game.
Feedback and rivals. A decision changes the environment that the next decision faces. A rival who studies your strategy and counters it teaches the lesson in Mike Tyson’s famous line: “Everybody has a plan until they get punched in the mouth.” In MarketMoves, the digital challenger classifies the player’s strategy every two quarters and picks a counter, so a plan that changes lever by lever gets punished, and a plan that imitates the rival stays a step behind.
Externalities. A decision that looks free in one department costs another. Closing a loss-making service line at a hospital frees beds and nurses, and a year later admissions, trust, and share fall together, because the loss-making line fed the profitable ones. A case can assert that interdependence. A simulation makes the student close the line, watch the year unfold, and live with the consequences.
Let us look at four examples of how this plays out in a course. In marketing, promotions borrow from the future and customers learn to wait, a stock-and-delay problem that eight quarters of play teach more directly than a retail case can. In operations, opportunity cost and the moving bottleneck are nonlinear, and students grasp them when the bottleneck moves under their own decisions. In supply chains, the beer game has shown for more than half a century that four reasonable people with a delay between them will produce a bullwhip, and causing one teaches more than hearing about one. In capital budgeting, the value of flexibility and the cost of waiting are real options that discounted cash flow handles poorly. In CAPARC, students commit capital to satellites that cannot be changed after launch, decide whether to pay for a repair port they may never use, and learn what an option costs and when it pays.
Where to put simulations in your courses
I do not argue that simulations should replace cases. They teach a different set of lessons. A case teaches students to reason toward a decision. A simulation teaches them what happens after they decide, and then makes them decide again with that knowledge. The first builds judgment. The second builds the habit of living with consequences and adjusting, which takes up much of a manager’s time between the decisions that get written up as cases.
The practical objection to simulations is that they take too much time to run and the course has no room. That was true when a simulation meant days of play and weeks of rounds, and it is no longer true. The titles we build at Pedagogix take a student one sitting of 25 to 75 minutes before class, and the class session is the debrief, so a simulation costs a course about what a case costs. Students play with advisers who disagree, a rival who counters, and a dashboard that misleads at the worst moment. Each student receives feedback written from the game they played, and the instructor receives the cohort’s results before the debrief begins.
So here is what I ask of instructors. Look at your syllabus and find the lesson that is really about a system, where the true content is a delay, a stock that drains, a bottleneck that moves, a rival that responds, or a cost that falls on another department. In that session, your students are learning a description of the thing when they could be learning the thing itself. Put a simulation there. Keep your cases for the decisions that deserve ninety minutes of argument, and give the systems lesson an instrument built to teach it.
Mohanbir Sawhney is Co-founder & Chairman, Pedagogix.