[dankhank]

the link to spicyf's post is excellent, but i think it's incorrect application because your delayed raise doesn't save any bets. let me see if i can explain why:

-if you raise on the turn and he has a Q, he is going to 3-bet you. so you'll put in four bets when behind.
-if the river comes non-diamond/straight card and he has a Q, and you delay raise, he is probably going to 3-bet you. so you'll put in four bets when behind.
-if the river completes a draw and he has it, and you raise, he might 3-bet you. so you'll put in four bets when behind.

so the only situation where the delayed raise saves you bets is when a draw completes and doesn't help him, but this is tempered by the real possibility the villain is betting a draw you want to charge him the max on. furthermore the value of this scenario is low because he still might have a hand that beats you.

in other words if you are going to play aggresively after the turn card, it's hard not to get hurt if he has a Q, whatever you do.

so, what if you were planning to only call the river if no draw completes (based on when ahead / when behind reasoning)? would you then be saving bets with a delayed raise plan? yes but you wouldn't be extracting the max from A7, 99, etc., and at its heart spicyf's theorem seemed to be about maximization not just saving bets.

the main thing is how confidently you can put this opponent on a Q (versus a draw or a smaller pair) when he checkraises the flop. after ten orbits some opponents are predictable enough where you can lean one way, in which case you can take the appropriate turn action. barring that, this makes good sense, and your seeing how a third diamond might protect your raise is a nice logical grab, and so is at least in the spirit of the theorem.