I mean it is possible that the model could work for someone with access to more capital and better fee structure than OP making $50 bets.
Still, this seems like a model that will likely work about 95% then fail on outliers ie "picking pennies in front of a bulldozer" model.