Bankroll Arithmetic, and the Second Bet You Make by Holding Crypto
An earlier version of this page described crash scenarios tested across «1,800 sessions», then «2,200 sessions», then «2,217 sessions» — three different figures for the same claimed research, inside one article. That is not a rounding difference; it is what invented numbers look like when nobody is keeping track. There was no session corpus. The figures are gone, along with the dozens of percentages built on them.
The subject is real, though, and it has two halves that the old version mixed together: the arithmetic of a bankroll, which is the same in any currency, and crypto price exposure, which is a separate risk you take on by holding one.
Half one: what actually sets the cost
Expected cost = stake per bet × number of bets × house edge. That is the whole of it. Variance decides any single evening; the formula decides the expectation, and four of its five inputs are yours:
| Input | Yours? | Effect on the cost of an hour |
|---|---|---|
| Stake per bet | Yes | Directly proportional |
| Bets per hour | Yes — pace, turbo, game format | Directly proportional |
| House edge | Yes — by choosing the game | Under 1% to about 8% depending on what you open |
| Session length | Yes | Directly proportional |
| Outcome of one bet | No | None you can plan around |
Published house edges differ by a factor of eight: under 1% on blackjack played to basic strategy with good rules, about 1.06% on the baccarat banker bet, about 1.41% on the craps pass line, about 2.70% on single-zero roulette against 5.26% on double-zero, and roughly 4–8% on most slots. No bankroll rule competes with that choice, because the choice is made before the first bet.
Half two: holding crypto is a second bet
A balance held in a volatile asset can be worth less when you withdraw it than when you deposited, and that movement has nothing to do with the games. It is market exposure, and you did not acquire it by winning. Two practical consequences:
- Decide in advance whether you are holding or converting. If a payout arrives in a volatile asset and you intend to keep it, the result of your session is no longer settled when the session ends.
- Network fees are yours, not a bank’s. On a busy network a small withdrawal can lose a meaningful share to them, which is the opposite of the usual claim that crypto is cheaper.
The one clean, verifiable datum in the crypto-energy debate, for what it is worth: Ethereum’s move to proof-of-stake in September 2022 reduced the network’s energy use by over 99% while carrying the same traffic. That is a measured change on a named date, unlike most claims in this area.
A bankroll plan that survives contact
- A total you are prepared to lose, decided before the first bet, in the currency you will actually withdraw in.
- A stake fixed in advance, and a limit on time as well as money — they run out at different speeds.
- Volatility matched to the money, read from each game’s own information panel rather than from a review. High volatility on a small bankroll is a faster route to zero, not a bolder strategy.
- Deposit caps set in the account, where something other than your own judgement enforces them mid-session.
- A written record of stake, duration and result. Memory over-weights wins, partly because a celebrated sub-stake return registers as one.
What a loss dressed as a win looks like is covered in the near-miss article, and the systems people reach for when a bankroll is shrinking are examined in the piece on betting routines.
Three different session counts for one claimed study is what invented research looks like. The real arithmetic is stake × bets × edge, and holding crypto adds a second, separate bet on the price.
Quick answers
How big should a casino bankroll be?
Large enough that losing all of it changes nothing in your life, because expected cost is stake times bets times house edge and the expectation does not change sign. Size it to what you can lose, not to a target.
Does crypto make bankroll management different?
It adds one thing: price exposure. A balance in a volatile asset can be worth less on withdrawal, which is market risk you did not choose by winning. Decide before the payout whether you hold or convert.
What happened to the crash-scenario research on this page?
It was invented. The article cited 1,800, 2,200 and 2,217 sessions for the same claimed study. No such sessions were run and the figures have been removed.
