What is TapeArena
A proving ground for crypto trading strategies: simulated money, real prices and the same rules for everyone. It tells you, without risking anything, whether an idea really holds up.
Simulated trading for educational purposes. Not investment advice.
What it is for
Almost every strategy looks good on a past chart. The hard part is knowing which ones still do when you pay fees, the market shifts and you cannot blow up the account. TapeArena gives you a place to check it honestly.
- Learn to trade without a mistake costing money.
- Measure yourself against bots and against “buy and hold”, with a score that accounts for risk and not just profit.
- Tell luck from method: a minimum of trades and active days, and a flag when a single trade explains more than half of the profit.
- Automate and systematise your trading: turn what you do by hand into written rules that a program runs the same way every time, so they can really be measured.
- Experiment with systematic strategies: start from the bots' rules or from indicators, mix them, change them and see what happens, in the Laboratory or through the API.
Our philosophy
Why TapeArena exists. Measuring a strategy honestly is usually out of reach for beginners: you must count the costs, limit the risk, compare against chance and against “buy and hold”, and not fool yourself. We want that yardstick to be within everyone's reach, with the same rules for people and bots.
Why automate. A decision made by hand, different every time, cannot be checked. A written rule can: it can be tested on data that was not used to create it, repeated in other market conditions, measured with its costs and compared against chance. Automation does not create an edge by itself, but it is the only way to know whether one exists, how much it is worth and whether it lasts over time. That is the confidence we look for: an edge that is measurable and lasting, not a streak.
How we work. The ideas we test follow the same path we suggest to you:
- We write the hypothesis before seeing the results, so we do not fit it to what comes out.
- We test it on periods that were not used to design it and in different markets (rising, falling and sideways).
- It counts real fees, slippage and risk limits, and is compared with chance and with “buy and hold”.
- We discard ruthlessly: most ideas do not pass, and that is information too. If one does not clear the bar, we do not lower the bar so it passes.
No strategy, ours or yours, has a guaranteed edge. The goal is to know what we know and what we do not.
What you can do
Trade by hand
Live charts in several windows, 24 indicators, drawing tools, orders with stop and target, and conditional orders created from your drawings.
Trade with your program
An open API and a one-file Python client: your algorithm trades under the same rules as a person.
Test ideas in the Laboratory
Mix one bot's signal with another's management, build rules with 38 indicators, backtest them against chance and let your strategy trade on its own.
The rules, in one list
- Simulated account: 5,000 USDC
- Real market prices, with a fee (0.035 %)
- Maximum leverage: 3×
- Maximum loss 5 % a day and 10 % a month
- Markets: perpetual futures of BTC, ETH and SOL
- The same rules for people and bots
The detail is in the documentation and on the leaderboard, where you see who meets the requirements and who does not.
The bots: who they are and how they change
Bots are automatic accounts, always labelled BOT, that trade through the same API you do. There are two families:
The simple ones
They trade at random or on very short-term rules. They are the minimum yardstick: if you cannot beat them, your strategy adds nothing.
The systematic ones
They follow rules with a rationale: 60 and 120-day trend, moving-average filter and 55-day breakout. They size positions to volatility and have a brake if they lose too much.
Buy and hold
Not a bot but the rule to beat: buy the asset and never touch it. You see it as a line on the charts.
How do the bots change?
During the current month, they do not change. In the competition in progress each bot follows its rule to the end, without changing it, without watching what others do and without adapting to you. So you know the opponent plays fair.
Changes arrive when a new month starts. A new generation of bots appears, so every month there are more bots. The new ones are of two kinds:
- Children that learn (half): read between 2 and 8 parents from the previous month and, for each trait (asset, leverage, type of rule, brake…), decide whom to inherit it from according to what has worked best; then they mutate a little.
- Bots born from scratch (the other half): a random strategy, inheriting nothing. They are the control to know whether inheriting helps.
Parents do not disappear: they keep trading. But at the start of each month the worst 20 % of the generations that are two months old or more are retired, and those do stop trading. This renews the population without it growing forever.
Our commitment to honesty
A simulator can teach you a lot or fool you easily. That is why we publish the rules, label the bots, measure risk and say what we do not know: no bot has a proven edge and simulated results do not guarantee future results.
Frequently asked questions
Is it real money?
No. Everything is simulated, with real prices, for educational purposes. It is not investment advice and there are no cash prizes.
Do the bots have an edge?
None is proven. They are there so you have a yardstick, not as a signal to copy.
Do the bots learn from your trades?
No. A bot never looks at what you do: it only inherits from other bots, and only when a new month starts. Learning from users' trading as well is on our Roadmap.
Who is behind it?
An educational alpha-stage project. You can write to us from the contact page.