In a cash game, I see no reason to make the blind pay to play if you don't have a raising hand.

I still think the bottom line on raises is whether or not you can effectively make a continuation bet. If you can't, I'm for raising JJ-AA, and that's it, maybe also AK from LP.

If you can make a continuation bet and are getting a very high fold percentage on the flop (I'd put the figure at somewhere around 75% or greater before I'd want to do this), then I think you can get by with raising a much wider variety of hands than you advocate. My current criteria for the loose LP raise: Any pair, any two J or bigger, any 2 suited T or bigger, suited connectors.

Actually, this gives me a good excuse to go for another calculation. Let's say that your callers will fold the flop with a probability of p. And let's just say that you need to raise to $25 in order to get 2 callers. Then you need to bet $60 to take down the pot with a probability of p.

Then you're risking $85 to win $50. So, this is break-even when p*50 = (1-p)*85, hence p*135 = 85. So, p needs to be at least 63% for this to be profitable essentially on any 2.

Obviously, though, if you're raising any 2, it isn't going to be long until you start running into trouble. I think it keeps people guessing much longer if you take hands with some good potential for this raise, mainly because they don't happen very often. Also, on many of the hands I've mentioned (the pairs particularly, but also suited connectors), there's a decent chance of flopping a monster or a monster draw. So, you have a backup plan if you do get an unexpected caller. Anyhow, maybe it's just me, but I don't see the "raise any 2 sometimes" plan as particularly healthy poker.

And, if you have a lot of LP raising hands as part of a particular strategy, then I think it's also a good idea to have a lot of LP limping hands to go with them.