Dominated Strategy
A strategy $m_i\in\Delta(S_i)$ is strictly dominated if there exists $\sigma_i\in\Delta(S_i)$ such that $$ g_i(\sigma_i,t_{-i})>g_i(m_i,t_{-i}) \quad\text{for all } t_{-i}\in S_{-i}. $$ It is weakly dominated if there exists $\sigma_i$ such that the same inequalities are weak for all $t_{-i}$ and strict for at least one $t_{-i}$.
For pure strategies, this specializes by identifying $s_i\in S_i$ with the Dirac mass at $s_i$.
Lean scope
Lean's StrictlyDominates / WeaklyDominates formalize domination between
pure strategies ($s,s' : G.\text{strategy}\,i$). The mixed-strategy
domination above — domination by some $\sigma_i\in\Delta(S_i)$ — is the more
general MFoGT notion, of which the Lean relations are the pure-strategy special
case. Mixed-strategy domination is not yet formalized, so this node stays
pending until the general notion is ported.
References
- [MFoGT, Section 1.3.2] Laraki, Renault, and Sorin, Mathematical Foundations of Game Theory. Strictly dominated and weakly dominated pure strategies.
- [MFoGT, Definitions 4.3.2 and 4.3.5] Laraki, Renault, and Sorin, Mathematical Foundations of Game Theory. Strict and weak domination for mixed strategies in finite games.
Used by
- Forward Induction
- Iterated Elimination Of Strictly Dominated Strategies
- Perfect Equilibrium And Undominated Equilibrium
- Proper And Perfect Equilibrium Examples
- Strict Domination And Never Best Response (via auxiliary zero-sum game)
- Weak Domination And Completely Mixed Tests (via auxiliary zero-sum game)