54 / 76
54
// PUBLIC GAMING INTERNATIONAL // July/August 2016
SHORT SUMMARY
Electronic draw systems based on RNG
have become an integral part of the lottery
industry and require utmost security to re-
duce exposure to potential fraud. Insider
fraud, where an employee of the lottery, an
operator, or vendor, takes advantage of their
system knowledge and access to the system
to fix a draw outcome, is the most danger-
ous and difficult to detect. There are existing security methods that
can protect lotteries against such fraud.
INTRODUCTION
Lottery drawings using electronic random number generators are
integral to the future of modern lotteries. Lottery industry leaders
have recognized that they must embrace electronic drawings: while
offering potential for growth, electronic drawings also offer major
cost savings to lotteries.
However, lotteries worry about potential fraud with the shift
from mechanical to electronic drawings. In a recent lottery fraud in
the US, the perpetrator was seized by coincidence, which raises the
question of how many other fraud cases were not detected.
In this article we discuss the costs of fraud and the probability
of RNG fraud in the US, and we provide some guidelines for
RNG security.
THE COSTS OF FRAUD
At the recent Public Gaming conference, we approached top-
level lottery management to obtain their perceptions of fraud in the
lottery industry
(see Table 1)
. The survey results suggest that fraud
would be damaging to lotteries by reducing their customers, restrict-
ing lottery’s freedoms to introduce new products, and through im-
posing large monetary costs on lotteries in the short and medium-
run. The average estimates of fraud cost to their lottery provided
by lottery leaders was $283 Million, for the first year, and $1.31
Billion, for years 2 to 5 after the fraud occurred. The survey did not
take into account other potential impact such as increased financial
risk of liability through class action, so the numbers may be under-
stated. These numbers are also based on a survey rather than actual
costs that were measured, so they are only suggestive. However, they
highlight that lottery managements believe fraud could severely af-
fect their institutions in many ways. To avoid this impact, the best
safety and security procedures should be introduced to protect lot-
tery systems and particularly electronic draw systems.
THE PROBABILITY OF RNG FRAUD
With large potential winnings, insider fraud is a true threat to
the lottery industry. In the US, many lotteries have been victims of
fraud in the electronic drawings for games played within or across
states.
Currently we can estimate that over the last 5 to 10 years elec-
tronic draw fraud has occurred in games played across 13 states. If all
50 states used electronic draw systems, this would put the probability
of electronic draw fraud at 26%.
This underestimates the amount of
fraud, because some states do not take part in electronic draws, and
we can only report the amount of fraud that has been discovered
and reported.
SECURITY MISCONCEPTIONS
Misconception 1:
Mechanical drawing machines are more secure.
Explanation:
With high potential costs and prevalence of fraud,
one approach would be to return to mechanical machines. However,
when the security measures are correctly implemented, an electronic
draw system allows the lottery to control and protect the gaming
environment better than with a mechanical system, making the elec-
tronic draw system also more secure than a mechanical system. Ad-
The Importance of Security and Fraud Detection
for Electronic Drawings
By Helena Pereira, PhD, Marketing Director, Szrek2Solutions
Table 1: LOTTERY FRAUD SURVEY RESULTS
In a survey of 9 lottery directors, 2 former directors, 1 CFO,
1 COO, and 6 other top level management (n=19):
• On average, respondents thought that over half of their custom-
ers may stop purchasing lottery tickets or buy fewer tickets if they
discovered that lottery fraud had been committed in that state.
• Lottery’s legislative and regulatory powers, including the
lotteries ability to develop new games, would be impacted by
the discovery of lottery fraud—95% agreed.
• The total short-run costs estimated were, on average, $283
Million for the first year in which lottery fraud was detected.
Total costs include reputational costs, legal fees, political costs,
losses in ticket sales, etc.
• The total medium-run costs estimated were, on average, 1.31
Billion for the first 5 years in which lottery fraud was detected.
Total costs include reputational costs, legal fees, political costs,
losses in ticket sales, etc.