Book Summary : Thinking Strategically By Avinash Dixit and Barry Nalebuff
The best way that I think to summarize this book would be to restate the examples. This summarises are not meant to replace the book for those that have not read the book. I think this summary would be very helpful to those who have read the book in their freshman year and just need to revisit and recollect the most important lessons that the authors wanted their readers to get.
Table Of Contents
Part 1
1. Ten tales of strategy
1. Hot hands - basketball and tennis: Players do not have a hot hand. If it is played out over a very long series of events, its probability would be no better than the flip of a coin. But yes game theory does provide an explaination, this can be explained by the interactions of defensive and offensive strategies of the players, which is why it sometimes seems a player has a hot hand.
2. To lead or not lead - Sailboat racing and business: In sailboat racing it is generally seen that the leader will change course in the same direction with that of the follower, because in case the follower is right he is still going to with him in catching the wind and would reduce the chance of loosing due to not catching the wind. But in case the race involves more than 2 racers than his decisions would not just be based on that of the follower. The game becomes very different when there are more than 2 racers. Similarly in business there might be strategic advantage in moving second. Thus it is observed that the smaller companies engage in new innovative and risky ventures and the leaders in that business tends to then follow them.
3. Go directly to jail - prisoner's dilemma, nuclear dismembarment, union bargaining: In this situations when each choses his dominant strategy, would lead to sub-optimal jointly preferred outcome. The problem is interdependence of decisions. If individually sub-optimal strategy is chosen than jointly preferred outcome could be achieved. These games are different as they are not zero-sum games hence one player's gain is not necessary other players losses.
4. Here I stand - Martin Luther, Charles de Gaulle (France to keep England out of EEC), Ferdinand de lesseps ( Suez canal , failed in panama canal) - "An inflexible personaility is not something that can be turned on and off. Although being inflexible can sometimes wear down an opponent and force him to make concessions, it can equally well allow small losses to grow into major disasters" . There is advantage in such a position as it denies the other party to come with a counter offer.
5. Belling the cat - Hostage dilemma, Khrushcher & Stalin, taxi & dispatcher, NY eviction, United Auto Workers - "accordion effect" (where each fold pushes or pulls the other) - Communication and coordination become important in solving this issues.There is another point that must be noticed is that there is generally superiority of punishment over reward.
6. The thin edge of wedge - Import restrictions, Income tax reforms of 1985-86, social interest - When decisions are made one at time could lead to different conclusions, if instead all the cases are clubbed together. Decisions case by case basis vs. together, when done case by case it could lead to demands for various concessions and complications, instead though inflexible decisions taken together can have strategic advantage.
7. Look before you leap - Commitment to job in a particular city, operating system of your computer, frequent flyer program, marriage is expensive to escape. So one should look before one takes a leap.
8. Mix your plans - Randomized strategies, football pass/run, tennis backhand forehand, IRS audit, gillette promotion coupon.
9. Never give a sucker even bet - Guys and Dolls sky masterson & Nathan detroit, at what odds will you take the other bet; market maker must always mention both sides bid/ask before knowing your side. "The strategic insight is that other people's actions tell us something about what they know, and we should use such information to guide our own actions.
10. Game theory can be dangerous to your health - Two economics in Beruit, ignore irrationality and pride at your own perils.
11. Stuck wheel gets the grease - Of course sometimes even gets replaced. There could be stragetic advantage in being like a stuck wheel because it could mean concessions and favors and extra greasing but should be careful not to over kill since sometimes the wheel can get replaced as well.
For every action there is a reaction.
Our behaviour affects others behaviour. (Theory of reflexivity - George Soros)
Not always an advantage to seize the initiative and move first.
2. Anticipating your rivals response
"Just because an action lies in the future does not mean one (Charlie Brown) should regard it as uncertain."
Bargaining - An essential feature of negotiating is that time is money. When negotiations become protracted, the pie begins to shrink. Still parties may fail to agree, each hoping cost of negotiating will be outweighted by a more formidable settlement.
Look ahead and reason backward - reason may determine the outcome before it starts. The time for strategic maneuvering may be earlier, when the rules of negotiations are being decided.
War and peace - My enemy's enemy's enemy is not my friend
Games the British play: Marget Thacher & Kinnok Neil, second level thinking, third level thinking....
Rule 1 - Look ahead and reason backwards. Assumptions: The moves must be observable to others and strategies must not be reversible.
One of the lessons from the chapter, it is better to take risks earlier (assuming the same amount of risk) than later.
3. Seeing through your rivals strategy
Linear reasoning - I do he does i do he does
Simulataneous reasoning - Place himself simultaneously in both his own actions and the others guys' shoes then figure out the best move for both sides.
Dominant strategies : Time and newsweek , love is a dominant strategy. A strategy that dominates your other strategies not necessarily other players strategies.
Over- Cover warfare : If you have a dominant strategy then use it
Dominated strategies: Eliminate any dominated strategies from considerations and go on doing so successively.
Equilibrium strategies : Having exhausted the simple avenues of looking for dominated strategies or ruling out dominated strategies, the next things to do is to look for an equilibrium of the game.
Games with simultaneously moves : 1. Look and use dominant strategies
2. Next look for and avoid donimated strategies
3. Assuming similar behaviour on part of your rivals look for and use equilibrium
Part 2
4. Resolving the prisoners' dilemma
Opec, KGB, Reagan vs Waster Mandale, Yossarian's position in Joseph Heller's Catch -22
The underlying problem is the players' incentive to cheat in any agreement
1. How to achieve cooperation?
2. Detection of cheating? - Example "competition breaking out into new dimensions" - airline industry
The law increasing opaqueness - collusion focuses on the more transparent dimensions of choice and competition shift to the less observable ones.
Identifing cheating, identifing the cheater, even tougher is get the cheater to admits his faults
Cheating may consist of remaining passive and may be difficult to isolate: There are numerous excuse for inaction like greater urgency of other issues, time needed to consolidate forces and so on...
3. Punishment of cheaters: Within the structure of the game there should be punishment of the cheaters.
There is no solution that achieves reciprocal co-operation in a one-time game only in an ongoing relationship is there an ability to punish
Thus while each side is waiting to take advantage of the other both are benefiting from mutual deception. Argument of niceness - present vs future
4. Punishment is guaranteed- Crazy Eddie & Newmark & Lewis stero wars, beat the competition, most favoured customer
5. A choice of punishment - Most important are simplicity & clarity so that a player thinking of cheating can easily and accurately calculate its consequences.
Players should have confidence that defection will be punished and cooperation rewarded.
Since the punishment threats succeds in sustaining co-operation, it should not matter how dire it is. The fear keeps everyone from defecting, hence the breakdown never actually occur and its cost is irrelevant.
To reduce the cost of mistakes the punishment should be smallest size that suffices to deter cheating.
Few qualities of a good strategy: Clarity, niceness, provocability and forgiveness
Tit-for-tat Strategy: there is flaw in this strategy which is there is even with slight threat of misinterpretation can lead to break down in the strategy.
A better strategy would be to forgive to some extent and if you find your opponent taking advantage of you. Punish with tit-for-tat strategy.
5. Strategic moves
Scorched earth strategy - Joseph Stalin's scorch earth, Rupert Murdoch attack on a magazine. - For scorched earth strategy to be effective you must destroy what the invader wants, which may not coincide with what the present occupants want.
Leaving options open strategy: May not be always be a preferable strategy. Your lack of freedom has strategic value. It changes other players expectations about your future response and you can turn this to your advantage
"It takes a carpenter to turn a tree into a table, a clever strategist knows how to turn a table into a tree"
Preempt a move could have advantage (one may question the credibility of the preemptive move but once assured has strategic value) - example of the competition between USA and Japan on HDTV
One must establish a response rule to threat (2 types of threat - 1. Compellent threat - Plane hijacker kill passangers 2. Detterent threat- US will attack with nuclear missiles)
(2 types of promises - 1. Compellent promise - induce someone to take action 2. Detterent promise - induce someone from taking an unfavourable action)
Respond to a compellent threat with detterent promise
Respond to detterent threat with a compellent promise
Both differ only in status quo
Warning - when it is in your interest to carry out the threat
Assurance - when it is in your interest to carry out the promise
Warning and assurance do not change response but provide information
Unconditional move must be observable if it is to influence your rival, his actions must be observable if you are to influence them by threats and promise
- Democrats and Republicans - Regan Tax cuts - Use of combination of threats and promise to change ones strategies.
Sledge Hammer to crack a nut? - Promise the minimum amount necessary - A promise should not cost you a lot.
You shouldn't threaten somebody more than necessay. One should strive for the smallest and the most appropriate threat that will do the job. Make punishment that fit the crime.
6. Credible commitments
The eight fold part to credibility: to make a strategic move credible, you must take a supporting or collateral action. Such an action is called a commitment
1. Change the payoffs of the game: The idea is to make it in your interest to follow through on your commitment. Turn a threat into a warning, a promise into assurance
2. A second avenue is to change the game to limit your ability to backout of a commitment. Cut yourself from retreat. Remove yourself from decision making to leave to chance.
- Establish and use reputation
- Write contracts
- Out of communication
- Burn bridges behind you
- Leave the outcome to chance
- Move in small steps
- Develop credibility through teamwork
- Employ mandated negotiating agent
Reputations effect is two-edged sword for commitment. **Sometimes destroying your reputation can create the possibility for a commitment. Example a hijacker can break his promise of negotiations by attacking.
A weak bond of words can be strengthened in two ways, 1. Fear of consequences of breaking one's word or a glory/pride in not breaking one's word. 2. Contracts another way to make threats and promise credible.
3. But one life to lay down for your country - army, battle of agincourt
4. An offer you can't refuse: Interview and offer and ranking
A decision made by a committee is sufficiently dependent on chance that it cannot promise that given the same inputs it will reach the same verdict.
A committies inability to commititself to rational decision making makes the take it or leave it threat credible
7. Unpredictability
Don't have much - simple idea randomize
Part 3
8. Brikmanship
- Strategy of taking your opponent to the brink of disaster and compelling him to pull back
- Brink is not a sharp precipice, but a slippery slope, gently gradually getting steeper
- The credibility of brinkmanship still needs a device of commitment only that device contains within it a coin toss or a die that governs what happens
- Brinkmanship is not just the creation of risk, but careful control of the degree of that risk
- Successful brinkmanship remains something of an art and an adventure
9. Cooperation and coordination
- The invisible hand at best applies only to situations in which everything has a price
because many unpriced or non-marked activities matter, it is no wonder that individuals acting selfishly often do too much harm to others and too little good
- The feature common to grading system is that success is determined by relative rather than absolute performance
- Source of a problem: Why some students study too much, is that they do not have a price to pay, a price or compensation to the others. Each student studying is akin to factory polluting. Because there is no market in business and selling student study time, the result in a rat race.
- When other peoples' perception of your ability matter, it might be better for you to do things that increase your chance of failing in order to reduce its consequences.
Phsycologist see this behaviour in other contexts, some individuals are afraid to recognize the limits of their own ability. In these cases they take actions that increase the chance of failure in order to avoid facing their ability.
example: A marginal student may not study for a test so that if he fails. The failure can be blamed on his lack of studying rather than intrinsic ability.
10. The strategy of voting
- People have an incentive to tell the truth about direction but exaggerate when it comes to intensity.
- If splitting the differnce is the way then one has an incentive to begin with an extereme position in the first place
- Hence choosing the median position does not depend on the intensity of the voters preferencec only their preferred direction
11. Bargaining
-Don't have any notes
12. Incentives
-Don't have any notes
13. Case studies
- Don't have any notes
