the robinhood gold card has a mystery box: 1,000 points per spin, for a shot at cash-back boosts or a 100,000-point jackpot. i ran the published odds and expected value. the average spin returns about 177 points, or 0.18 cents per point, against the 1 cent you'd get redeeming straight to your brokerage account. 94.5% of spins return less than the stake you put in.
most card reviews stop at "3% cash back." they don't mention robinhood quietly added a game.
The Feature Hiding in the Rewards Tab
open the banking app, tap rewards, and there it is: spend points to open a mystery box and try to win a prize. robinhood's help page gives you one flat line: "for 1,000 points, you can open a mystery box to earn a reward." it lists no odds.
for the odds you have to read the rewards program rules pdf, the legal document rather than the marketing page.
the bottom line: the mystery box is a points-to-cash-back lottery, and the published odds say it pays back about a sixth of what you'd get skipping the game entirely.
there's a daily spin limit, and the terms reserve the right to change every number below without notice.
Every Prize and Its Exact Odds
the complete published table, straight from the rules.
| Prize | Odds |
|---|---|
| 100,000 points | 0.001% |
| 10,000 points | 0.20% |
| 7,500 points | 0.30% |
| 5,000 points | 0.50% |
| 2,500 points | 1.00% |
| 1,500 points | 1.50% |
| 1,250 points | 2.00% |
| 10% grocery (next run) | 5.00% |
| 5% grocery (next run) | 8.00% |
| 10% fill-up | 2.00% |
| 5% for 30 days | 0.25% |
| 5% for 1 week | 2.00% |
| 10% for 24 hours | 0.50% |
| 5% for 24 hours | 6.00% |
| 4% for 24 hours | 17.00% |
| 15% next purchase | 3.00% |
| 10% mobile-wallet purchase | 4.00% |
| 15% / 10% / 5% next restaurant order | 5.00% / 25.749% / 16.00% |
Data table
| Item | Value | Share |
|---|---|---|
| 100,000 points | 0.001% | 0.0% |
| 10,000 points | 0.2% | 0.4% |
| 7,500 points | 0.3% | 0.6% |
| 5,000 points | 0.5% | 0.9% |
| 2,500 points | 1% | 1.9% |
| 1,500 points | 1.5% | 2.8% |
| 1,250 points | 2% | 3.8% |
| 10% grocery (next run) | 5% | 9.4% |
| 5% grocery (next run) | 8% | 15.0% |
| 10% fill-up | 2% | 3.8% |
| 5% for 30 days | 0.25% | 0.5% |
| 5% for 1 week | 2% | 3.8% |
| 10% for 24 hours | 0.5% | 0.9% |
| 5% for 24 hours | 6% | 11.3% |
| 4% for 24 hours | 17% | 31.9% |
| 15% next purchase | 3% | 5.6% |
| 10% mobile-wallet purchase | 4% | 7.5% |
every cash-back prize is capped: $500 on most, $300 on the fill-up, $2,500 on the week, $3,000 on the 24-hour boosts, $10,000 on the 30-day one.
read the shape, not the jackpot. the odds sum to exactly 100%, so every spin pays something. the modal outcome is a 10% restaurant-order boost at 25.749%; the 100,000-point jackpot at 0.001% is one in 100,000 spins.
The Expected Value Math
the odds table is a probability distribution, so we can do expected value. points prizes are simple: 1 cent each. cash-back prizes need an assumption, so assume a $500 qualifying purchase, which hits most caps. multiply each prize by its probability and add them up.
| Outcome group | Probability | EV contribution |
|---|---|---|
| Point prizes (1,250-100,000) | 4.5% | 141 points |
| Cash-back boosts (capped at $500) | 95.5% | ~36 points |
| Total EV per 1,000-point spin | 100% | ~177 points |
177 points back on a 1,000-point spin is 0.18 cents per point. redeem those same points to your brokerage account and you get a locked-in 1 cent each: $10, no odds, no caps.
the point prizes carry the average: a 2% chance at 1,250 points is worth 25 points of expectation by itself, and the jackpot at 0.001% adds 100. the cash-back boosts look flashier but contribute only ~36 points across 95.5% of outcomes, because most land on a capped 5% boost a real purchase rarely maxes out. halve the cash assumption and the total drops to ~159 points.
the distribution is brutal: only 5.5% of spins return more than the 1,000 points you paid. the other 94.5% come back short.
the honest read: the mystery box has negative expected value against the 1-cent floor. you are paying 1,000 points to receive, on average, 177 back.
The Trap: Points Are Worth Less as a Stake
here's the catch nobody talks about, and it's the reason the feature exists.
you earn points at 3 points per dollar of spending. redeem them straight to your brokerage account and each one is worth exactly 1 cent. that's the floor. 1,000 points is a certain $10.
spin those same 1,000 points and you're buying a lottery ticket with $10 of guaranteed value. expected value says the ticket returns about $1.77 on average, and only 5.5% of outcomes beat the stake.
robinhood's terms make the asymmetry explicit: cash-back point redemptions are non-reversible. commit points to the box and the decision is final.
so the question isn't "is the mystery box +EV?" it's "do i want a certain $10, or a lumpy bet worth about $1.77?" for almost everyone, the answer is the $10.
When a Spin Actually Makes Sense
there's one situation where the math flips: you have points you'll never redeem.
the 1-cent floor only exists if you hold a robinhood brokerage account. the help page says you need an individual investing account to redeem cash back. if you don't have an account, your points are worth zero, not $10. spin them and any return beats nothing.
the second case is narrower. if a $500+ grocery run or restaurant order is already in your plans, a spin is a cheap hedge. you'd spend the money either way, so the boost is pure upside if it lands. the odds say it usually won't.
beyond those two, the mystery box is a novelty, not a strategy.
Pros
- published odds let you do real expected-value math (rare for a card feature)
- every spin pays something: the odds sum to 100%
- the 10% restaurant outcome is genuinely high-value if it lands
Cons
- 1,000 points is a guaranteed $10 you give up to play
- expected return is ~0.18 cents per point vs the 1-cent floor
- 94.5% of spins come back short of the stake
- cash-back prizes are conditional: no qualifying purchase, no value
- redemptions are non-reversible
- robinhood can change every number without notice