Control Data Australia Memories compiled by Brian Membrey

RIMFIRE, RIOTs and Restrictions

A statement issued by VicTAB General Manager Jack Rutter late in 1971 suggested that five new types of betting were being considered under the roll-out of RIMFIRE – he suggested a Treble (about to be introduced in Queensland), a Tierce, now today's Trifecta where all three place-getters must be selected in order; a bracketed Quinella where the field was divided into eight selections (said to be popular in Japan, but which would never have worked in Australia), a four-place card similar to a Quadrella, but placed horses only (very common in the 1960s with illegal bookies in pubs around Melbourne); and a “concession daily double” as used in South Australia where a consolation dividend was paid to punters selecting first and second.  It wasn't clear from the statement whether the Quadrella had already been decided on, but it was implemented just six or seven months later which suggests it was already “on the cards”.

This may have been Rutter’s wish-list or perhaps he was only nominating what was available on other systems.  In fact, the RIOT design would not allow for the Trifecta and would have messy for the Treble.

The success of the Quadrella introduced in 1972 had “pros and cons” – the lure of a dividend that could run into the thousands was great and attracted high levels of revenue, but given that four separate races were involved, it was logically restricted to one per meeting and the pool money was effectively tied up for the day.   

The early days of promoting a “non-betting shop” atmosphere had long since passed, and the concept of a Trifecta where Joe Punter had to select the first three place-getters in the correct sequence offered similar "wind-fall" dividends, but on a race-to-race basis and thus generating continuous revenue saw the introduction of the new bet type on-course as early as the Spring of 1974, but RIMFIRE (or perhaps more correctly the RIOT terminal) was unable to handle three double-digit selections and it wasn't until 1981 when the heavily modified CRISP system finally came on-line that off course punters could invest in the Trifecta.

RIOT ; The keyboard

Although six keys were available on RIOTs and allowed the Quadrella via brackets, the layout of two banks of three keys each meant that the second selection in a Trifecta would have had to have been split; the two numbers keyed as the lowest digit of the Win bank and the highest digit of the Place section, which was found impossible with the software available and also extremely confusing for the operator.

The fall-back of bracketing was not an option in this case as it was possible that two bracketed horses could have ran in the placings with rank confusion as a result - e.g. if the placings were 8, 3 and 11, then under the Quadrella bracketing system, numbers  8 and 11 would be combined, and the Trifecta would have been 8, 3 and 8!

(I'm told this still a problem for TabCorp on the few occasions on which they operate on races from North or South America when stable-mates are bracketed as a matter of routine as nobody in Oz has any idea of the “form” and are only betting by “lucky numbers”).

In 1979, the TAB began offering the Trifecta on the major race of the day, but only through telephone betting. The long queues at the Trifecta windows at the racecourse showed that there was a huge demand for a bet type that could pay in the hundreds, even thousands, on a single race.

A Treble/Tierce would have been possible, but again would have required brackets to prevent splitting of a selection - it was probably outmoded by the Quadrella anyway.

The other major shortcoming of RIMFIRE/RIOTs was the technology didn’t allow for multiple selections to be made on a single ticket.


Whether this was identified as a potentially major problem when the system was conceived is unknown, but it should have been!

The TAB had no other options for single race betting - I think race-by-race Quinellas were already in place - and it should had been apparent that any new bet type would involve multiple races and hence the likelihood of multiple selections.

(I’m a bit hazy on this, but I think Double and later Quadrella “multiples” during RIMFIRE days were still handled manually with a handwritten ticket, collated at the agency and totals transmitted to a Control Centre).

One of the major advantages of GWS (and thus CRISP) was that multiple selections could be made on a single ticket, although there was a slight restriction.

All had to be for the same amount - some punters like to back more than one horse in a race for a win - one as their first selection, another or others for varying lesser amounts as “savers” to cover the potential loss on the primary selection - thus for different amounts and requiring different tickets.  This restriction remains even today in agencies, not sure about on-line.

The other major advantage of CRISP was the introduction of mark-sensing technology whereby Joe Punter could mark down his selections on a pre-printed slip, feed this via a reader and collect a second ticket with the details of the wager printed.  This had the great advantage of reducing the need for verbal interaction between the punter and the selling clerk, thus considerably speeding up the process (estimated by 50%) and eliminating the potential for misunderstandings - did you ask for race 5, number 2 for 3 units, race 2 number 3 for 5 units OR race 3, number 5 for 2 units?  PLEASE CHECK YOUR TICKETS.

Punters were always thus warned “Please check your tickets”, but the mark-sense concept immediately placed the onus on the punter to ensure that had he or she had marked their selections correctly.

The proposed CRISP technology was not entirely new – a similar system called GART (General Amusement & Recreational Terminal) had been developed over the previous 12 months by CDC in the U.S. with a market target of American states looking at lotteries, lotto-type systems and off-course betting as methods of raising revenue.

A GART configuration included both a mark-sense reader and a ticket printer as well as other peripherals, but VicTAB had struck upon an alternative concept where the wager details would be printed on the same slip as the punter used to record his bets. I seem to remember technicians at the Moorabbin plant rigging up a reader and printer which could emulate this - for the purposes of the experiment, I think that after passing through the reader, the bet was held in memory and the marked slip had to be hand-fed a second time into  the printer.   

Other than providing the dummy terminal as “proof of concept”, CDA hadn’t  been  involved in the experiment and the idea was ultimately abandoned as legal opinion was that the TAB would be liable to payout on the original mark-up if it and the printed bet were somehow different and there was no evidence that the ticket had been tampered with.


(The obvious connotation is someone attempting to erase the original mark-up and replace it with a winning combination - the TAB were convinced that this wasn’t possible without detection through the use of a special light meter that would highlight the inevitable difference in the thickness of the ticket where an erasure had been made and the surface altered.  I remember inviting a visiting CDC engineer who had worked on the GART terminal to a meeting at VicTAB to discuss the idea and he destroyed the “one-slip/ticket” concept in about two minutes by marking a betting slip with a Letraset transfer, registering the bet with the printed details, then peeling off the Letraset and re-creating a totally different wager with a ballpoint - the Letraset system was designed for invisible removal simply by placing the adhesive side over an existing transfer and gently rubbing until it was removed without trace).

The problem was one of legislation only - the "one-card" system was adopted by some other states including NSW, and later used in Hong Kong, France, England and the Netherlands where machines developed by International Totalisator Systems (ITS) were in place in the 1980's

These installations would probably have needed betting rules to be drafted to say, in essence, that the bet as recorded in the totalizator computer system was the one that was accepted as being valid where there was a disparity - little consolation to the punter, as tickets that misprint with missed or added selections due to some sort of read/write error are always a (albeit hopefully remote) possibility.

And, of course, the reverse chances of Joe Punter actually checking a ticket after a losing bet which may have in fact reflected a winning ticket were in common vernacular … “none, nil. Zilch, zap-all” and probably another dozen other synonyms which all could be summarised into the acronym SFA, “Sweet” the opening word and “Fresh Air” in the politer and lesser used variation.  

My Free-of-Charge Other Than Free Bottle of Red at Lunch RIOT Re-design And Patented Solution For Multiple Selections

I’m sure there are a few survivors that could shoot down the possible solution in flames given what was achievable some 45 years ago compared to today. All-in-all, possibly no cigar, but I’m willing to accept that bottle of red at a Third-Friday lunch as a well-earned consolation prize for at least making you think about it!

BIG update : April, 2017

The only image I could find of a RIOT keyboard is poor quality - hence the facsimile above. Some of the Meeting (Melbourne, Sydney, etc) and Bet Type (Win-Place, Double) legends are indistinguishable and not in use, but would have changed over time anyway as additions were made.  There was also a Sell/Pay and other administrative functions not shown.

Although not marked, the left column of Win and Place was the Race Number (0 (non-functioning), I doubt 11 or 12 were ever used - it was noted that CARBINE could handle 10 races per meeting with no plans for expansion.

 The Win and Place banks 0-9 were simply the horse/greyhound number, the Race and high-order 0 Selection irrelevant to SELL but potentially part of the Security Code on a PAY transaction.

The Unit values covered probably 90% of requirements; two tickets were issued if someone wanted, say, 15.  Two were also issued if Joe Punter wanted an “each-way” bet, but with varying amounts for the win and place.

The three keys in the First and Second Leg allowed for early Feature Doubles (Type F) such as the Caulfield-Melbourne Cup where there could have been up to 300 entries. RACE was irrelevant in this case and used as the high-order digit in selections over 99.

BAL produced a summary ticket of the punter’s Sell and Pay tickets to alert the operator of what was due in either direction; the balance was maintained by the CPU and downloaded when the key was pressed.

The ticket was printed with a security code which comprised the meeting, bet type and a scrambled number across the Win, Place and Units which very, very few knew the entire algorithm for.  I seem to remember from later CRISP days it was based on a random number entered by TAB Security staff entered for each meeting, so knowing either individually or by collusion the algorithm didn’t help!

Check out SPECIAL RIMFIRE Report for Staff and Agents - ya know you always wanted to run a TAB!