OR Insight Vol. 6 Issue 1 January - March 1993

Spreadsheet gaming and management skills development Converting a simulation game to a spreadsheet style of operation

Jim Freeman

The use of computers in management training is

being developed all the time.

very common nowadays, and still evolving. Computer simulation games are an important aspect of this development but, as with most applications, need to

A challenge for the new generation of simulation game designers is how to adapt this powerful software technology effectively to their own field of operation. This article focuses on a project which

be updated from time to time as a result of major improvements in background technology. This article describes how a stand-alone business game developed in 1987 was converted to run on the LOTUS 1-2-3 spreadsheet system. The resultant

attempts to do just this.

By way of a backdrop to the project, a brief outline on spreadsheets, their evolution and use ¡s given next. Following on, the said business game is introduced and the experience of converting this to a LOTUS 1-2-3 format is described. In the last sec-

package has been found to have a number of significant advantages over the original game - but there are costs also.

tion, the pros and cons of the kind of translation Management gaming has grown remarkably over the last few years, with applications not only more

exercise are summarised.

numerous (reflecting the continuing popularity of the approach) but also diverse (as packages have been developed to deal with an increasingly wide range of business situations).

Spreadsheet modelling The first electronic spreadsheet package, commercially available was VisiCalc, written in 1979 for use

For computer-based games in particular, significant advances in systems and software capability have led to demands by users for much greater sophistication. On-screen editing, flexible graphics, help keys etc. are features, now largely taken 'as read' by the computer community (where, of course, only a comparatively short time ago, most had hardly been

on the Apple Il microcomputer. This enormously

heard of).

more advanced spreadsheet systems, tailored to meet the new hardware standard e.g. Microsoft's

successful software product is credited with having made the greatest contribution to Apple's (spectacular) sales growth in the early eighties.

The emergence of the IBM Personal Computer (PC) in 1981 was quickly followed by the development of

Modern spreadsheet packages provide many of the facilities often sought of the latest simulation game packages (indeed it can be argued they are the main reason why user expectations are now so high (Li,

EXCEL, Borland's QUATTRO PRO, Lotus' l-2-3. Of

these integrated software packages (so-called because their spreadsheet, data management and graphical analysis facilities operate in unison with each other), 1-2-3 - the market leader - has been predominant. The CTN-X package described in the next section, was developed to run with 1 -2-3 re-

1991).

More importantly, spreadsheets, because of their widespread availability and popularity, have fundamentally altered the nature of computer modelling. What was once an exclusive concern of mathematical specialists is now routinely conducted by busi-

leases 2.01 and later. (At the time of writing, the most advanced version of 1-2-3 is version 4.01).

ness staff of all types. As a result, spreadsheet

An example of 1-2-3 worksheet from CTN-X is given in Fig.1 below. lt is a typical spreadsheet screen,

models have become pervasive in many areas of

consisting of a matrix of cells, each addressed

management (Jackson 1989) - with new applications

according to the letter and number of the appropriate

Copyright é 1993 Operational Research Society.

9

OR Insight Vol. 6 Issue 1 January - March 1993

column and row. For example, the spreadsheet

The game application

'pointer' (represented by the shaded area) is located in cell G6 here. With Version 2 of 1 -2-3, up to 256 columns and 8192 rows of information can be held

The CTNTRAIN simulation game was written at the

Manchester School of Management in 1987, in conjunction with the Retail Confectioners and Tobacconists Association (RCTA), to support the

in a single 'worksheet'. Each cell can hold a label (e.g., a piece of text), a value (i.e., a number) or a

training of Confectionery, Tobacco and Newsagency (CTN) shop managers and related staff (Yiu,1987),

formula.

Figure 1: Introductory Screen

The package (released as a compiled TURBO BASIC code) runs on IBM PC and compatible microcomputers. CTNTRAIN is based on actual

in CTN-X game

operating data supplied by RCTA (Key Note Publica-

tions, 1986): a high degree of realism was thought

vital if the game was to be perceived by RCTA G6:

A

B CD E

members as credible and relevant.

REA F

G

H

With CTNTRAIN, as many as five CTN shops, each

1

represented (ideally) by a team of between three and eight members, compete for a given week's trade - assumed to take place in a large town or conurbation. All teams start off on an equal basis and teams have to make a variety of decisions reflecting their shops' retailing policies, in respect of gross margins, stock re-ordering, special offers, staffing etc. An example of printed output from

2

3 4 5 6 7 8 9 10

**INTRODUCTION**

1. Previous game to load?

NO

CTNTRAIN is given in Fig. 2.

2. How many retailers? (1-5)

11

12 13

3. Enter your team name

14 15

16 17 18 19

20 21

(1) (2) (3) (4) (5)

Figure 2: Example of CTNTRAIN printed output CTN MANAGEMENT GAME

WEEK 1

NB. If a previous game is loaded, no further questions on this screen need be answered. **Re..start entering information by pressing

RETAILER 1

YOUR DECISIONS FOR THIS RUN WERE AS FOLLOWS:

AltR**

27-Jul-92 03:O7PM

NUM PRODUCT CATEGORY GROSS MARGIN

STOCK

SPECIAL

ORDRES

OFFER CODES

Values, unlike labels, can be manipulated arithmeti-

cally. Formulae are used for this purpose and

TOBACCO

9.0

1250.00

3

commonly act on ranges of values held in different areas of the worksheet.

NEWS & MAG

28.0

650.00

0

CONFECTIONERY

22.0

400.00

1

SAVOURY SNACKS

24.0

50.00

1

Another type of cell entry, possible with 1-2-3 (but

ICE CREAM & SOFT

not every spreadsheet system) is the so-called 'macro'. A macro is a series of instructions which

DRINKS

24.0

100.00

2

STATION ERY,CA ROS

35.0

50.00

2

can be used to automate a spreadsheet task in the same way that a function key on the computer is programmed to invoke a standard procedure (see "Automated spreadsheets" by Colin Lewis in this issue and O'Leary & O'Leary, 1990). By using macros, simulations can be made to run dynamically in 1-2-3. This facility has been exploited very exten-

OTHERS

20.0

100.00

3

ADVERTISING () 6.00 RECRUITMENT-FULL TIME

O

PART TIME

3

LAYOFFS FULL TIME PART TIME

sively (and in fact, proven indispensable) in the development of CTN-X from the original source

YOUR FIXTURES ARE 6 YEARS

program, CTNTRAIN.

REFIXTURING?

1

O

25 WEEKS OLD

o

N

OR insight Vol. 6 issue 1 January - March 1993 SALES/PROFIT ANALYSIS

STOCK ANALYSIS

1 TURNOVER BY PRODUCT CATEGORY

5 INVENTORY BY PRODUCT CATEGORY

1482.67

45.6

TOBACCO

NEWS AND MAG.

768.15

23.6

NEWS & MAG

CONFECTIONERY

591.22

18.2

CONFECTIONERY

533.85

TOBACCO

SAVOURY SNACKS ICE CREAM & SOFT DRINKS STATIONERY, CARDS

OTHERS

0.00

78.52

2.4

SAVOURY SNACKS

338.32

121.22

3.7

ICE CREAM & SOFT DRINKS

305.87

94.49

2.9

STATIONERY, CARDS

438.08

118.42

3.6

OTHERS

404.77

3254.69

TOTAL

1148.27

3169.17

TOTAL

100

2 COSTS OF GOODS SOLD BY PRODUCT CATEGORY

6 STOCK/SALES RATIOS BY PRODUCT CATEGORY

WEEKS

TOBACCO

TOBACCO

0.77

1351.73

NEWS & MAG

553.92

NEWS & MAG

0.00

CONFECTIONERY

466.15

CONFECTIONERY

0.90

SAVOURY SNACKS

62.16

SAVOURY SNACKS

4.31

ICE CREAM & SOFT DRINKS

93.83

ICE CREAM & SOFT DRINKS

2.52

STATIONERY, CARDS

63.20

STATIONERY, CARDS

4.64

OTHERS

95.23

OTHERS

3.42

TOTAL

0.97

TOTAL

2686.23

3 GROSS MARGINS BY PRODUCT CATEGORY 7 ADDITIONAL INFORMATION TOBACCO

130.94

8.8

NEWS & MAG

214.23

27.9

CUSTOMERS

CONFECTIONERY

125.07

21.2

STAFF-FULL TIME

SAVOURY SNACKS

16.36

20.8

- PART TIME

ICE CREAM & SOFT DRINKS

27.38

22.6

HOURS WORKED

58

STATIONERY,CARDS

31.29

33.1

IMAGEFACTOR(%)

96.00

OTHERS

23.18

19.6

STAFF COSTS RATIO (%)

3.90

SALESISQ.FT.()

4.65

TOTAL

568.46

5166 O

3

17.5

MARKET SHARE SUMMARY 4 OPERATING EXPENSES

RETAILER

%'S

PRODUCT CATEGORY

1

TOTAL

FIXED COSTS

215.00

6.6

LABOUR COSTS - NORMAL

128.36

3.9

TOBACCO

100.0

100

0.00

0.0

NEWS&MAG

100.0

100

CONFECTIONERY

100.0

100

SAVOURY SNACKS

100.0

100

ICE CREAM & SOFT DRINKS

100.0

100

STATIONERY, CARDS

100.0

100

OTHERS

100.0

100

-OVERTIME ADVERTISING

6.00

0.2

TOTAL COST

349.36

10.7

PROFIT

219.10

6.7

11

/-l n r ifl.Lf&L. vot. u issue I (11%

January-March 1993 Superficially however the games look anything but alike. In part this can be explained by CTN-X offerIng a number of features - particularly graphing (see Figs. 3 and 4 below) - not available with CTNTRAIN. More generally, the apparent incompatibility arises from the dominant nature of 1 -2-3 which seems to imbue the new application with a power which, in

The game, despite its innovativeness at the time, demonstrates typical operating limitations of many of the one-off computer systems written in the early to middle eighties. In particular, the package has no graphical output facilities. Given the large amount of

quantitative information generated by CTNTRAIN this is now recognised to be a serious shortcoming, especially when used with less numerate trainees. On-screen editing is also lacking and although the

fact remains mostly untapped.

routine provided in CTNTRAIN for changing or

Figure 4: Example of graphical output from CTN-X

correcting inputs ¡s reasonably straightforward, the feel ¡s somewhat dated and unfamiliar.

WEEK ONE Marke r Shares

To put right these and other omissions, it was decided to re-orient the package along spreadsheet lines. The enhanced version of CTNTRAIN - code-named CTN-X - was deliberately programmed to mimic the original software package as closely as possible in order that existing CTNTRAIN users could switch to

the revised game system with a minimum of fuss and discomfort. Thus, for example, the sequence of prompts and inputs in the new package is exactly the same as in the old, the logic is completely consistent between the two programmes, the games serve broadly identical training aims etc. A team strategy in one game would, all other things being equal, produce equivalent results if tried in the alter-

Product Category a

bi

g

native game.

A summary of the main differences between the two packages - according to whether they are judged to be advantages for CTN-X or otherwise - is given in

Figure 3: Graph options with CTN-X

Figure 5.

Al: A

B CD E

REA F

G

Some of the differences here admittedly derive from the way the project was managed. For instance, the total space occupied by the many (sixty plus) CTN-X

H

1

2 3 4 5 6 7 8 9 10 11

12 13 14 15 16 17

files created, might have turned out to be a good deal less (though still greater than CTNTRAIN) if more experience in using 1 -2-3 macros had been gained at the start of the work. On the other hand, CTN-X's graphics might have made the situation even worse if they had been more ambitious. At present, only 1-2-3's bar charts are used so that four

of the spreadsheet's graph options are effectively ignored. (Of course, the potential exists for these and other 1-2-3 options to be used by CTN-X at a later date but only at the cost of a considerable

This spreadsheet will plot the results for **THREE RETAILERS**

amount of extra programming.)

Conclusions

18 19 20

27-Jul-92 03:O7PM CMD

To sum up, the project has demonstrated the feasibility of converting a computer simulation game over to a spreadsheet style of operation. There is nothing

NUM

12

OR Insight Vol. 6 Issue ¡ January - March 1993 special about the game chosen here so the process

more sense to update CTNTRAIN, along stand-

is probably one that can be applied generally.

alone lines. The additional facilities could than have

However the reprogramming work required is potentially so onerous - given the present state of the art and demands such a degree of technical skill, that it

been achieved with considerably less effort and without some of the more glaring disadvantages experienced.

really cannot be judged to be cost-effective. With Spreadsheet-based packages of course, continue to serve a very important role in the world of simula-

hindsight, it would have made much

tion-gaming. However, in the author's opinion, at

Figure 5

least, certain applications are predisposed for Advantages

Disadvantages

CTN-X is broken down into a

CTN-X occupies just short of 1Mb of disk space making it some 13 times larger than

number of smaller files, making it easier to understand the structure of the package and hence to debug and update it.

spreadsheet treatment and others are not. Bridging this gap is still a problem.

Figure 6: Example of Input Decision Screen with CTN-X

than CTNTRAIN. Inevitably

this brought about some degradation in run-time performance but on 386-based machines and upwards, the problem is not a noticeable

REA

12:

J

K

M

L

RETAILER 1

**

O

N

P

Q

ENTER GROSS MARGIN (0-99%)

one.

CTN-X exploits many of

CTH-X depends on users having

# TOBACCO

10

1-2-3's standard options, particularly, graphs, menus

a basic awareness of spreadsheet processing and an

# NEWS & MAG

25

and random number generators

initial familiarity with l-2# CONFECTIONERY

29

#SAVOURYSNACKS

26

#ICECREAM&SOFTDRINKS

19

#STA11ONERY, CARDS

18

#OTHERS

22

etc which it uses to good effect,

CTN-X - with its on-screen

3; CTNTRAIM, on the other

hand, can be run by virtually anyone, requiring no

o 1

specialist computer knowledge

2

whatsoever.

3

CTN-X is considerably less

4

editing facility (see Figure secure than CTNTRAIN - which since its completion has 61 - allows for more flexible data correction during decision input,

5

6

been obtainable only in

7

executable form.

8 9

CTN-X offers much greater

o

guidance on data input e.g. confirming the ranges of values acceptable for a particular decision.

Press Alt R after correcting input

7 - Jul-92 03:58 PM

NUM

For the interested reader

CTN-X can be viewed as a

Jackson, M (1989) Creative Modelling with LOTUS 1-2-3, 2nd Edition.

friendly induction aid to the

John Wiley and Sons

l-2-3 system. Keynote Publications Ltd (1986) Confectioners, Tobacoenists, Newsagents.

Additional features with

Keynote Publications Limited.

CTN-X e.g. enabling teams to be identified by names

Li, J KT (1991) Management Simulation using LOTUS 1-2-3, Unpublished

instead of just numbers help make the related exercise more personal and friendly.

M.Sc. Dissertation.University of Manchester Institute of Science and Technology.

O'Leary, T J & O'Leary, L 1(1990) The Student Edition of Lotus 1-2-3 Release 2.2 User's Manual. Addison Wesley Publishing Co.

Yiu, P T P (1987) Computer-based Training for CTN Retailing Management, Unpublished M.Sc. Dissertation. University of Manchester Institute of Sdence and Technology.

13

OR Insight Vol. 6 Issue 1 January - March 1993

JIM FREEMAN specialises in computer simulation. Before taking up his lectureship at UMIST in 1981, he was statistician at the Distributive Industry Train¡ng Board (DITB) where he held responsibility for computer based training. He has published widely in his field and been involved in developing a large number of management simulation packages - particularly business games - for organisations such as Cyanamid, Makro Self-service Wholesalers and the Greater Manchester Police.

L4