Challenge settings
Later releaseChallenges are fixed and public: the conditions are written here, and a player can read all of them before starting. The system never generates a challenge from someone's behaviour, and never suggests a next one after somebody finishes.
Challenges are not part of the first release. The settings live here so the two rules that matter are fixed before anything ships: a challenge reward is never tokens, and the “share of people who have ever done it” figure is a required part of every challenge, not an optional marketing line.
Challenge definitions
| Code | Name | Type | Fixed conditions | Reward | % who have ever done it | Where the figure comes from | Status |
|---|---|---|---|---|---|---|---|
| CH-01 | Seven days without a bet | Taking a break | duration = 7 days tracks no_bet_since |
🎖 “Seven clear days” badge | 41% | Estimate hard-coded, nobody has finished one yet | Approved |
| CH-02 | Thirty days without a bet | Taking a break | duration = 30 days tracks no_bet_since |
🎖 “One month” badge + profile title | 12% | Estimate hard-coded, nobody has finished one yet | Approved |
| CH-05 | One weekend without opening the betting screen | Taking a break | duration = 2 days tracks no_bet_since |
🎖 “One clear weekend” badge | 74% | Estimate hard-coded, nobody has finished one yet | Approved |
| CH-03 | Fifty bets, at least 55% winners | Long betting run | games_required = 50 win_rate_target = 55% |
🎖 Badge + reconciliation report | 15% | Estimate from a 10,000-run simulation at the expected win rate of 47.6% | Approved |
| CH-04 | One hundred bets, at least 65% winners | Long betting run | games_required = 100 win_rate_target = 65% |
🎖 Badge + reconciliation report | 0.034% | Estimate from the exact binomial probability at the expected win rate of 47.6% | Approved |
| CH-06 | Two hundred bets, at least 60% winners | Long betting run | games_required = 200 win_rate_target = 60% |
🎖 Badge + reconciliation report | 0.1% | Estimate from a 10,000-run simulation at the expected win rate of 47.6% | Awaiting review |
The “% who have ever done it” figures are estimates, and players have to be able to see that
There is no accumulated real data yet, so none of the figures above come from users: the long-run figures come from a simulation, and the break figures are hard-coded. The player-facing screen has to print the word estimate beside every one of them, with a line explaining how it was worked out. Once there are enough real completions, the figure is replaced by the observed rate and the sample size is stated.
Why the reward field will not take tokens
Paying a large token reward for taking a break creates exactly the loop this product exists to break: stop for a week to build a stake for the next one. So the reward field here only offers badges and titles. There is no token amount to enter, and no way to convert a badge into tokens.
Rewards sit outside the token economy.
Three behaviours the system is not allowed to have
- No API and no screen generates a “next challenge” from what just happened. Choosing a new challenge is always something the player does deliberately.
- A challenge ends when the required number of bets is reached, whether the player is ahead or behind. There is no extension and no “play ten more to save the streak”.
- There is no option anywhere called double, all-in or double-or-nothing.
The report at the end of a challenge always compares the actual win rate against the theoretical house edge, even when the player hit the target.
Edit definition — CH-06
Nothing on this page affects players until challenges are included in a release.