Re: Beer Game Variations LO2302

Richard Karash (rkarash@world.std.com)
Tue, 1 Aug 1995 21:09:29 -0400 (EDT)

Replying to LO2300 --

On Tue, 1 Aug 1995, Steen Martiny wrote:

> Those of you who use the beer game regularly: What variations can be
> performed ?
>
> The standard way of gaming is to make one step move in retail demand (from
> 4 to 8 at the retailer) and nothing more. Peoples overreactions then
> set fluctuations ahead.
>
> But once participants have experienced this, the excitement is over. So,
> for those who want to play a second round, or for those who wants to go
> on for longer periods, I assume that the game can / should be performed
> in a somewhat changed fashion. With more variations in demand or what ?

Hello Steen --

I do two variations, but first let me suggest the mode of failure which
occurs with people who have played the game before (also sometimes with
people who have read closely the description of the game in The Fifth
Discipline).

The situation we find is that people *don't order*, they are so sensitive
to getting into an oversupply situation. They simply order eight cases
each week, ignore their backlog position, ignore the customer order
stream, and the whole team stays in backlog for the whole game. If the
whole team does this, orders just what is ordered of them, the score can
actually be moderately good, much better than the teams who attempt to
adjust and overshoot.

The problem is that the experience is boring, just what you're worried
about. The dynamics are actually pretty interesting to a systems
theoretician, and there is a valuable debrief, it's just that it's not a
compelling experience that stays with people.

So, here are two variations:

1) Make the customer demand 4 for the first four weeks, 8 for the next
four weeks, 12 in week 9, then 8 per week for the duration of the game.
When people see the 12, they worry about what's coming next, and the
interesting dynamics begin.

2) Create a penalty in the event that the retailer stays in backlog for a
long time. With technical groups, I make the penalty 10 x N squared,
where N is the number of weeks the retailer is in backlog. More often,
I'll make it zero up to ten weeks, 500 for 11 to 15 weeks, 1000 for 16 to
20 weeks, and 2000 for more than 20 weeks. Then I tell the whole team
that their mission is to get their retailer out of backlog!

I often use both devices and tell people that we have varied the game,
that it's not the same as described in the book (or that they played in
the past). And, I put the more "experienced" players nearer the factory.

Using these, I have no hesitation to do the beer game even with a
moderate number who have played the game before. Some of the wildest
scores have occured on teams including "experienced" players.

Richard Karash ("Rick") | <http://world.std.com/~rkarash>
Innovation Associates, Inc. | email: rkarash@world.std.com
3 Speen St, Framingham MA 01701 | Host for Learning-Org Mailing List
(508) 879-8301 and fax 626-2205 | <http://world.std.com/~lo>