Crash game myths and misconceptions

Key takeaways

• No round is “due”: bustabit’s documentation states that after five losses in a row, the chances on the sixth round are what they were on the first.
• Wang and Pleimling simulated Martingale-type strategies in crash games and concluded that they increase the risk of huge losses.
• In June 2026, Check Point Research reported on a fake “Aviator Predictor” whose downloads delivered clipboard-hijacking malware.
• Whether Spaceman’s demo uses the same math as the real-money game is not published; no source we found says either way.

Crash game myths about streaks and patterns

A crash game shows its recent results on screen, and a list of numbers invites the eye to find a shape in it.

Myth: after a run of low rounds, a high one is due

No round is due. The documentation for bustabit puts it plainly: after five losses in a row, the chances on the sixth round are what they were on the first. Expecting the run to turn is the gambler’s fallacy.

Runs of low rounds are also ordinary. Under the constant-RTP model at an assumed RTP (return to player) of 95.5%, five rounds in a row below 2.00x occur about 3.9% of the time (our calculation), and the chance that the next round reaches 2.00x is the same after such a run as before it. The losing-streak arithmetic covers other targets.

Myth: the recent results show a pattern you can read

A results screen records; it does not forecast. According to the Game Help, Spaceman’s Statistics panel lists up to 500 recent crash multipliers and charts their distribution by range, and the same document describes each round’s end point as a random, predetermined value. Clusters of low rounds and tidy-looking alternations are what random sequences look like when enough of them are laid side by side.

Myth: the game reacts to a player’s bet or to other players

The crash point is fixed before the multiplier moves, so it is not adjusted during the round for the size of a stake or the moment someone cashes out. According to bustabit’s documentation, results were decided before the player showed up and cannot react to how the player plays.

A round is a shared event, but each player bets against the house, not against the others; Wang and Pleimling’s study of online gambling data, which covers crash games, notes little or no interaction among players. For games offered under a British license, UK Gambling Commission standard RTS 7 also bars adaptive behavior, meaning a game that compensates for earlier results. The page on how an RNG (random number generator) is tested covers that standard.

Myths about cashout plans and systems

Myth: cashing out early is safer in the long run

A low target wins more often and swings less, and that is all it does. Under the model, the chance of reaching a target falls in proportion to what the target pays, so the expected return is the same at every target; bustabit’s documentation says of its own game that neither low nor high targets overcome the house edge. The trade-off is tabulated under cashout timing.

Myth: doubling after a loss wins the money back

Doubling the stake after each loss, the Martingale approach, leaves the expected return of every round unchanged. Wang and Pleimling simulated Martingale-type strategies in crash games and concluded that the strategy increases the risk of huge losses, and Wizard of Odds reaches the general conclusion that no betting system withstands the long run. The page that examines Spaceman strategies works through this approach and others.

Myth: a predictor app can see the next crash point

In our reasoning, there is nothing for such an app to see: the result is random and predetermined, and it cannot be derived from past rounds or from anything public before the round. What has been documented is harm to the people who download them: in June 2026, Check Point Research reported on a fake “Aviator Predictor” whose downloads delivered clipboard-hijacking malware. The page on predictor apps has the detail.

Myths about RTP and fairness labels

Myth: the RTP is what comes back in each session

RTP is an average over a very large number of rounds, not a refund rate for a session. A session of a few hundred rounds is far shorter than that, and its result can sit well above or below the stated figure. The page on house edge and RTP explains the difference between theoretical and actual return, with a regulator’s worked example.

Myth: every crash game is provably fair, and the label means better odds

Provably fair verification is not universal. Some games document it in detail: bustabit publishes its hash chain, and Spribe describes a seed-based check for Aviator. Others rest on a certified RNG, so the label has to be confirmed in the information inside each game; what the documents say about Spaceman is set out under Is Spaceman fair?

The label is also about tampering, not generosity. A successful check shows that a result was not altered after it was committed, which says nothing by itself about the return: bustabit describes its game as provably fair and publishes a house edge of 1%.

Myth: demo rounds show how real-money rounds will go

Whether a demo uses the same math as the real-money game is not published: Pragmatic Play’s product page offers a Spaceman demo behind an 18+ confirmation (as checked on 2 October 2026), and no source we found says either way. Even with identical math, a run of demo rounds says nothing about later rounds, for the same reason the results history does not.

Sources

• Wang and Pleimling, Online Gambling of Pure Chance: Wager Distribution, Risk Attitude, and Anomalous Diffusion, Scientific Reports 9, 14712 (2019)
• UK Gambling Commission — RTS 7: Generation of random outcomes
• Check Point Research — report on fake tools delivering a clipboard hijacker, 17 June 2026
• Spribe — Provably Fair
• Pragmatic Play — Spaceman product page (as checked on 2 October 2026)
• bustabit — published documentation (cited by name, not linked)
• Wizard of Odds — betting systems analysis (cited by name, not linked)
• Pragmatic Play — Spaceman Game Help, version 1.7 (change log to 10 July 2024), as distributed to licensed operators (cited by name, not linked)