How to Read a Nash Equilibrium in a Litigation Game
A plain-English guide to reading a payoff matrix and understanding what an equilibrium strategy actually tells you about a case.
A Nash equilibrium in a litigation game is a pair of strategies — one for the plaintiff, one for the defendant — where neither side can improve its own expected outcome by unilaterally switching to a different move, given what the other side is doing. It is not a prediction of what will happen. It is a description of a stable posture: the point where both sides' choices are mutual best responses to each other under the assumptions fed into the model.
Litigation lends itself to this framing because it is genuinely strategic. Discovery aggressiveness, motion practice, and settlement timing all change in value depending on what the other side does. Juricratic treats these choices as a solvable game so you can check whether your current posture is defensible, or whether a rational opponent could pick it apart. Reading the equilibrium correctly means understanding what it does and does not claim.
What a litigation game actually models
A litigation game reduces a stage of the matter — say, how aggressively to pursue discovery, or whether to make an early settlement offer — to a small set of discrete strategic choices for each side. Each combination of choices produces a payoff for the plaintiff and a payoff for the defendant, usually expressed as an expected value net of litigation cost.
The payoffs come from the same dials that drive any Juricratic simulation: probability of prevailing on a claim or element, estimated damages, litigation cost to reach each stage, and time. The game format adds one thing a single expected-value calculation cannot: it accounts for how the other side's rational choice changes your own best move.
Reading the payoff matrix before you read the equilibrium
Before looking for the equilibrium cell, make sure you understand the matrix itself. Rows are typically the plaintiff's available moves, columns are the defendant's. Each cell holds a pair of numbers: the plaintiff's expected value and the defendant's expected value if that particular combination of moves occurs.
- Litigation is not zero-sum — both sides can lose value simultaneously through fees, delay, and deadweight cost, so a cell's two numbers do not have to sum to a constant.
- A higher number in a cell is better for that party only; compare within a party's own row or column, not across the two parties' numbers.
- The matrix is only as good as the moves it includes — a coarse move set (e.g., only 'aggressive' vs. 'cooperative') will miss a real strategy that sits between the two.
What 'equilibrium' means here — and doesn't
A pure-strategy equilibrium is a single cell where each side's move is already its best response to the other's. A mixed-strategy equilibrium instead prescribes a probability of playing each available move — for example, escalating discovery some fraction of the time and holding back the rest — because no single fixed move is unbeatable given the other side's incentives.
Mixed strategies tend to appear when the underlying game has imperfect information: if your move were fully predictable, the other side could always counter it optimally. Randomizing (in the game-theoretic sense of a considered mixed posture, not literal coin-flipping) removes that exploit. Either way, the equilibrium is a stable point given the assumptions entered — it is not a claim about what will actually happen, and it is not a verdict.
Using the exploitability gap to judge how solid your current strategy is
The equilibrium is most useful as a benchmark. Juricratic can also report the exploitability gap: how much better an opponent could do by best-responding to your actual current strategy, compared to what they could achieve against the equilibrium strategy.
- A large exploitability gap means your current posture leaves value on the table — a well-informed opponent could improve their own outcome meaningfully by countering it.
- A small or zero gap means your posture is close to equilibrium — hard for a rational opponent to exploit even if they know exactly what you plan to do.
- Track the gap over time rather than once — as facts develop and dials are updated, a posture that was near-equilibrium early in the matter can drift and become exploitable later.
Common misreadings to avoid
The most common mistake is treating the payoff in the equilibrium cell as a guaranteed outcome. It is an expected value conditioned on the dials you entered, not a promise. The second most common mistake is forgetting that the equilibrium moves if you change the underlying assumptions — a different damages estimate or cost dial can shift which cell is the equilibrium entirely.
- Don't confuse a Nash equilibrium with a historical base rate of how similar cases have actually settled — the model is descriptive of the assumptions you gave it, not of past outcomes.
- Don't assume the equilibrium is unique — some games have more than one equilibrium, and which one is 'selected' in practice can depend on factors outside the formal model.
- Don't skip re-solving the game after new evidence changes a probability or cost dial — an equilibrium computed on stale assumptions is stale advice.
- Does an equilibrium tell me the best move to make right now?
- It tells you the mutual best-response point given the payoffs you entered — a useful benchmark, not an instruction. Whether the assumptions behind those payoffs are right is a separate judgment call you still have to make.
- What if my case doesn't fit neatly into a matrix of a few discrete moves?
- Most real matters unfold over several sequential stages rather than one simultaneous choice. In that case a sequential decision tree, solved stage by stage, is usually a better fit than a single payoff matrix.
- Why would a mixed-strategy equilibrium appear in a litigation game?
- It typically shows up when the game has imperfect information — if one side's move were fully predictable, the other side could always counter it optimally, so an unpredictable (mixed) posture removes that exploit.
This page is an educational explainer, not legal advice, and creates no attorney–client relationship. Juricratic is a simulation engine: every probability-like figure is a dial you set, not a calibrated prediction. Verify every rule, deadline, and figure against the authorities and orders that govern your matter.
Stop estimating one number at a time.
Juricratic models the whole matter as a solvable game and runs it thousands of times — so the settlement value, the risk, and the optimal line all move together when the facts do.
Request access →