7: Designing the ticket system
<< 6: Inventing Willowbrook || 8: Locating the stations >>
Designing the ticket system
The understanding with the Government of Ontario was that this was to be a test installation,
to demonstrate, in the age of the private automobile, whether the public would be prepared to
park the car at a station, and use a good, modern commuter train to make the journey into town.
The plan was to make on-going changes to the nature of the service offered, change the train
schedules, vary the fare structure, try different types of feeder service, and generally measure the
response of the public to these various changes.
Clearly, the need was for some means of measuring the influence of these changes day by
day, as they affected the level of carryings on the trains. So the commuter group needed to
develop a system of ticketing that would yield all of the statistics to enable this kind of analysis
to be made.
The primary purpose of any ticket system is to ensure that passengers pay for their journey, in
effect, control of revenue. The design of commuter ticket being used on the existing commuter
service was intended for multi-trip use, and was sold at ticket counters, on a basis of being valid
for one month at a time. So there were records of the quantities and types of tickets sold, and of
the cash flow received in payment, but not of what train services the tickets were being used on.
It was a system of on-train ticket collection, where the conductor punched or clipped off small
squares of the ticket, one for each trip made. If the passenger happened to have some of these
squares left unused at the end of the month, the ticket expired, and the trips that these squares
might have paid for were lost. A few passengers used single trip tickets that the conductor
punched, but very few passengers did this. The only way there could be of recording carryings
on a train would be if the conductor counted the passengers as he punched each ticket, and
turned in a report at the end of the trip. Mostly, the conductors became experienced at estimating
the load on the train, and turned in an estimated figure. It was doubly difficult to have statistics,
not only recording the total carryings on a train, but also to know the stations of origin and destination of each passenger, so that the
traffic generated at the various stations might be known.
The system was open to fraudulent travel, where a passenger could alight from the train before the conductor reached his car, or a
new passenger might board the train after the conductor had passed, and so complete his journey without losing the validity of the ticket.
It was saved by the automatic expiry of the ticket at the end of the month of issue, but it was quite doubtful if this control could be
The commuter group considered the practice on the subway
systems, where payment for each trip was made upon entry to the
platform. Revenue control was on the ground at the stations, and
train crews were not involved. One token or counterfoil was
deposited each time a passenger entered a station, so there could easily be a count of passengers
entering. There was no special control at exit. The passengers exited through turnstiles, that
could count the number of exits made, but the primary purpose was to ensure that passengers
could re-enter only by depositing tokens at the entry turnstiles. So the total usage of each station
could be known, but not the relationship of origins and destinations.
This was essentially a flat-fare system that was best applicable to transit that extended only
over short distances. Riders over short distances paid the same rate as long distance travelers, so
in effect, the short-haul passengers were subsidising the long-haul passengers. The TTC had
already extended the network to longer journeys that ought to cost more than the flat fare, and
some forms of zoning had been tried, but always it was difficult in practice, and politically
The intent of the Government was to start with a service over a distance of 21 miles on each
side of Toronto, with the possibility for passengers to travel over the whole length of line, 42
miles, and later extensions had to be accommodated. Taking into account that there also would
be two trains in and out again serving Hamilton at a further 20 miles, clearly a flat-fare system
would not be a good idea. The need was for a system that would accommodate different fares for
different combinations of origins and destinations.
In the United States and in Europe, some experimental systems were under test, using tickets
carrying a magnetic strip. The strips could be coded, to be valid for entry and exit at specific
stations, and the turnstiles canceled each entry and exit until there was nothing left, when the
turnstile would retain the ticket. This should give an electronic count of entries and exits at each
station, and could have been elaborated to report the station of origin for every exiting passenger.
Another concept of magnetic ticket was to have a stored value on the ticket. Passing it through a
turnstile at point of entry recorded on it the identification of that station, then at exit, the turnstile
would compute the fare due, and deduct it from the stored value. So the exit turnstiles would be
computers, totaling trips completed, revenues earned, and reporting origins as well.
The problem was, that these experiments were still in their early stages, and considerable
problems were being reported to us, concerning the unreliability of the machines under test, and
the difficulty for ticket holders to know how many trips, or how much stored value, remained on
their multiple trip tickets. The new commuter service was already breaking new ground for train
equipment and territory served, so we were not willing to inaugurate it with risky problems with
electronic machines for revenue control.
Developing the two-part ticket
This did persuade us to consider a system of on-ground ticketing that should yield the
statistics we were asked to produce. So we abandoned the current practice of having open
platforms at the stations, and accepted that the commuter stations would be fenced in, with
admission only through controlled entry points.
In the absence of such an electronic system, we examined alternatives for a manual ticketing
system, with on-ground sales and collection. There was a long-established practice for using
small tickets printed on heavy card, that would be perforated by a ticket taker using a punch at
entry, and then collected at the destination station. The process of punching was to cancel the
validity of the ticket for multiple entries. This is a slow process, not attractive for an intensive
commuter service, and requiring a lot of workers and passengers lining up at the gates of the
Once the service would be in operation, it would be the responsibility of Jack George to
oversee the whole accounting process, so he was deeply involved in the process of creating a
ticket system that would meet our needs. He had been in a clothing store somewhere, and had
seen a special form of price ticket, that had colours on it, and was perforated across in two lines,
so that it could be torn into three parts. When an item was sold, one part of the ticket stayed on
the garment, one part went to the cash wicket, and the other part went to stock control, for them
to record the turnover on the item sold. The colours served as a code, so that the proper
destination of each part could be clearly recognised. Jack brought one of the tickets into the
office, and we considered how we might use this kind of system. We saw that we had no need
for a three-part ticket, but maybe we could make good use of a two-part ticket.
The thought was to have a first part to be dropped at point of entry, that would have the effect of
canceling the ticket, just as the punched ticket had done. Then the other half dropped at point of
exit would terminate the journey, and the ticket takers there could count the tickets taken, and
sort them by point of origin. We would know how many passengers
arrived off each train, and where they came from.
Testing the concept
Somehow we had to have a demonstration for our own benefit, to
convince us that the two-part ticket could be applied to a commuter
service. Jack knew of a supplier of rubber stamps, and suggested
that we get a few stamps made up, then we could make a few two-
part tickets to play with. So we designed the words for some
samples, and took them to the rubber stamp maker. When we
explained our purpose, we had an immediate proposal that they
would print 100 samples of each, for us to practice with. That is
where our first two-part tickets came from.
We designed the words for only two stations, Port Credit and Scarborough, so that the names
would be printed in different colours of ink. We had Jack sit behind a desk, and the rest of us
circled passed him, dropping half-tickets, and expecting him to sort them as they fell. A single
file of pedestrians can walk at the normal rate of 60 per minute, and he collected tickets at that
rate, with no delays as we walked passed his desk. George had a problem recognising the
differences in the colours of the inks. There was not enough colour to stand out.
Once we had found that the colours of ink alone were not sufficiently prominent for quick
recognition, so we coloured them in by hand using coloured crayons. The colours of the crayons
were the same as we had chosen for the inks, but now each half ticket was of a dominant colour.
We mixed up the colours of the halves as we dropped them into his sight. Jack immediately
recognised which stations the half-tickets applied to, and separated them easily as they came in,
so we knew we could use this concept for a high volume operation.
So Jack asked in the store where the sales tickets came from, and
we contacted the supplier directly. A representative came to visit us
from Kimball Systems, a division of Litton Industries, bringing
samples of their products. We could have the colours imprinted on a
long strip of ticket paper, then have the text printed on it in black
ink. Each half of the ticket could be colour coded for one station,
and the other half for another station. At the station of origin, the
ticket taker would collect ticket halves, but only those colours good
for his station. The passenger would carry the other half for exit at
the destination station, where only the proper colour would be
To make it quicker for the ticket takers at the exiting stations to
sort and count the exiting passengers by station of origin, we added
a small strip of the opposite colour on each end of the ticket, and serial numbers for revenue
control. We assigned different colours to each station, but we would have to ensure that ticket
takers, when we got to them, would not be colour blind! The proposed system was ready in
October 1966, to be explained to the passenger department of CN and to Bill Howard's group in
the Government, who accepted that it would meet the needs of statistics collection. So quantities
were estimated and orders placed, to come into our hands in time for the day of inauguration.
Our Office Manager, Jessie Smith, selected the different colours, and worked with the printing house on the choice of inks. Basically she chose
pale colours for the east end, and darker colours for the west end. The Government decided to have a flat fare of $0.25 for children, so we had a
simple ticket printed in black ink for children, with no colour coding added. For some tickets we had the words printed in green ink to
The system offered the potential, at some time in the future, to be mechanised. During the
printing of the tickets, we could add a magnetic strip to match the colour of the station at each
end. This would identify the station where the passengers would be dropping their ticket halves
into a turnstile for entry or exit. The turnstiles would accept only tickets coded for that station,
and would recognise whether it was being dropped for entry or exit. Computers attached to the
turnstiles could compute numbers of passenger exiting from each train, their stations of origin,
and the revenue earned by each train on each trip. So for the future we would be able to automate
ticket control, even one station at a time.
As it turned out in the long term, GO-Transit did not need to mechanize this ticket control
system, because an alternative was found in the self-validated time-lapse system now used. In
this system, there is no ticket check at entry or exit. Passengers are required to carry a valid
ticket that may be inspected at any time by inspectors traveling on the train. When the ticket is
purchased it is not yet valid. The passenger validates it by passing it into a stamping machine
that imprints the date and time of day. After a specific time-lapse, the ticket ceases to be valid.
Normally the ticket is never collected. It remains for the passenger to dispose of it, since it has
This system loses the facility of counting the exiting passengers by station of origin. With the
long-term experience of the service in full operation, there is no further need for this detail of
<< 6: Inventing Willowbrook || 8: Locating the stations >>