At what age does your taste in gambling harden?

In 2024, data journalist Daniel Parris published an essay called “When Do We Stop Finding New Music?” that went all over the internet. The gist: Spotify data shows that at around age 33 musical taste “closes up” — playlists stop refreshing and spiral away from the mainstream. A Deezer study put the point even earlier, at 31. And taste is formed at 13–16: whatever you listened to as a teenager is what you'll be replaying at 60.

The professional question asks itself: how does this work in gambling? Do betting and casino have their own “age 33”? A complete answer requires operator data (that's a separate methodology piece), but there is more in the public domain than you'd think. We took three sources: the freshly released Gambling Survey for Great Britain 2025 (fieldwork January 2025 — January 2026), the GSGB trends report for 2023–2025, and Patterns of Play — a NatCen / University of Liverpool study built on the transactions of Britain's seven largest online operators.

The age map: every product has its own peak

The first thing the map shows: overall gambling does not fade with age. The share who gambled in the past 4 weeks rises from 33% at 18–24 to a peak of 54.8% at 55–64 — and only then declines. Britons don't quit gambling. They change what they gamble on.

The products line up into an age ladder. Online casino is a young person's game in its purest form: 8.6% at 18–24 versus 0.5% at 75+, a seventeen-fold gap. Slots — 7.8% versus 1.0%. Sports betting peaks at 25–34 (13.5%). Further along the axis come scratchcards, maxing out at 35–44, and finally lottery draws, which reach 48% among the 55–64s and single-handedly keep the older groups' overall numbers high. Every age has its product, and the peaks barely overlap.

The repertoire narrows for life

Parris's central metric is “openness of ear”: how much new stuff a person samples. The closest public analogue is the number of different game types played in a month. The gradient is mercilessly monotonic: 3.6 → 2.9 → 2.6 → 2.3 → 1.9 → 1.8 → 1.7. No plateau, no second wind — the repertoire narrows with every decade of life.

The tail is even more expressive: 13.8% of the young play 7 or more types at once — versus 0.4% of players aged 75+. “Omnivorousness” is a property of the age at which you enter the industry, exactly as musical omnivorousness is a property of adolescence. Same story with account counts from the Commission's 2019 survey: 25–44s averaged 3.3–3.4 accounts with different operators, the 65+ group — 1.5; one in five 18–24s held five or more accounts, against one in twenty at 65+.

Sports vs racing: this is where the local “age 33” hides

The most beautiful chart came out of the transaction data. The researchers split each age group's betting money between sports and racing. In every group under 35, sports (mostly football) takes 75–78% of losses. At 35–44 the share drops to 66%, at 45–54 to 53%, and between 45–54 and 55–64 the lines cross: at 55–64 racing already takes 61%, at 65–74 — 73%, at 75+ — 81%.

But this isn't “old people love horses” — it's plain imprinting. Today's 65+ entered betting in an era when a bet in Britain meant the shop on the corner and the racecourse, while the football coupon was a niche. Their taste locked in back then — and they carried it across half a century while the market turned around. Today, industry-wide, football brings online bookmakers £1.3bn in GGR versus £0.77bn for racing — but inside the wallet of a 75+ player the proportion is reversed, four to one in favour of the horses.

The crossing point of the lines — roughly age 50–55 — is the closest publicly measurable analogue of Parris's “33”: the age at which the taste recorded in youth overrides the current mainstream.

The young bring accounts, the old bring the money

Participation and revenue live at different ages. Active betting accounts per 1,000 population: 346 at 21–24, 329 at 25–34 — then a cliff down to 10 at 75+. Meanwhile annual GGR per customer rises almost monotonically: £64 for the youngest → £345 for the oldest. The under-25s are 13% of the population, 21% of active accounts and only 11% of betting GGR. Youth means many trials and little money; maturity means few trials and a lot of money. The industry quite literally lives off players whose taste has already hardened.

Explore → Exploit: fewer games, more gambling

Parris explained musical stagnation through the explore/exploit dilemma: with age it pays more to exploit the proven than to search for the new. The gambling data reproduces the mechanism with unsettling precision. While breadth halves, intensity within the chosen product multiplies: among casino players playing days per year climb from 7 to 31, sessions from 11 to 64, session length from 18 to 29 minutes, and game rounds per year from under a thousand to over ten thousand. With age the player doesn't leave gambling — they dig into a single product. For the product it's loyalty; for responsible gambling it's a risk profile, and that deserves to be said plainly.

The generational scissors

Three GSGB waves on a single methodology give us the first piece of dynamics. Participation of 18–24s in gambling beyond lotteries has fallen three years running: 33.8% → 32.9% → 29.5%. Among the 65–74s it is rising, statistically significantly: 16.4% → 17.2% → 19.2% (p = 0.033). Scissors.

Part of the explanation is surely a period effect: since April 2025 Britain has age-based online slot stake limits — £2 for 18–24s versus £5 for 25+; the regulator has literally institutionalised the age boundary. But the growth among the older group looks like that very cohort conveyor: people who took up online gambling at 45–55 have ridden into the retirement bracket with their habits intact. The older cohorts of the future will not play like today's — they'll bring their slots and their football with them.

So — at what age, then?

Here is the chronology the public data allows. Peak openness — 18–24: the maximum of game types, accounts and casino experiments. By 35–44 the portfolio is assembled: the betting mix is still mainstream, but breadth is down by a third. Between 45 and 55 the mainstream breaks: preferences recorded in youth outweigh the current market. Beyond that — pure exploitation: one or two products, high frequency, a growing average spend.

Meanwhile, a question for readers in the spirit of Deezer: how old were you when you placed your first bet or spun your first slot — and does your current profile match what you played back then? Judging by the data, for most of us it does.