A bitcoin skill game is a contest where crypto goes to whoever plays better, and the operator earns from a fee rather than from a built-in edge against you. You will meet the label on PvP lobbies with crypto stakes, on sponsored tournaments that pay in BTC, and on a good number of casino games that have no right to it.

The label decides two things. The first is whether getting better at the game can raise what you earn. The second is legal. In the US, games that pass the skill test are treated as competitions, not gambling, and that is why so much is marketed under the word.

Where the money comes from

There are three ways a skill game can pay, and you should know which one you are in before you deposit.

Pooled entry fees. Players stake into a match and the winner takes the pot. The operator takes a rake, which is a percentage of the money, and it does not care who wins. Bitcoin Chaser describes this as the structural difference from banked casino games, where the house plays against you at a fixed edge. A rake operator gets paid either way, so it has no reason to tilt the match.

Sponsor-funded prizes. A brand puts up a fixed prize. No pool of entry money exists, and the prize is the same size whoever wins.

Tradeable rewards. You win in-game items or tokens and sell them on a market. This is the play-to-earn version, and the payout depends on buyers being there when you sell.

The legal footing comes from the "predominance of skill" test. If player ability drives outcomes more than chance does, US courts and regulators treat entry fees as competition fees, not wagers. The federal UIGEA exempts contests that reflect the "relative knowledge and skill of participants." As of September 2026, 45 states allow cash skill play. Arkansas, Connecticut, Delaware, Louisiana and South Dakota bar it. Indiana and Maine limit skill card games to free play, and New Jersey does the same for domino games. Free-to-play is legal everywhere.

That framework comes from a guide to skill gaming in general, and the guide never mentions crypto. Whether a state treats a token prize the same way it treats a cash prize is a separate question, and the sources do not settle it. Regulators in the US, UK and EU are also examining whether in-game tokens count as securities.

Three places it shows up

A sponsored bracket: Mundo Dodgeball. Coinbase and the official League of Legends organization ran a 1v1 tournament with 1 BTC for the winner, worth about $83,937 when the report ran on September 22, 2026. Both players pick Dr. Mundo, fight in the Baron pit and may use only the Q projectile. That removes items, team composition and most of the randomness, which leaves something close to a pure skill test. The payout is a flat prize for one person. The top 8 also get tickets to the 2026 Global Finals match. Everyone else gets practice.

A rake lobby. Money Slither, a guide affiliated with a platform, describes the platform it features this way: a 10% rake charged only on cashouts, not on every round, stake lobbies at $1, $10 and $100, and instant withdrawals. All of that is the vendor's description. You can check it two ways: read the cashout terms, and withdraw a small amount before you deposit a large one. Assuming the rake works as described, the arithmetic looks like this.

Hypothetical: 100 matches in a $10 lobby

Say you deposit $100 and play 100 one-on-one matches at $10 each. The winner takes the $20 pot, there is no fee per match, and 10% comes off whatever you withdraw. Gas and network fees are left out because the sources do not give figures for them.

Win rate 55%: 55 wins x $10+$550
45 losses x $10-$450
Balance before cashout$200
10% rake on $200 withdrawn-$20
Net profit on the $100 deposit$80
Same player at a 50% win rate: withdraws $100, pays $10 rake-$10

A player who wins exactly half their matches loses money once the rake is taken. Skill pays only when you are better than the people in your lobby by more than the operator's cut.

A tradeable-reward game. In Gods Unchained, every card is an NFT you can trade. CryptoNews counted over 1,200 active players, with mixed reviews about slow matches. It counted 46,000+ for Gunz, a battle royale with PvP and PvE modes. Two things limit what that means for a player. A card you win is worth only what the market pays for it. And no one is independently checking these counts: DappRadar, which supplied most of the active-wallet figures the industry cited, shut down on November 17, 2025. A figure printed in a listicle is a claim until a project dashboard backs it up.

Crash, Plinko and the word "strategy"

Bitcoin Chaser's test fits in one question: can a better player consistently earn more than a worse one? Most crypto casino staples fail it. A crash game sets its bust point by RNG at the start of each round, so your timing does not change the odds. The "settings" in Mines, Plinko and dice change variance, not expected value. Video poker asks you to make decisions, but you make them against a house algorithm with a fixed edge. Money Slither puts the cutoff bluntly: if 95% of a game's variance is RNG, calling it skill-based is marketing.

If a game pits you against the house and the house has a built-in edge, no amount of in-game decision making makes it skill based in the meaningful sense.

Jon Fields
Writer, Bitcoin Chaser
Skill game Chance game sold as skill
Who you play against Other players The house
How the operator earns Rake or sponsor budget Fixed house edge
Does better play raise expected value Yes No, only variance changes
Examples from the sources PvP poker, Mundo Dodgeball, PvP crypto games Crash, Mines, Plinko, dice, video poker

Blackjack sits on the border. Bitcoin Chaser puts the house edge under 1% with correct strategy, but the edge still exists.

Before you deposit

Ask three questions. Who is on the other side of the match? How does the operator get paid, and on which transactions? Is cash play legal in your state? Then run a small test withdrawal before you trust any number on the lobby screen.

The pitch for "skill-based earning" is also a scam template. In 2025 the FBI's IC3 recorded $11.366 billion in crypto fraud losses, and investment fraud made up $7.228 billion of that. The category includes fake earn apps that ask you to deposit into a wallet before you can "unlock" winnings. The average crypto fraud victim lost $62,604. Recovery services that promise to get your crypto back cost victims another $1.4 billion. A legitimate skill game pays out from a pot or a sponsor, never from a fresh deposit you are asked to make.

Keep winnings in a wallet you control. Our guide to setting up a separate wallet covers funding and bridging without exposing your main stack. If the balance grows into real money, move it to a hardware wallet for NFT games before you let it sit.