Bull vs bear debate
Two agents argue for and against; a judge decides the size. It's cloned turned off in your practice account so you can review it before turning it on.
How it works
- 1It triggersAccording to its frequency: daily.
- 2It reads the dataPrices and signals for BTC.
- 3It evaluates the ruleTwo agents argue for and against; a judge decides the size.
- 4It acts and notifiesIt trades in practice or sends you the alert, depending on the template.
About this strategy
A well-known problem with AI models is that they tend to agree with you. This template tackles it with a debate: one agent argues the bullish case, another the bearish one, both using the same market data and news.
A third agent acts as the judge: it reads both arguments and decides whether there’s a trade and how big, from zero up to the maximum you set. The more balanced the debate, the smaller the size.
Like every AI bot, it costs credits per call, and it’s worth reading the arguments in the first few weeks to understand how it decides. Orders always go through the risk engine with your limits.
The workflow
Click a node to see what it doesCustomize it before cloning
What you'll receive
Questions about this strategy
To force the counter-arguments to be considered before trading. It reduces the bias of a single model talking itself into an idea.
Yes. Every run stores both agents’ arguments and the judge’s decision in its trace.
It depends on the frequency: each run makes three AI model calls. The page shows a daily estimate before you clone it.
It's cloned in practice. Try it for a few weeks and move it to real money only if you're convinced, with your risk limits.
Learn before you switch it on
- AI trading agents vs trading bots: what’s the difference?What an AI trading agent is, how it differs from a classic trading bot, what it can really do and what it can’t. With examples and the risks to watch for.
- How to backtest a trading strategy without fooling yourselfWhat backtesting is, how to test a trading strategy on historical data and the mistakes that inflate results: overfitting, look-ahead bias, fees and slippage.