I should add that chess and poker have virtually nothing in common, because in chess you got complete information. Other than the human factor there is no random event. You can measure mistakes in centi-pawns, but many ways lead to Rome, which means there is no difference if you convert a winning position in 10 moves or 49 moves. Except for blitz games in a coffeehouse there is no hourly rate either.

If you want to compare chess hustlers to other hustlers, then pool billiard is the game.

Playing vs. a weakened engine is not the best idea as they play too strongly but when the second best move is a weak move.

Chess is not a complete information game in practice; poker is. The poker bots always play the best move, the chess engines do not, as the information is incomplete.

Playing vs. a much stronger player, bot or not, that rarely makes obvious mistakes, the weaker player is best to play whatever max the chance to get a draw, like Petroff where I spotted only a 1 in 3 chance for a strong 150 to 200 points weaker player to lose with black. Compare this to Najdorf (-e5, English attack mostly) where the stronger engine scored 3 wins out of 4 with white according to my simulations.

It is possible to exploit in many ways in chess also but it is often not successful.