The short version

In 2025-26, Victorians lost $3.32 billion on poker machines in pubs and clubs (not counting the casino). That’s a record. But it isn’t spread evenly. I wanted to see where that money actually comes out of, and what happened when COVID shut every venue down and then they all reopened.

What I found:

  1. The most disadvantaged fifth of council areas lose about $838 per adult a year. The least disadvantaged fifth lose $429.
  2. When the venues reopened after lockdowns, losses came back much higher in disadvantaged areas than in well-off ones, especially in Melbourne.
  3. A small group of venues, mostly big suburban pubs, take a big share of the money.

A note before start: losses are recorded where the venue is, not where the player lives. So this shows which areas the money is lost in, not that people in poorer areas gamble more as individuals.

1. The map: where the money is lost

Source: VGCCC (2025); boundaries from ABS (2022).

Try this: switch between the two layers in the top right corner. The dark areas on the “money lost” map (Melbourne’s west and north, and regional cities like Mildura and Latrobe) mostly line up with the dark areas on the disadvantage map. Hover over your own council to see its numbers.

The highest loss per adult is Brimbank at $1,091 per adult in 2024-25. In Boroondara, the lowest in Melbourne, it was $135. Grey areas have no pokies venues at all.

2. The pattern: poorer areas lose more per person

3. The pause: what happened when the venues shut

Source: VGCCC (2026a, 2026b).

Each line is one fifth of Victoria’s council areas (about 16 councils each), grouped by disadvantage. I’ve shown each group compared with its own 2018-19 level, the last full year before COVID, because the groups have very different populations so raw dollar totals would be misleading.

For a decade the lines stayed fairly close together, although if you look carefully the gap was already opening: the two least disadvantaged groups were slowly dropping while the others crept up. Then in 2019-20 and 2020-21 lockdowns closed every venue for months, and statewide losses fell to $1.57 billion in 2020-21. That’s the dip in the middle.

When venues reopened, the lines split. The three most disadvantaged groups ended up 37%, 31% and 29% above their 2018-19 level (Q2, Q1 and Q3). The two least disadvantaged groups only went up 14% and 11%. These are dollar figures without inflation adjustment, but every group went through the same inflation, so comparing the groups against each other is still fair.

Try this: double-click a group in the legend to show only that line, then click another to compare the two.

4. The comeback wasn’t equal

Source: VGCCC (2026a, 2026b).

This breaks chart 3 down to each council area: the change in losses from 2018-19 to 2025-26. The dashed line is the state average, 23%.

In Melbourne the pattern is really clear. The dark dots (more disadvantaged) sit on the right and the yellow-green dots (least disadvantaged) sit on the left. Across Melbourne, losses in the most disadvantaged group rose 31%, while the least disadvantaged group rose 10%. Boroondara (0%), Stonnington (-4%) and Bayside (-5%) basically didn’t grow at all.

Some of the biggest jumps are in fast-growing outer suburbs like Melton (56%) and Hume (41%), so part of this will just be more people moving there. I couldn’t fully separate that out with this data. But established suburbs show it too: Brimbank, an older western suburb rather than a new growth area, still went up 30%.

In the country it’s messier. Small shires with only one or two venues can swing a lot from one venue changing, so I wouldn’t read much into single dots there.

Try this: hover over a dot to see the before and after dollars for that council.

5. Your local: the venues behind the numbers

Source: VGCCC (2026a). 6 venues with no machines recorded at the end of the year are left out.

Every dot is one of the 478 pubs and clubs with pokies in 2025-26. Higher up means each machine took more money, and bigger dots are venues with more machines. The grey boxes show the middle half of venues in each group. Three things stood out to me:

  • Machines in poorer parts of Melbourne work harder. The typical (median) machine in Melbourne’s most disadvantaged group lost players $174,228 in the year, compared with $108,844 in the least disadvantaged group.
  • Hotels take far more than clubs. On average a pub machine took $156,756, compared with $93,397 for a club machine.
  • The money is concentrated. The top 10% of venues (48 venues) took 28% of all losses in the state.

Try this: click “Club” in the legend to hide clubs, then hover over the highest dots in the Metro panel. Most are big pubs in Melbourne’s north, west and south-east.

What I take from this

Pokies losses in Victoria aren’t random. They are highest where people have the least, they came back hardest in those same places after the lockdown pause, and a fairly small number of big pubs collect a large share. The lockdowns were basically an accidental experiment: for a few months the machines were off everywhere, and when they came back on, the gap between Melbourne’s richer and poorer areas grew.

Limitations. Losses are counted by venue location, so people who travel to gamble get counted in someone else’s council. Dollar figures are not adjusted for inflation, which is why I only compare areas and groups against each other over time. Part of the growth in outer suburbs is population growth, which I couldn’t fully separate out. SEIFA scores are from the 2021 Census, and the adult population figures in the VGCCC file are projections. All of this shows patterns between areas, not what any individual does.

If gambling is affecting you or someone you know, free and confidential support is available from Gambler’s Help on 1800 858 858 (24/7).

References

Australian Bureau of Statistics. (2022). Local Government Areas - 2022, GDA2020 [Data set]. Australian Statistical Geography Standard (ASGS) Edition 3. https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs/edition-3-july-2021-june-2026/access-and-downloads/digital-boundary-files

Cheng, J., Schloerke, B., Karambelkar, B., & Xie, Y. (2024). leaflet: Create interactive web maps with the JavaScript ‘Leaflet’ library [R package]. https://CRAN.R-project.org/package=leaflet

Pebesma, E. (2018). Simple features for R: Standardized support for spatial vector data. The R Journal, 10(1), 439-446. https://doi.org/10.32614/RJ-2018-009

R Core Team. (2025). R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org

Sievert, C. (2020). Interactive web-based data visualization with R, plotly, and shiny. Chapman and Hall/CRC. https://plotly-r.com

Victorian Gambling and Casino Control Commission. (2025). Annual density update: Population density and gaming expenditure statistics, 2022-23 to 2024-25 [Data set]. https://www.vgccc.vic.gov.au/for-community/gambling-victoria/gambling-data

Victorian Gambling and Casino Control Commission. (2026a). EGM venue level expenditure (current), 2023-24 to 2025-26 [Data set]. https://www.vgccc.vic.gov.au/for-community/gambling-victoria/gambling-data

Victorian Gambling and Casino Control Commission. (2026b). EGM venue level expenditure (historical), 2008-09 to 2022-23 [Data set]. https://www.vgccc.vic.gov.au/for-community/gambling-victoria/gambling-data

Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer-Verlag. https://ggplot2.tidyverse.org

Acknowledgements

Gambling data © State of Victoria through the Victorian Gambling and Casino Control Commission, used under a Creative Commons Attribution licence. SEIFA scores in the VGCCC file come from the ABS (2021). Boundaries © Australian Bureau of Statistics.

I used Claude (Anthropic, 2026) to improve the grammar, clarity, and style of my writing.