Independent scholarship · Open accessThe monthly digest

Indexed text attachment

evolutionary-game-theory-in-higher-categories.tex

Source for Evolutionary Game Theory in Higher Categories

View paper · Download original file

Lines 481–580 of 2037. Source text is displayed literally.

481\Scl[Hofbauer, Schuster and Sigmund 1979; \cite{Wei95}] The three conditions are equivalent.482\end{theorem}483 484\noindent The source presents the equivalence as evidence that the definition is not arbitrary. In the present485language two thirds of it is one line: by \Cref{exer:biaffine},486\begin{equation}\label{eq:ray}487\Q_x(y_\epsilon)=\epsilon\,\bigl(\pay{x}{y_\epsilon}-\pay{\mu}{y_\epsilon}\bigr),488\end{equation}489so the barrier inequality along the ray from $x$ towards $\mu$ \emph{is} local superiority along that ray. What490remains is uniformity in $\mu$, which is compactness of the simplex, and the passage from (i) to a491neighbourhood, which uses that $\epsilon\mapsto\pay{x}{y_\epsilon}-\pay{\mu}{y_\epsilon}$ is affine with value492$\pay{x}{x}-\pay{\mu}{x}\ge0$ at $0$ and slope governed by the second clause of (i).493 494\begin{example}[Hawk--Dove]\label{ex:hd}495With $A$ as in \Cref{exer:hd} and $V<C$, let $x=(p,1-p)$ with $p=V/C$ and $y=(q,1-q)$. The payoff advantage of496Hawk over Dove against $y$ is $(V-qC)/2$, so497\[ \Q_x(y)=(p-q)\,\frac{V-qC}{2}=\frac{C}{2}\,(p-q)^2 . \]498The mixed state is globally, and not only locally, superior. \chk{S1}499\end{example}500 501\begin{example}[Drift]502For $A=\begin{psmallmatrix}2&1\\1&1\end{psmallmatrix}$ and $x$ the second pure strategy,503$\Q_x(y)=1-(1+q^2)=-q^2<0$ when $y$ plays the first strategy with probability $q>0$. The equilibrium is not504merely not superior; it is inferior to every nearby state. \chk{S4}505\end{example}506 507\section{Properties}508 509The source collects five facts \src{Thms.~1--5}. Each is a statement about $\Q$.510 511\begin{theorem}\label{thm:essprops}512\Scl[Bishop and Cannings 1976; Haigh 1975]513\begin{enumerate}[label=(\alph*)]514\item A game need not have an ESS: Rock--Paper--Scissors has none. \chk{S2}515\item A weakly dominated strategy is not an ESS. \chk{S3}516\item If $x$ is an ESS and $y\ne x$ is a Nash equilibrium, then $\supp y\not\subseteq\supp x$.517\item A game with $n$ strategies has at most $\binom{n}{\lfloor n/2\rfloor}$ ESS.518\item An ESS in the interior of the simplex is the only ESS, and it is globally superior.519\item A quasi-strict equilibrium $x$ is an ESS if and only if $z^{\T}Az<0$ for all $z\ne0$ with520$\supp z\subseteq\supp x$ and $\sum_s z_s=0$.521\end{enumerate}522\end{theorem}523\begin{proof}524(a) By \Cref{cor:uniform} the uniform state $u$ is an equilibrium, and it is the only one. For a pure strategy525$e$, $\pay{u}{e}=0=\pay{e}{e}$ and $\pay{e}{u}=0=\pay{u}{u}$, so both clauses of (i) fail.526(b) If $\mu$ weakly dominates $x$ then $\pay{\mu}{x}\ge\pay{x}{x}$ and $\pay{\mu}{\mu}\ge\pay{x}{\mu}$, so527$\Q_x(\mu)\le0$ with $\mu$ an alternative best reply.528(c) If $\supp y\subseteq\supp x$ then $y$ is a best reply to $x$, since every pure strategy in the support of529$x$ is. The second clause of (i) gives $\pay{x}{y}>\pay{y}{y}$, contradicting that $y$ is an equilibrium.530(d) By (c) the supports of distinct ESS form an antichain of subsets of $S$; apply Sperner's theorem.531(e) If $x$ is interior, every $y$ is a best reply to $x$, so $\Q_x(y)>0$ for all $y\ne x$; then no other state532is an equilibrium by (c).533(f) For $y$ near $x$ write $y-x=z+w$ with $z$ supported in $\supp x$; quasi-strictness makes the first-order534term in $w$ strictly positive, and for $w=0$, $\Q_x(y)=-z^{\T}Az$ because $z^{\T}Ax=0$.535\end{proof}536 537Criterion (f) is the one used by the companion script to decide evolutionary stability exactly.538 539\section{Sets, and why they are forced}\label{sec:essets}540 541A population may be stable as a whole while drifting inside a set of equivalent states. The source surveys542Thomas's evolutionarily stable sets and three variants due to Balkenborg and Schlag \src{\S2.3}. We use the543following form.544 545\begin{definition}\label{def:esset}546\Scl[Thomas 1985; \cite{Wei95}] A non-empty closed set $L\subseteq\Dist(S)$ is an \term{evolutionarily stable547set} if every $x\in L$ has a neighbourhood $U$ with $\Q_x(y)\ge0$ for all $y\in U$, and $\Q_x(y)>0$ for548$y\in U\setminus L$.549\end{definition}550 551\noindent A one-point set $\{x\}$ is an evolutionarily stable set exactly when $x$ is an ESS, and every point552of an evolutionarily stable set is a Nash equilibrium.553 554\begin{theorem}[Stability is a property of sets]\label{thm:esset}555\Sob Let $f\colon(S,A)\to(T,B)$ be a payoff-preserving surjection and let $L\subseteq\Dist(T)$ be an evolutionarily556stable set. Then $f_*^{-1}(L)$ is an evolutionarily stable set of $(S,A)$. If $f$ is not injective and $x$ is557an ESS of $(T,B)$ whose support meets the set of duplicated strategies, then no point of $f_*^{-1}(x)$ is an558ESS.559\end{theorem}560\begin{proof}561By \Cref{prop:push}, $\Q_\sigma(\mu)=\Q_{f_*\sigma}(f_*\mu)$ for all $\sigma,\mu\in\Dist(S)$. Let562$\sigma\in f_*^{-1}(L)$ and let $U$ be the neighbourhood of $f_*\sigma$ from \Cref{def:esset}. Then563$f_*^{-1}(U)$ is a neighbourhood of $\sigma$ on which $\Q_\sigma\ge0$, with strict inequality off564$f_*^{-1}(L)$; and $f_*^{-1}(L)$ is closed and non-empty. For the second statement, the fibre over $x$ contains565a segment through each of its points, on which $\Q_\sigma$ vanishes identically.566\end{proof}567 568\begin{example}569Duplicate Dove in the Hawk--Dove game. The fibre over the ESS is the segment570$L=\{(p,s,1-p-s)\}$, and $\Q_\sigma(\mu)=\tfrac C2(p-\mu_{\mathrm H})^2$ for $\sigma\in L$: zero exactly on $L$,571positive elsewhere. \chk{S6}572\end{example}573 574\Cref{thm:esset} explains a development that the source records historically. Duplicating a strategy changes575nothing about the biology or the sociology of a model; it is a change of description. A solution concept576that is destroyed by it cannot be the right concept, and the least repair is to pass from points to sets.577The three variants of Balkenborg and Schlag differ in the order of quantifiers over the neighbourhood (one578threshold for each invader, for each incumbent, or for all) and so in how they behave when $L$ is not compact579in the relevant topology; for the closed sets of \Cref{def:esset} we do not need the distinction.580