Game of skill or chance? An 8-question test for any competition

In a game of skill, better decisions reliably produce better results. In a game of chance, they don't, because the outcome is set by something no player controls. Most games sit somewhere between the two, which is why the label on its own tells you so little.

The quickest check comes from the investor and author Michael Mauboussin: ask whether you can lose on purpose. You can deliberately lose at chess or poker, so skill is involved. You can't deliberately lose at roulette or a lottery, because your choices don't move the result. In the United States, many states put the same idea in legal terms and ask whether skill or chance is the dominant factor in the outcome.

Ask eight questions about how the game actually works.

1. Is the underlying activity already accepted as skill? Playing chess, running a business and trading are ordinary, skilled things adults do. A competition built on one of those starts from a far stronger position than one built on a game whose status is already in dispute.

2. Does the player act during the contest, or at the start? If you make one choice and then wait, something outside you decides the result. If you keep deciding throughout, you do.

3. Do the player's decisions settle the result? If a team's score, a card, a die or a stock's weekly move settles the outcome, skill can improve your chances of picking well, and that's all it can do.

4. Is the randomness the same for everyone? Some randomness can't be removed. The question is whether it's shared. In duplicate bridge every table plays the same deals, so the deal stops deciding anything and the play decides everything.

5. Does the scoring punish all-in play? If the best plan is to stake everything on one swing and hope, the game rewards chance whatever its marketing says.

6. Do the same players keep winning? This is the signature of skill. When researchers picked poker players as skilled before a World Series of Poker, that group returned more than 30% while everyone else lost 15%. In a game of chance, last month's winners are no likelier to win this month.

7. Can anyone verify the result? If the organizer alone sees the prices, the cards or the scores, you're taking the skill on trust.

8. Is the organizer neutral? An organizer who earns the same whoever wins has no reason to tilt the game. One who profits when players lose does.

Here is how common formats score on the test.

Chess is the pure case, with no randomness at all. Poker passes most of the test but deals different cards to different players. Fantasy sports pass the "winners repeat" test, but once the lineup is set the athletes decide the result.

Trading is a game of skill when you're measured against other traders. In almost every country it's an ordinary, lawful activity for adults, regulated as finance rather than as a game, because the result depends on the trader's judgment: what to trade, how much, which direction, and when to get out.

Against the market on a single day, any trade can look like chance. Against other traders, on the same prices, over many rounds, the same people keep finishing near the top. That persistence is what separates skill from chance in any game, and it's why an active trading tournament scores as well as chess on the test while starting from an activity nobody disputes.

No, and the difference is question 2. A fantasy stock game asks you to pick stocks and wait. A trading competition asks you to trade.

The best known fantasy stock case is the SEC's 2016 order against Forcerank, an app whose players ranked ten stocks and then waited a week for the stocks to decide the winners. The SEC treated each paid entry as a security-based swap, fined the company $50,000, and did not accept its description of the game as skill-based. Run through the test, Forcerank fails questions 2 and 3: after entering, the player did nothing, and the securities settled the result. A contest like that works as a side contract on how particular securities will move, which is exactly what securities regulators oversee, because contracts of that kind can be used to hedge or speculate on securities outside regulated markets.

An active trading contest is a different activity. Nobody in a 30-minute match rides a price, because prices rarely move far enough in that time for buying and waiting to win anything. Traders decide whether to trade, which market, long or short, how big, how much leverage, when to cut or take profit, and whether to go again, many times in one match. Two traders can watch the same market make the same move and finish first and last. In our view that's why the Forcerank reasoning doesn't fit active trading tournaments, and the counsel we've consulted read it the same way.

Even in active trading, one plan can look like a shortcut: maximum leverage on a single all-in trade that happens to come off. The Sivex score is built to make that a losing plan. It is your return minus 1.5× any drawdown beyond 10%, published before every match and the same in all of them.

Leverage makes the penalty hard to escape. At 100x, a 0.1% move against an all-in position is a 10% drop in the account, and a 0.2% move is 20%. Bitcoin moves that much within minutes on an ordinary day. Two sessions scored with the live formula show the effect:

The bigger profit finishes below the smaller one. All six sessions are worked through in how much drawdown costs in a trading tournament.

Three more rules point the same way. A new position is locked for 30 seconds, so trades can't be flipped like coins. Every result moves a chess-style Sivex Rating that carries over between matches. And the Drop pays $5,000 every two weeks to the highest-rated qualifying players, so the reliable way to the top is to keep beating other traders.

It's early, but the first signs are there. In Sivex's first four tournaments, 6 of the 36 traders who played three matches finished in the paid places twice, against about four that chance alone would predict (checked 4 November 2026). That's far too little to prove anything yet. As matches add up, the same comparison becomes the kind of evidence researchers used to show that poker and day trading reward skill, except gathered in the open, with every trade publicly verifiable.

On any platform, including this one. Ask the eight questions, or start with the shortest version: can you lose on purpose, and do your decisions keep changing the result after you enter? If a competition calls itself skill-based but you act at the start and then wait, something outside you settles the result, the scoring rewards going all in, and nobody can check the outcome, the label is doing work the game isn't.

Active trading, scored against all-in play and verified in public, passes all eight. That's what 100% skill-based should mean. For how a Sivex match compares with a game of chance in more detail, see skill versus chance in a trading tournament.