Introduction to the Competitive Strategy Game Ian Larkin & Evan Rawley MBA 299: Strategy Section 1 (March 18-28, 2005)
Dec 20, 2015
Introduction to the Competitive Strategy Game
Ian Larkin & Evan Rawley
MBA 299: Strategy
Section 1 (March 18-28, 2005)
Why are we playing this game? Game is obviously highly stylized, but it does have quite
a few merits: Gets you thinking about some topics covered in class
(first-mover advantage, cost/differentiation strategies, effects of competition)
Simulates team-based approach used in strategy departments of most firms (and consulting)
Requires some degree of number crunching and analytical thinking
Demonstrates power of “psychology” Not meant to be taken TOO seriously, but definitely
should be looked at as an opportunity to learn and have fun!
This game consists of…..
8 Firms 4 Markets 3 Choices 9 Periods Each class plays two separate games
Firms
8 teams with 3-4 players Each firm has a set of distinct “capabilities”
Costs of entering a market, building capacity, and marginal costs of production
Each firm has a million dollars and access to external financing
Markets
4 separate markets with distinct supply and demand differences
Supply Different distributions of costs across firms
Demand Different market sizes Different types of “tastes” across markets
Choices & Periods
Each period a firm makes three choices in each of the four markets Do you enter, how much capacity do you build (or
sell), and what do you price at? Opportunities costs and trade-offs motivate
these decisions The game is played over 9 rounds
Today’s choices impact tomorrow’s outcomes (and vice versa)
What happens in each period…
Choices are privatelymade by firms
(due by NOON on due date)
The game aggregates the choices
Outcomes revealedand time moves forward(Usually late afternoon
of due date)
After the game compiles you see the following information for each market… How much you sold
Approximation of what others sold What prices everyone charged How much capacity everyone built/sold in
each market You get an update of your financial health
(but not anyone else’s!)
And time moves forward…
Capacity depreciates in value and is eventually scrapped (four periods after it is built)
Interest is paid to you if you still have money in the bank
Interest accumulates if you owe the bank money (interest rates escalate as your total borrowing goes up)
Important rules about choices You can only enter one market each period Capacity takes one period to build; therefore price entered
is valid ONLY for EXISTING capacity, not any additional capacity added in a period (What does this mean about pricing in the first period?)
Capacity has 4 periods of productive life; e.g. capacity built in period 1 is productive in periods 2,3,4 and 5
Capacity can be sold anytime after building it, and is automatically “scrapped” at the end of the 4th productive period; it depreciates according to the schedule in the market profile document At the very end of the game (period 9), all capacity is sold
automatically
Basic flow of game
1 2 3 4 5 6…Old capacity 0 100 100 200 200 100Price -- 20 30 20 15 15Capacity change 100 0 100 0 0 100-------------------------------------------------------------------------------------Units sold -- 100 50 120 150 100Revenue -- 2000 1500 2400 2250 1500
Questions: Assuming no more capacity is built, how much capacity will this
team have in period 7? Period 8? Period 9? Do you think this team is playing the game well?
Team choices in each round
What you are responsible for each period… Period 1: Enter into one market and choose
capacity You can add or subtract capacity later in the game
Periods 2-8: Manage entry, capacity, and pricing in all four markets
Period 9: Choose price only
Some notes on game play
You don’t ever formally enter or exit a market – just build capacity in the market you want to enter.
You should only enter capacity CHANGES in each round. If you’re happy with your current capacity, only enter a market price.
There are no inventories in this game. You build to order depending on how many units are demanded at your price. What does it likely mean if you sold all your capacity? If you sell less than capacity, you are not charged the MC of production for the unused capacity in your factory.
How to play the CSG
Go to the website: http://csg.haas.berkeley.edu
Scan through the list and choose the game corresponding to the game number I sent you earlier in the week
Clicking on that game will show you the following screen…
Here is where yourprivate information is.
Notice the use ofthe word change
Four Important Ideas
Understanding your cost structure and where you fall in each industry
Understanding demand Developing a coherent initial strategy based
on these two forces Using information gained in each round to
update your strategy (marginally or radically….)
Where should we get started? You have 3 pieces of information:
Individual firm cost information (private) Market profiles (public) Sample/simulated data on market demand with
monopoly seller (public) You don’t know others’ costs but get a good
idea of your cost position Markets are quite different in their economics
and probable competitive dynamics
Three types of costs Entry cost (EC): one-time sunk cost to enter an
industry Note, if you ever completely leave an industry, you are
charged the EC again if you re-enter! (May be worth leaving one unit of capacity as an option)
Capacity cost (K): cost paid per unit to build capacity, which has 4 units of productive life Again, K can be sold back according to depreciation
schedule, and is automatically sold back (e.g., scrapped) after 4 productive periods
Marginal cost (MC): unit cost to produce a good “Build to order” system so there are no inventories; you
only build what is demanded in a certain period, up to your capacity constraint
What does my cost data mean? The market profile document lists mean and SD for all
costs in a market. Costs are distributed normally, which means that, on average, 68% of teams SHOULD have costs within one SD of mean. (Of course, with only 8 teams in a game, weird things can happen…)
Example: Market B and fictional “Team 501” Mean EC=50,000, SD=3,000 Team 501 EC=52,541 Mean K=20, SD=4 Team 501 K=8 Mean MC=220, SD=40 Team 501 MC=212
Question: Is Team 501 well positioned in Market B?
Be careful! Costs are distributed normally, but with only 8 cost draws in
any game, anything is possible! Normality only holds with LOTS of draws
Example from last year’s game: capacity costs in market D (capital intensive market)
Mean K=$1,600; SD=400 Actual team draws:
1. 12462. 11093. 17644. 9575. 10026. 15327. 12618. 1054
You can and should use scenario generators in Excel or other packages to determine where you stand in expectation; for example, the chance that team 1 was the lowest-cost team in this market was only ~5%!
Thinking about economics Total costs, assuming entry in period 1 and constant capacity Y (and orders of
level N per period): EC + K*Y + MC*N*4(for periods 2-5) + K*Y + MC*N*4(for periods 6-9)
Market economics look VERY different depending on levels of EC, K, MC
Total revenue (assuming constant price): N * 8 * price (for all periods) + capacity scrap costs (in periods 5 and 9)
Note: When you spend money on EC, K, Y etc, you incur opportunity cost of 2% per period (assuming you have money in the bank!)
You really need to do your homework to understand whether EC, K or MC are most important in markets, and also how much capacity you should build (unused capacity has a cost, both real (wasted K) and opportunity (because you can always sell it back for greater than scrap value))
Understanding the Differences Between the Four Markets Market A: modest differentiation, moderate capital
requirements, moderate marginal costs Like refrigerators or other consumer durables (?)
Market B: low capital cost, high MC, differentiated Like apparel (?)
Market C: moderate capital cost, low MC, highly differentiated market Like business software (?)
Market D: high capital / sunk costs, low MC, undifferentiated Like memory chip production (?)
Analogies are not meant to be taken too seriously.
How can I start to think about demand and pricing?
Market demand and appropriate market prices are not as intuitive as the cost side, so we’ve given you additional information
For each market, we have put together 20 simulated “rounds” of demand, using one seller offering a single price (e.g., a monopolistic market – this is the “monopoly pricing exercise” that’s on the website)
You can use this data to calculate what your company would charge, and what capacity it would build, if it were a monopolist
What complicates the use of these results in the actual game? Would you ever charge more (and produce less) than you would under
a monopoly?
Sample price and quantity demanded
Market A Demand CurveLinear Fit
R-squared=92%
050
100150200250300350400450500
0 1000 2000 3000 4000 5000 6000 7000 8000
Quantity
Pric
e
Market A sample dataPrice Quantity
157 4624208 3187168 4306299 1430331 1043
89 6927441 311300 1414203 3361213 3108244 2356359 807399 485188 3778255 2212281 1730335 1096350 857175 4146111 6254
The P/Q relationship might not be linear
Market A Demand CurvePolynomial Fit
R2 = 0.9815
050
100150200250300350400450500
0 1000 2000 3000 4000 5000 6000 7000 8000
Quantity
Pric
e
Note: in Excel’s chart formatting toolbox, you can ask it to automatically fit the data to these and many other types of functions
Market A Demand CurveExponential Fit
R2 = 0.993
050
100150200250300350400450500
0 1000 2000 3000 4000 5000 6000 7000 8000
Quantity
Pric
e
Regression estimates of the P/Q relationshipRegressing P=a+b*ln(Q)
Notes: Regression command in Excel is under “Tools > Data Analysis > Regression”The P=a+b*ln(Q) form fits the data well and is easy to interpret, but feel free to use other functional forms if you deem fit. Doing so could make the analysis more complicated.
SUMMARY OUTPUT
Regression StatisticsMultiple R 0.9931415R Square 0.98633Adjusted R Square0.9855706Standard Error11.532812Observations20
ANOVAdf SS MS F Significance F
Regression 1 172742 172742 1298.8 3E-18Residual 18 2394.1 133.01Total 19 175136
Coefficients SE t Stat P-value Lower 95%Upper 95%Lower 95.0%Upper 95.0%Intercept 1104.1011 23.6935 46.599 3E-20 1054 1154 1054 1153.88X Variable 1 -111.72244 3.10011 -36.038 3E-18 -118 -105 -118 -105.209
X Variable 1 Residual Plot
-30
-20
-10
0
10
20
0 2 4 6 8 10
X Variable 1
Res
idu
als
Leading to easy calculation of monopoly price Our functional form assumption was P=a+b*ln(Q)
In our regression estimate, a=1104, b=-111.7
Because of our functional form assumption, the optimal monopoly price is MC-b. So if your MC in Market A is 50 (which is the market average), the optimal monopoly price would be 161.7
We can put this P into our function to get an optimal monopoly quantity, in this case (with a little math, which Excel can do for you) optimal quantity = 4605
Note: if you do not use the P=a+b*ln(Q) functional form assumption, the optimal price will not be MC-b. You have to set up a profit function, take FOCs, etc. It’s actually not too hard, if you want to work with other functional forms.
Usefulness and problems with this approach Why is this result useful? Would you ever price above your optimal monopoly price? Would you expect capacity in a competitive market to be higher or lower than in a monopoly?
What are the problems with this? Not everyone has the same costs…… so your monopoly P and Q are
not the same as everyone else’s If there are multiple firms in the market, the monopoly P and Q does
maximize joint profits. However, it doesn’t maximize profits for any individual firm, so everyone has incentive to defect (cartel problem)
In our example regression, assuming 3 firms split the market evenly, they can each make a profit above MC of $170,000 per period if they collude on monopoly P and Q. If one firm drops its P by 5%, it will steal demand from the other firms and increase profits by over $20,000, while the other two firms both lose over $15,000 in profit. Note that total profits went down, but defector’s profits went up. Hallelujah competition!
Other notes on demand and pricing Consumers have some degree of “brand loyalty,” and are willing to pay a price premium for their preferred brand. The degree to which they are willing to do so differs by market and its exact nature is unknown You can get an idea of the “brand loyalty” premium in each market
by looking at the “brand substitution” matrix in the “Market Profiles” document. Market D has almost no brand loyalty!
If a customer’s preferred brand is not offered, she only looks at price
Demand “expands” somewhat with additional firms E.g. total market demand for two firms both charging X will be
greater than total market demand for only one firm charging X You can get an idea of how much demand expands by looking at
the “market demand & effect of brand proliferation” matrix
Information is critical – good teams adapt flexibly to ongoing information given In each round, you’ll get important pieces of information
How many units you sold (exact figure, only known to you) Your financial position (can use it to update your model, if you have
one) Capacity changes by each competitor Prices charged by each competitor Total approximate market demand
Since you have the same information as they do, except their cost profiles, you can try to “guestimate” competitors’ cost profiles
There’s also an opportunity to make a “public statement,” which can be very strategic Using public statements to collude will result in SEC fines Mentions of suggested pricing or capacities strictly not allowed
How you are evaluated You are required to write a 1-page “journal” of your team’s actions and
strategies at 3 specific times: Directly after periods 1, 5 and 9 (due next section/class) First journal entry due Friday, April 1 in section
“Journal” explains rationale for your strategic moves (especially recent ones) and what you expect going forward Last journal entry will be a bit different, but we’ll explain it later Should be done as a professional-type “memo” to public, board of
directors, etc. (making format professional not important but making content professional is)
Journals only returned at the end of the game; no comments will be given before that; also, journals will remain completely private (so you can talk about your costs, etc)
Your CSG grade also depends on the end financial state of your firm, but to a lesser extent than the journal Initial cost position will be taken into account
Money Previous students have suggested adding a real financial element
to the game to help incentives and make the game even more fun
Note: participation is voluntary – if your team doesn’t want to play for money, let your professor know; it won’t affect your grade
Idea: you pay $20 to play the game. You will get $1 for every $100,00 in the bank at the end of the game, but will pay $1 for every $100,000 in losses $2 million = breakeven Losses capped at $40 ($4 million); no upside cap Last year this would result in the class “breaking even” on
average; in other years teams would average as much as $10-15 in winnings per team
Expectations and Odds and Ends Take the game seriously, and try to use
concepts you learned in micro and the current strategy course (because many of them are actually helpful)
You do not need the perfect model Psychology is very important
Good strategies don’t always map to good outcomes, and vice versa
Cheating’s no fun (Bernie Ebbers is going to jail, and so will you)