Game Theory

1
Game Theory
2
Strategy vs. Game Theory

• Strategy is about getting more than your fair
  share.
• Game theory is about maximizing your return and
  minimizing the return of your competitors.
• Therefore, your strategic moves do not occur in a
  vacuum, but are based on what moves your
  competitors make or might make
• and you want to get your competitors to make
  the wrong moves.

3
Poker

• Who plays?
• Do you play the probabilities?

4
John von Neumann

• The best mind of the 20th Century
• 1928 - Mathematical genius von Neumann, 25, plays
  poker and invents game theory.
• After someone bluffs.
• 1944 - Von Neumann is a major force in inventing
  the atomic bomb and the modern computer.
• 1950 - Two Rand Corporation scientists invent the
  Prisoners Dilemma game.

5
The Prisoners Dilemma

• In 1950 a conductor on a train to Kiev rehearses
  for a Tchaikovsky concert.
• KGB arrests him for subversive activity.
• KGB arrests Boris Tchaikovsky, a worker, on the
  streets of Kiev.
• KGB puts them in separate cells so they cant
  communicate.
• KGB offers them both a deal.

6
The Prisoners Dilemma

• If the conductor rats and Boris doesnt, he gets
  one year in a gulag and Boris gets 25 years.
• If the conductor doesnt rat and Boris does, he
  gets 25 years in a gulag and Boris gets one year.

7
The Prisoners Dilemma

• If both rat, each gets 10 years.
• If neither rats, each get three years.
• The silent auction begins.

8
The Prisoners Dilemma

• Each serving 10 years, they meet in the gulag,
  begin talking and discover they ratted on each
  other.
• While talking they realize that if each had said
  nothing, they would only have been in for three
  years.

9
Payoff Matrix
Boris
Rat
Not Rat
Rat
10, 10
1, 25
Conductor
3, 3
25, 1
Not Rat
Conductor, Boris
10
Game Theory

• See Game Theory - Programming case on
  http//www.charleswarner.us/cseindex.html
• See Game Theory Sales case in the above link.

11
Scenario

• KAAA-TV, on the West Coast, is considering
  switching from its current prime time (8-11 p.m.)
  to early prime time (7-10 p.m.). KAAA is 2 in
  prime time, and because of KBBBs very strong
  10-11 p.m. lead-in to its late news, KAAA is 2
  in late news even though its news product is
  competitive. KBBB is 1 in late fringe also.

12
Scenario

• KBBB-TV is 1 in prime time and has excellent
  10-11 p.m. network lead-ins to its 11 oclock
  news, which puts it 1 in the late news race.
  KBBB is also 1 in late fringe.
• KCCC-TV is a weak 3 in prime time and late news.
  It is a network-owned station and will not switch
  to early prime.

13
KAAA Decision Tree
KBBB Go
KAAA Go
KBBB No Go
KAAA
KAAA No Go
KBBB Go
KBBB No Go
14
Payoff Matrix
KBBB
Go
No Go
Go
4, 2
3, 4
KAAA, KBBB
KAAA
2, 1
1, 3
No go
Assigning weights is the most difficult decision.
15
KAAAs Payoff Weights

• 4, 2 If KAAA switches (go) to early prime and
  KBBB also switches (go), both gain more revenue
  from higher ratings for 10-1030 p.m. late news.
  KBBB doesnt gain as much as it would if KAAA
  switches and KBBB doesnt (3,4).

16
KAAAs Payoff Weights

• 3,4 If KAAA switches (go) and KBBB doesnt
  switch (no go) , KAAA gains revenue with its
  10-1030 p.m. news, but the news is up against
  KBBBs strong prime and KBBBs late news gets
  higher ratings than before because KAAA has
  dropped news from the time period.

17
KAAA Strategies

• 1,3 If KAAA doesnt switch (no go) and KBBB
  switches (go), KAAA loses big because its weaker
  10-11 p.m. prime is up against strong KBBB news
  which has strong lead-ins and strong late fringe.
  
• 2,1 If KAAA doesnt switch and KBBB doesnt
  switch, nothing happens, but the outcome isnt as
  bad as if KAAA doesnt switch and KBBB switches
  (1,3)

18
KAAA Strategies

• Adding KAAA go weights (43 7) shows switching
  is the best strategy, because its no-go weights
  (12 3) are much worse.
• KBBBs judged weights are the same with either
  decision (41 and 32 5).

19
KAAA Strategies

• KAAAs best strategy is to announce its staying
  with its current schedule, hoping KBBB will
  switch to gain an advantage and hurt KAAA (1,3).
• Then, at the last moment, KAAA switches to early
  prime to gain its maximum outcome (4,2), assuming
  KBBB stays with its decision to switch.
• Secrecy is critical.

20
The Prisoners Dilemma

• If the prisoners had been able to communicate,
  what would have happened?
• If they had been given a chance to play the game
  again and again, what would have happened?

21
The Prisoners Dilemma

• The rules for the game changes when you play
  repeatedly, as the Rand Corporation scientists
  discovered.
• And if the other side gets greedy (which is
  inevitable), you must use tit-for-tat.
• You must teach the other side cooperation (to
  accept three years in the gulag) and to do whats
  best for both.

22
Strategies

• Tit-for-tit, while effective, leads to escalation
  in many situations.
• The dollar auction
• A dollar bill is put up for sale minimum bid one
  cent.
• Proceeds as a regular auction.
• One exception Auctioneer must be paid by highest
  bidder, but also by the second highest bidder.
• Winning bidders pay what they bid and receive
  dollar.
• Second-highest bidders pay what they bid and
  receive nothing.
• Escalation
• Game introduced in 1971. Thousands of games
  played, average paid was .3.41

23
Strategies

• Bluffing
• Bluffers make statements, show behaviors, and
  perform activities that would be perfectly all
  right if they were not completely unfounded.
• Bluffing is like vitamins. It is essential in
  small amounts, but harmful if used excessively.
  
• In the game Chicken, the proper bluff is to act
  drunk, crazy irrational.

Laszlo Mero, Moral Calculations, Copernicus, 1998.
24
Mixed Strategy

• Cant bluff all the time, no one will believe
  you.
• I you never bluff, youll never win big because
  competitors know you tell the truth.
• Have to bluff occasionally.
• A mixed strategy.
• On a random basis no identifiable pattern.
• Chicken?

25
Strategies

• Trial balloon
• Prior announcement
• False announcement
• Secrecy
• Preemptive strike
• Threat, warning, or promise
• Fighting brand
• Guerrilla marketing (tactic)

26
Games

• Zero-sum games
• Assume a winner (1), and a loser (-1)
• Multi-player games
• Marathon - 1000 racers, different niches
  (men-women), many prizes.
• Business a marathon.
• Strategies are about achieving objectives (not
  necessarily winning in a zero-sum game).
• Profit
• Market share
• Dominate a niche

27
Co-opetition
Customers
Complementors
Competitors
Suppliers
Business Ecosystem
28
Co-opetition

• Who are Google Newss complementors?
• Competitors?
• Who are The Huffington Posts suppliers?
• Complementors?
• Remember, ecosystems co-evolve.
• Wolves and caribou

29
Co-opetition

• It might be a smart strategic move to change the
  game and the players.
• In fact, it might be a smart strategic move to
  pay someone to compete (co-opetition).
• Coke, Pepsi and NutraSweet
• Lin, McCaw and Bell South
• Google, Microsoft and Twitter

30
Auction Theory

• Auction theory evolved from game theory.
• Types of auctions
• English auction
• Dutch auction
• http//www.charleswarner.us/artindex.html
• Vickrey auction
• http//www.charleswarner.us/artindex.html

31
Business Insanity

• Doing the same thing over and over and expecting
  different results.
• Think strategically and come up with new and
  different solutions and use game theory to weigh
  your moves to maximize your advantage and
  minimize your competitors advantages.