When Does a Rollover Jackpot Make a Lottery Ticket a Good Bet?
Almost never. But the math is interesting — and the threshold is higher than most players think.
The naive model says: when the jackpot is large enough, expected value of a ticket becomes positive. The truth is more interesting and considerably less optimistic.
The naive break-even
Ignoring taxes and shared jackpots, a Powerball ticket has positive gross EV when:
Jackpot × (1 / 292,201,338) + $0.32 ≥ $2.00
Jackpot ≥ ($2 − $0.32) × 292,201,338
Jackpot ≥ ~$491 million
So the naive threshold is about $491M advertised jackpot. Past this, an idealized rational player should buy tickets.
The lump-sum adjustment
Advertised jackpots are 30-year annuities. The lump-sum cash option is typically 50–55% of that. Recomputing with 0.55× factor:
Jackpot_lump ≥ $1.68 × 292,201,338 / 0.55
Jackpot_lump ≥ ~$892 million advertised
Accounting for the cash option, the threshold rises to about $892M.
The tax adjustment
Federal tax on lottery winnings: 37% top bracket. State tax: typically 5–10% (some states like Florida and Texas have none). Using 42% US-average:
Jackpot ≥ $1.68 × 292.2M / (0.55 × 0.58)
Jackpot ≥ ~$1.54 billion advertised
Now we need a $1.54B advertised jackpot before a single ticket has positive expected value, ignoring shared-jackpot risk.
The shared-jackpot adjustment
Here's the part most analyses miss. As the jackpot grows, more people buy tickets, and the probability of multiple winners rises sharply. Empirical data from large jackpots:
- At $500M: ~140M tickets sold; P(at least 1 winner) ≈ 38%; expected co-winners | win ≈ 1.2
- At $1B: ~280M tickets; P(winner) ≈ 62%; expected co-winners | win ≈ 2.0
- At $2B: ~500M tickets; P(winner) ≈ 82%; expected co-winners | win ≈ 3.5
Your conditional payout, given a win, is divided by (1 + E[co-winners]). At $2B, you take home about 1/4.5 of the advertised jackpot — bringing the after-tax, after-split value of a 'winning' ticket down to:
≈ $142 million
So the effective post-tax, post-split jackpot at a $2B advertised level is about $142M. Break-even:
≈ $0.81 per $2 ticket
Still negative even at a $2 billion jackpot — by about $1.19 per ticket. The shared-jackpot effect almost completely offsets the apparent jackpot-size advantage.
When does EV genuinely turn positive?
Realistic models suggest that net EV-positive lottery plays happen rarely or never in practice. The only edge cases:
- A jackpot rollover so high that the cash-after-tax-after-split value exceeds the ticket cost × C(n,k).
- Tickets bought in states with no state income tax (raises after-tax payout by ~5%).
- Combinations specifically chosen to avoid common picks (reducing E[co-winners] by 30–40%).
Even at $2B+ jackpots, the math is marginal. Treat the lottery as an entertainment expense, not as an EV-positive investment.