Expected Payoff

You already know Expected Value - probability-weighted average of outcomes. And you know Game Theory - your best move depends on someone else's move.

Expected Payoff is what happens when you combine them. If a competitor doesn't play one strategy every time but mixes - sometimes undercutting, sometimes holding price - then your payoff from any strategy you pick isn't a fixed number. It's an expectation.

Formally: if the other player plays strategy A with probability p and strategy B with probability (1-p), your expected payoff from your strategy S is:

E[payoff(S)] = p × u(S, A) + (1-p) × u(S, B)

where u(S, A) is the value you get from your Utility Function when you play S and they play A.

This is the bridge between 'I know what outcomes are possible' (Expected Value) and 'I know the other side is optimizing too' (Game Theory). It gives you a concrete number to compare strategies even when you face genuine strategic uncertainty.

Operators face mixing opponents constantly:

• •Competitive Pricing: A rival alternates between holding ARR targets and running aggressive discounts to grab Market Share. Your Revenue forecast depends on which one shows up.
• •Vendor Negotiations: A supplier sometimes holds firm on pricing, sometimes caves. Your Cost Structure planning needs a number, not 'it depends.'
• •Hiring Targets: A strong candidate might accept your offer or take the Outside Option. Your Time-to-Fill and Hiring Targets hinge on how you bid.

Without expected payoff, you either plan for only one scenario (fragile) or throw up your hands and guess (expensive). Expected payoff turns the P&L line items that depend on other people's moves into numbers you can actually Budget against, run Sensitivity Analysis on, and hold yourself accountable to.

The operator who computes expected payoffs across competitor scenarios builds a Budget with realistic Variance baked in. The one who doesn't gets surprised every quarter.

Step 1: Map the game

Identify your strategies and the other player's strategies. You already did this in Game Theory - this is the payoff matrix.

Step 2: Estimate mixing probabilities

This is where operator judgment matters. You rarely know exact probabilities, but you can estimate from:

• •Pipeline data and history: The competitor undercut pricing in 3 of the last 8 quarters. Start with p = 0.375.
• •Incentive analysis: Their Cost Structure or Capital Structure might force discounting when they miss targets. Weight accordingly.
• •backtesting: Run your model against historical periods and see if your probability estimates would have predicted actual outcomes.

Step 3: Compute expected payoffs for each of your strategies

For each of your possible moves, multiply each outcome by the probability the opponent plays the corresponding move, then sum.

Step 4: Choose the strategy with the highest expected payoff

Or - and this is the important part - if no single strategy dominates, compute your own mixing probabilities. You might find that randomizing between two pricing tiers yourself yields a higher expected payoff than committing to either one. This is equilibrium thinking: if both sides are mixing optimally, neither can improve by changing their mix.

The key mental shift

In Expected Value, uncertainty is about the world (will this deal close?). In Expected Payoff, uncertainty is about a strategic actor who is also optimizing. The probabilities aren't nature rolling dice - they're a rational competitor choosing a mix that makes you indifferent between your options.

Use expected payoff when:

• •No Dominant Strategy exists. If one move always wins regardless of the competitor's action, just play it. Expected payoff matters when your best move changes depending on what they do.
• •You face a repeated interaction. Pricing wars, recurring Vendor Negotiations, quarterly Budget cycles against competitors. The mixing probabilities stabilize over repeated play.
• •The stakes justify the math. A $50K decision? Estimate and move. A $2M ARR pricing decision against an aggressive competitor? Compute the expected payoffs.
• •You need to justify resource allocation. When your CFO asks why you're splitting Marketing Spend between two channels instead of concentrating, expected payoff against competitor responses is the answer.

Don't bother when:

• •The other side has a clear Dominant Strategy (they'll always do X, just plan for X)
• •You have an Informational Advantage that resolves the uncertainty (you know their move before deciding)
• •The decision is irreversible and one-shot with catastrophic downside - here Risk Tolerance and Tail Risk matter more than the expectation

Expected Payoff lives at the intersection of your three prerequisites. Expected Value gave you the machinery - probability-weighted averages - but applied it to scenarios where outcomes are uncertain because of the world (will this deal close, will this market grow). Game Theory introduced a different kind of uncertainty: outcomes depend on what a strategic actor chooses. Utility Function told you that the 'value' you're weighting isn't always dollars - it's whatever your internal scoring system says matters. Expected Payoff fuses all three: you compute probability-weighted utility over the strategies another rational player might use.

From here, the path branches. Zero-sum Game is the special case where your gain is the opponent's loss - expected payoffs sum to zero, which radically simplifies the math. Dominant Strategy is what you hope to find when computing expected payoffs - a move that wins regardless of the opponent's mix, making the calculation unnecessary. Outside Option sets a floor on your expected payoff - if no strategy in the game beats walking away, you walk. And Sensitivity Analysis is the operational discipline that keeps expected payoff honest: stress-test the mixing probabilities, find where your decision flips, and decide if your estimate has enough margin of safety to act on.