NSW poker machine data - Alliance for Gambling Reform

NSW poker machine data

How the Alliance pulls together data around poker machines in NSW

It’s only possible to address a problem when you know what it is. So getting the best quality data about poker machine operations helps us analyze the situation in suburbs and towns, and to provide that information to organisations we work with on gambling reform. 


We don’t have good quality data in NSW. We have to do a lot of estimating, which is frustrating, because the data exists, but is currently shielded behind some industry commercial in confidence veil. The Alliance thinks that data on gambling losses should be publicly available, to inform good policy decisions.Each state and territory publishes data differently - the ACT doesn’t publish anything, Victoria has venue by venue data, Queensland publishes by council area every month, NSW publishes by council area half-yearly. If you don’t mind being about 18 months out of date, you can get data on all gambling via the annual national statistics.


How do we calculate the losses in a venue in NSW? 

Victoria publishes venue by venue data. NSW does not. At the Alliance we have to estimate the venue losses - first we turn to the data about losses per council area, then we have to separately track down information about the venue to find out the number of machines it has.

In NSW the data for clubs and hotels is published in separate spreadsheets, every 6 months. Within each category, the venues are grouped by the council area (or LGA). Most of the time. If there are less than 5 venues in an LGA, Liquor and Gaming bundles the data with a neighbouring LGA. [They’ve told us this is to stop armed robbers identifying venues with lots of cash (at AGR, we’re guessing they think that Victorian robbers can’t read spreadsheets)]. 

Unfortunately, that bundling isn’t consistent particularly in the sparsely inhabited areas of the state west of the dividing range or in very small council areas like Mosman in Sydney. As an example, when we were trying to establishes losses in the electorate of Orange we found that Liquor and Gaming had 

  • combined data for Blayney and Cabonne for hotels in 1st half 2020; 
  • Hotels in Cabonne were previously bundled with hotels in Weddin until June 2018, then for FY2019 only, hotels in Cabonne reported just for that LGA; 
  • Clubs, but not hotels, in Forbes are combined with clubs in Weddin every half year since 2017

So, unless it’s bundled with a neighbouring LGA,  what we do is calculate the average amount of losses per poker machine (EGM) in a given period for each local government area for the 6 month period. We usually also work out the daily and weekly losses for each LGA, as organisations we work with often need this information. 

And we do that once for clubs and once for hotels, as the difference in NSW is stark - the average pub pokie takes more than double the amount of an average club poker machine.

It does mean we cannot attempt to estimate the losses in venues in LGAs that are combined with neighbouring LGAs. It also means it’s not possible to calculate the losses in venues within state or federal electorates, because these boundaries don’t match council/LGA boundaries.

{Don’t tell anyone, but the Alliance and a bunch of researchers, ClubsNSW employees, some MP staffers, and a few club and hotel managers have found ourselves on a list which receives this data quarterly for clubs and hotels. It’s not clear to us why this data isn’t published on the main Liquor and Gaming webpage for everyone to see, so we’ll be taking that up with the Minister shortly}

How do we find out the number of machines at a venue? That can be a challenge. Liquor and Gaming have an app called “Find my LIA Banding” which they created to help venue operators apply for more machines - but we can use it to look up data for each venue, including the number of machines they have. Liquor and Gaming also publish most months a list of venues with liquor licences, and those spreadsheets also show how many machines are in a venue. They can be found on the L&G website, or the NSW Customer Service data portal.

So, we have the average losses per EGM in an LGA, both for club and pub pokies, per day. We then multiply that by the number of machines at a venue, multiplied by the number of days we’re looking at, which is usually 365 for the year.

As an example, we were recently analysing pubs owned by the Endeavour Group, formally the ALH group, partly owned, still, by Woolies. 

The Greenhouse Tavern in Coffs Harbour has 25 machines. The average losses taken per machine in Coffs Harbour LGA pubs for the first three months of 2021 was $23,851. We estimate that taking that quarterly amount x 25 machines x 4 = the annual losses taken at the Greenhouse = $2,385,100.

part of spreadsheet showing venues and losses


What do the Bands mean?

For gambling purposes, NSW has a very weird system of categorizing the state by ranking areas according to an arbitrary formula that factors in socio-economic disadvantage, poker machine density and losses. Until 2018, the formula was applied to council (LGA) areas which were then ranked from 1 to 127. The review of the Local Impact Assessment system in 2017 received numerous submissions pointing out that using entire council areas obscured pockets of significant disadvantage. The legislative changes of 2018 replaced councils with statistical units called SA2s, which the Australian Bureau of Statistics uses, and which cover a few suburbs in metro areas, or towns in regional/rural NSW. These SA2s are then ranked based on the formula.

Then, arbitrarily, the state is divided into 3 bands, which are used to determine the level of community consultation, the size of the cheque a venue needs to write to get more machines and whether any more poker machines can be allowed. The 20% of SA2s most severely impacted by pokies are called Band 3 (new machines are banned), the next 30% are called Band 2 and fully 50% of the state’s SA2s are then called Band 1 - where no community consultation is needed.

There is no known rationale for choosing this 50/30/20 division of the state. Before the 2018 reforms it had been a 50/25/25 split, so the effect of the changes is that even less of the state than before is actually characterized as severely impacted (Band 3) by poker machines.

This map shows part of Western Sydney with pokie clubs and hotels marked. Green areas are Band 1, pale orange are Band 2, pink are Band 3

Map of part of Western Sydney showing bands and gambling venues