The WAGER, Vol. 16(7) – Guesstimates: Can venue staff accurately identify patrons with gambling-related problems?


Although some researchers suggest that there are visible indicators (e.g., gambling for long periods of time, using ATMs, placing high risk bets, avoiding social contact) associated with problem gambling, they also recognize that, in real-time, it might be difficult for gambling venue staff to correctly identify these indicators (Delfabbro, Osborn, Nevile, Skelt, & McMillen, 2007; Schellinck & Schrans, 2004). This week, The WAGER reviews a study that compares patrons’ self-reported gambling status with gambling venue staffs’ estimation of their gambling status (Delfabbro, Borgas, & King, 2011). 


  • The researchers interviewed patrons and staff from seven gambling venues in South Australia during different times of the day and week.
  • Patron survey: A small number of all patrons approached by the researchers agreed to participate in the survey and (n=303) completed the past year Problem Gambling Severity Index (PGSI).[I]
  • Staff survey: The staff rated patrons’ gambling status on a 4-point scale (1 = No problems, 2 = Might have some problems, 3 = Suspect patron is a problem gambler, 4 = Confident patron has severe gambling-related problems) for patrons they recognized (i.e., patrons they had seen at least once or twice at the venue). All staff in South Australia receives responsible gambling training, which purportedly includes visual indicators for problem gambling. 


  • Self-report (n = 303) PGSI indicated that 40% were not at risk for problem gambling, 29% at low risk, 22% at moderate risk, and 9% were classified as problem gamblers.
  • The staff recognized 76% (n = 230) of the patrons. Table 1 compares the patrons’ PGSI scores with staff’s rating of patron gambling among these 230 patrons.
    • Staff members made many false negative ratings for individuals who identified themselves as having problems (i.e., 64%) and were even more inaccurate for those who identified as having low or moderate problems (i.e., 83%).
    • Staff made a number of false positive ratings for individuals who identified themselves as being problem-free (i.e., 6%).

Table 1: Patron PGSI Classifications Compared to Staff Ratings (adapted from Delfabbro, et al., 2011)



  • Using the PGSI, the patrons self-reported their gambling status. The researchers could use more than one screen to identify problem gambling.
  • The researchers approached only patrons using machines in small or medium-sized venues. Therefore, patrons did not come from larger venues and/or venues that offer other forms of gambling.
  • Only a small percentage of the people approached agreed to participate. Therefore, the results might not be representative of the Southern Australian gambling population.


The results indicate that staff ratings were not a reliable indicator of problem gambling among patrons. It is unclear whether these low ratings are the result of the training, or the difficulty of completing this task in real-time. Perhaps future research will lead to the ability to better train staff to identify patrons’ problematic gambling. Similarly, bartenders also have difficulty correctly identifying which customers are intoxicated (Brick & Erickson, 2009), and this phenomenon is likely even more difficult for identifying problem gambling. These results suggest that staff might be unable to identify customers with a problem using visible indicators.


Brick, J., & Erickson, C. K. (2009). Intoxication is not always visible: An unrecognized prevention challenge. Alcoholism: Clinical and Experimental Research, 33(9), 1489–1507. doi: 10.1111/j.1530-0277.2009.00979.x

Delfabbro, P., Borgas, M., & King, D. (2011). Venue staff knowledge of their patrons' gambling and problem gambling. Journal of Gambling Studies, Online First. doi: 10.1007/s10899-011-9252-2

Delfabbro, P., Osborn, A., Nevile, M., Skelt, & McMillen, J. (2007). Identifying Problem Gamblers in Gaming Venues: Final Report. Melbourne, AU: Gambling Research Australia.

Schellinck, T., & Schrans, T. (2004). Identifying problem gamblers at the gambling venue: Finding combnations of high confident. Gambling Research, 16(1), 8-24.

[I]The researchers did not obtain an exact record of the number of patrons who refused to participate because the environment and patrons made it difficult to capture this information. The researchers aimed to report about a working sample/convenient sample of PGSI scores rather than a fully representative sample of South Australian gamblers.

Leave a Reply

Your email address will not be published. Required fields are marked *