Refactor DeployableTraderAgent to 1. predict a probability for a market then 2. place a bet based on its BettingStrategy

[evangriffiths]

May need a refactor of DeployableTraderAgent.process_bets (see

prediction-market-agent-tooling/prediction_market_agent_tooling/deploy/agent.py

Line 281 in 69ec60e

def process_bets(self, market_type: MarketType) -> None:

)

Currently it does:

result = self.answer_binary_market(market)
amount = self.calculate_bet_amount(result, market)
market.place_bet(amount=amount, outcome=result.decision)

The abstraction is a bit off here if we want to use get_kelly_bet here, as we want to calculate the direction and amount together, so it would look like:

prediction = self.predict_binary_market(market)
bet = self.get_bet_from_prediction(market, prediction)
market.place_bet(amount=bet.amount, outcome=bet.direction)

[evangriffiths]

@gabrielfior says "don't change the default for all agents... the dune dashboard will get messed up. Try on one or a few agents first, see how it changes accuracy(?) and ROI"

[evangriffiths]

Re-opening, as #352 didn't actually change the default

[gabrielfior]

Re-opening, as #352 didn't actually change the default

@evangriffiths do you mind closing this ticket in favor of #365?

[evangriffiths]

Sure. Have renamed to make it clearer what we're closing as solved 😄