Skip to content
Tutorials Center
My learning

Topics

Lesson 6 of 37 ยท Games and Strategies

Dominant Strategies

Learn strict and weak dominance, dominant strategy equilibrium and the prisoner's dilemma, and write Python functions that find dominated and dominant strategies.

In some games you do not need to guess what the other player will do. Whatever they choose, one of your options is simply better than another. Spotting such options is the easiest, most reliable tool in game theory, and it requires no assumption about what the opponent believes. This idea is called dominance.

You will learn the difference between strict and weak dominance, how a game with a dominant strategy for every player resolves immediately (and why the result can be disappointing), and how to find dominated strategies with a few lines of NumPy. We continue with the notation from Normal-Form Games and Strategies.

Strict dominance

Strategy a strictly dominates strategy b for player i if a gives a higher payoff than b against every strategy profile of the other players:

u_i(a, s_{-i}) > u_i(b, s_{-i}) for all s_{-i}.

Then b is strictly dominated. A rational player never plays a strictly dominated strategy, because switching to a is better no matter what the others do. If one strategy dominates every other strategy, it is a strictly dominant strategy, and a rational player plays it.

Take the prisoner's dilemma. Two suspects are questioned separately. Each can stay silent (Cooperate, C) or confess (Defect, D). Payoffs are years of prison, negated so that larger is better:

C (Bob) D (Bob)
C (Ann) (-1, -1) (-3, 0)
D (Ann) (0, -3) (-2, -2)

If Bob stays silent, Ann gets -1 from C and 0 from D, so D is better. If Bob confesses, Ann gets -3 from C and -2 from D, so D is better again. D strictly dominates C for Ann, and the same is true for Bob. So both confess and each serves two years, even though both staying silent would give each only one year.

Python Strict dominance in the prisoner's dilemma

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Notice that the column player's check works on columns of B: Bob compares two of his own options with Ann's action fixed, which means comparing two columns.

Dominant strategy equilibrium

When every player has a strictly dominant strategy, each plays it, and the resulting profile is the dominant strategy equilibrium. It is the most convincing prediction a game can offer, because it needs no belief about the opponents at all.

The prisoner's dilemma shows why that does not mean the outcome is good. The equilibrium (D, D) pays -2 to each, while (C, C) would pay -1 to each. Individually rational behavior produces a collectively worse result. The tension between individual incentive and group benefit appears all over the world: arms races, overfishing, advertising wars. You will meet it again in Pareto Efficiency and Social Dilemmas.

Weak dominance

Strategy a weakly dominates b if a is at least as good as b against every profile and strictly better against at least one. Weak dominance is a softer statement. A player might still play a weakly dominated strategy, because in cases where it ties, nothing is lost.

Python Strict versus weak dominance

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Row 0 weakly dominates row 1 (better in column 0, equal elsewhere), and strictly dominates row 2. A careful player avoids row 2 for certain, but may be unsure about row 1.

Warning: Weak dominance makes order matter when you eliminate strategies repeatedly. Iterated elimination of strictly dominated strategies always gives the same result no matter the order; iterated elimination of weakly dominated strategies can give different results in different orders. See Iterated Elimination of Dominated Strategies.

Dominance in action: the second-price auction

A famous weakly dominant strategy appears in the second-price sealed-bid auction. Every bidder submits a bid privately. The highest bidder wins but pays the second-highest bid. Suppose you value the item at v. Claim: bidding exactly v weakly dominates every other bid. The reason is short: if you bid more than v, you could win at a price above your value and lose money. If you bid less than v, you could lose an auction you would have profitably won. Bidding v never hurts and sometimes helps. We can check this by brute force.

Python Honest bidding weakly dominates

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Because the truthful bid is at least as good against every rival bid and strictly better against some, it weakly dominates all other bids. This property is the reason second-price auctions are favored in mechanism design; Mechanism Design and Auctions returns to it.

Finding dominated strategies in a general game

Let us put everything into one helper that finds all strictly dominated strategies of both players in a bimatrix game. Rows belong to the row player, columns to the column player; for the column player we apply the same row logic to the transpose of B.

Python A dominated-strategy finder

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

A strategy may be dominated not only by another pure strategy but by a mixture of two others, which pure comparison misses. The next lesson shows how to check that.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Strategy a strictly dominates strategy b if...

    Strict dominance requires a strictly higher payoff for every strategy profile of the other players.
  2. 2. Why is the prisoner's dilemma outcome (D, D) considered disappointing?

    Defecting is dominant for both, so (D, D) is the prediction. But each player gets less there than at (C, C), which shows the conflict between individual and joint interests.
  3. 3. What is true of weak dominance?

    Weak dominance allows ties, as long as there is at least one profile where a is strictly better.
  4. 4. In a second-price sealed-bid auction, what bid weakly dominates all others?

    Bidding your true value never loses money and never forgoes a profitable win, so it weakly dominates every other bid.

Summary

  • Strategy a strictly dominates b when a pays more against every profile of the others; a rational player never plays a strictly dominated strategy.
  • A strictly dominant strategy beats all alternatives; if every player has one, the result is a dominant strategy equilibrium.
  • The prisoner's dilemma shows that dominant-strategy play can be worse for everyone than cooperation.
  • Weak dominance allows ties and can depend on the order of elimination.
  • Truthful bidding is weakly dominant in a second-price auction.

Next: Iterated Elimination of Dominated Strategies