Parthasarathy's Theorem¶
A minimax theorem for bounded unit-square games with finitely many curve discontinuities, obtaining a mixed value when one player is restricted to Lebesgue-absolutely-continuous strategies.
Core Idea¶
Parthasarathy's theorem treats a continuum zero-sum game whose payoff need not be continuous everywhere. Boundedness and confinement of discontinuities to finitely many continuous curves keep the irregularity controlled.
The key asymmetry is strategic: one player may use any probability measure, while the other must use a distribution absolutely continuous with respect to Lebesgue measure. Under those exact conditions, the max–inf and inf–max expected payoffs coincide.
Structural Signature¶
Sig role-phrases:
- Unit-interval action spaces — Provide X=Y=[0,1]. It is action domain. Counterfactual: Changing compact spaces requires another theorem.
- Bounded payoff kernel — Maps action pairs to a finite zero-sum payoff. It is game rule. Counterfactual: Unbounded kernels can defeat integrability and compactness arguments.
- Finite discontinuity curves — Localize noncontinuity along graphs of continuous functions. It is regularity exception. Counterfactual: Arbitrary dense discontinuities are outside the result.
- Arbitrary mixed measures — Give one player the full probability-measure space. It is first strategy class. Counterfactual: Pure strategies are included as point masses.
- Absolutely continuous measures — Restrict the other player relative to Lebesgue measure. It is asymmetric strategy class. Counterfactual: Allowing arbitrary point masses changes the theorem.
- Expected payoff and extrema — Integrate the kernel and compare max-inf with inf-max. It is value criterion. Counterfactual: Pointwise payoff does not establish a mixed value.
What It Is Not¶
- It is not the unrestricted finite-game minimax theorem.
- It does not allow both players arbitrary measures under the same conclusion.
- It is not a pure-strategy saddle-point result.
- Arbitrary discontinuities are not covered.
- Closest near-miss. Von Neumann's finite-game theorem allows all probability vectors; Parthasarathy handles this continuum/discontinuity setting by restricting one strategy class.
Scope of Application¶
- Game theory. Establishes values in discontinuous continuum games.
- Minimax analysis. Shows how strategy restrictions restore equality.
- Probability measures. Uses absolute continuity and integration.
- Economic theory. Models randomized continuous actions under density constraints.
Clarity¶
State action spaces, payoff sign, boundedness, every discontinuity curve and continuity domain, measure spaces, which player is restricted, reference Lebesgue measure, expected-payoff integral, and max/inf attainment convention.
Manages Complexity¶
The theorem balances payoff irregularity against a restriction preventing one player from concentrating mass on problematic thin sets.
Abstract Reasoning¶
- Verify unit-interval action domains.
- Prove payoff boundedness and locate discontinuities.
- Confirm finitely many continuous graph curves.
- Assign full and absolutely continuous strategy classes correctly.
- Apply the exact minimax equality without extending it beyond hypotheses.
Knowledge Transfer¶
The result transfers only when transformed domains, reference measures, regularity of exceptional sets, boundedness, and allowed strategy classes preserve the proof's measure-theoretic structure.
Examples¶
Canonical¶
A bounded kernel on [0,1]² has jumps only on finitely many continuous graphs; X uses any probability measure and Y uses a density, so the theorem equates max over X then inf over Y with the reverse order.
Mapped back: domain → unit square; payoff → bounded; discontinuities → finite curves; X → all measures; Y → absolutely continuous; result → minimax equality.
Applied / In Practice¶
If both players may use arbitrary point masses against a discontinuous kernel, a displayed counterexample can lack a value; ordinary compactness alone does not restore the theorem.
Mapped back: both strategy classes → all measures; restriction → absent; value → may fail.
Structural Tensions¶
T1 — Strategy Generality versus Existence Of Value. Forbidding one player’s singular measures narrows play but restores minimax under controlled discontinuities.
Diagnostic: Which strategic conclusions depend on the absolute-continuity restriction?
T2 — Localized Discontinuity versus Integration Smoothing. Curve jumps can be negligible to densities while point masses can target them.
Diagnostic: Can either player's measure concentrate on the exceptional set?
Structural–Framed Character¶
Parthasarathy's Theorem is structural as a restricted-strategy minimax equality and framed by measure and discontinuity hypotheses.
Structural Core vs. Domain Accent¶
The core is action spaces, payoff, strategy measures, expectation, and extrema. Game theory supplies zero-sum interpretation; analysis supplies absolute continuity and regularity.
Instantiates / Related Primes¶
This entry is a kind of Minimax Theorem.
-
Approved root. No reviewed parent entails this minimax result.
-
Related — von Neumann minimax theorem, mixed strategy, absolute continuity, zero-sum game, and saddle value. They provide ancestor, carrier, restriction, setting, and conclusion.
Relationships to Other Abstractions¶
Current abstraction Parthasarathy's Theorem Domain-specific
Parents (1) — more general patterns this builds on
-
Parthasarathy's Theorem is a kind of Minimax Theorem Domain-specific
Parthasarathy's Theorem is a strict kind of Minimax Theorem: it is a minimax existence result for a specified class of discontinuous unit-square games.Every reviewed Parthasarathy's Theorem instance satisfies Minimax Theorem because it is a minimax existence result for a specified class of discontinuous unit-square games. The child adds the domain-specific restrictions stated in its frozen identity. Minimax Theorem is broader and can occur without the restrictions that define Parthasarathy's Theorem.
Hierarchy path (1) — routes to 1 parentless root
- Parthasarathy's Theorem → Minimax Theorem → Minimax Strategy → Optimization
Neighborhood in Abstraction Space¶
Parthasarathy's Theorem sits in a moderately populated region (48th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Combinatorial Optimization & Game Problems (12 abstractions)
Nearest neighbors
- Maximising measure — 0.87
- Radon Measure — 0.87
- Graphical Game Theory — 0.86
- Path Integral Formulation — 0.86
- Topological Dynamical System — 0.86
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
- Von Neumann minimax. Tell: Classically covers finite games without this curve-discontinuity setup.
- Sion minimax. Tell: Uses convexity and semicontinuity hypotheses.
- Pure-strategy equilibrium. Tell: Does not follow from a mixed value.
- Absolute-continuity of payoff. Tell: Is different from absolute continuity of a strategy measure.
References¶
- Frozen Wikipedia discovery revision: https://en.wikipedia.org/wiki/Parthasarathy%27s_theorem (revision 1297519748).
The frozen Wikipedia revision is discovery provenance. The retained source set was reviewed for identity, formal or operational relation, and scope. The encyclopedia's structural synthesis is bounded to those claims; a thin authority surface is recorded as a nonblocking source-strengthening repair rather than concealed.