International Conference on Gambling and Risk-Taking
Price sensitivity and jackpot seeking in blockchain lotteries.
$2.6B
single-operator revenue in 2022, higher than many regulated alternatives.
15.1%
share of all Litecoin transactions in Dec 2022 attributable to one crypto-gambling site.
≈ 0
peer-reviewed studies of consumer behavior inside a blockchain-native game of chance.
"Aside from decentralized finance, gambling may be the industry most benefiting from onchain technology" — yet behavior inside these markets has been treated as a black box.
Do the established behavioral drivers of lottery demand persist when the institutional wrapper is stripped away?
Wallets — not names. Social observability and stigma effects effectively removed.
No licensing, no spending limits, no advertising rules, no jurisdiction.
Ticket price plus a volatile blockchain gas fee, both denominated in cryptocurrency.
| Platform | PancakeSwap — largest DEX on Binance Smart Chain |
|---|---|
| Game | 6-digit number match, top prize for full match-of-6 |
| Ticket | $5 USD-equivalent, denominated in the CAKE token |
| Cadence | Two draws per day, every 12 hours |
| Operator | Pseudonymous developer collective; smart-contract enforced |
| Period | 18 months · July 2021 → December 2022 |
Two volatile costs in two different assets — a pricing structure that simply does not exist in a regulated fiat lottery.
Tickets sold ↑ with jackpot size.
Skewness preference and the convex region of Friedman–Savage utility — the most consistent finding in lottery demand research.Forrest et al. 2002 · Kearney 2005
Tickets sold ↓ with the price of a ticket.
Classical price sensitivity — but the onchain "price" is now ticket cost plus a variable gas fee, and we can identify both.Baker et al. 2020 · Gallet 2015
Tickets sold ↑ after a prior-draw winner.
Representativeness heuristic, amplified by blockchain's transparent, verifiable winner records.Guryan & Kearney 2008
Prize pool and total cost are jointly determined with ticket sales: every additional ticket grows the next pool, and gas fees shift with the same network activity that brings players in.
We need cost and jackpot variation that is exogenous to lottery demand itself.
F. Stock–Yogo 10% threshold is 13.43.J overidentification — instruments are exogenous.1.339, 95% CI [1.246, 1.433], p < .001. ≈ +6.3 tickets at the wallet mean.−0.086, 95% CI [−0.127, −0.046], p < .001. Players respond to ticket and gas.0.159, p < .001, but only n = 2 jackpot events observed — estimate noisy.Control-function residuals on both cost and prize pool are significant (p < .01), confirming endogeneity and validating the IV strategy. Effect magnitudes hold across both estimators.
Jackpot seeking and price sensitivity persist under pseudonymity, deregulation, and crypto-denomination. The gambling product itself — not its institutional environment — appears to be the primary behavioral driver.
The cost of participation is now product cost + infrastructure cost, each in a different volatile asset. Players respond to both consistently with classical price theory — a degree of economic rationality cognitive-bias accounts underweight.
If the behavioral risks of lottery consumption are not mitigated by the absence of regulation, consumer-protection frameworks must extend to decentralized platforms — even when operators are pseudonymous and borderless.