[rhm]

awm, on 2013-June-21, 19:26, said:

In all I think Justin's theory is reasonable, but since I tend to bid 3♥ on these hands anyway it leaves me without a lot of relevant experience (yes, I do pretty well bidding 3♥ in this situation).

I subscribe to this philosophy.
I believe standard requirements for a jump raise are too high in a strong notrump framework with frequent three card raises.
This is reinforced if you open light, what most people do nowadays. Weak notrump systems have an advantage here in my opinion and you can not just copy their logic here.
After a single raise what is responder supposed to do with a borderline hand, when opener's strength can vary and the degree of fit is unclear?
If responder passes, a lot of good, even cold games are missed and often are. On the other hand if responder game tries, he will often leave the last plus behind him.
Game tries are over-judged.
Much better for opener to be aggressive when he has four card support. That way a lot of good games are reached, without giving opponents a lot of information and going down in three is remarkably rare in my experience.
When it happens opponents often had a contract of their own.
I do not mind bidding game with four card support and 18-19 balanced and I also do not mind splintering with a slightly weaker hand than in standard. It occurs more often and seems to work well.
That way responder does not stretch to invite when he receives a raise, because opener is really minimum or has only three card support.
With a four or a weak five card major responder will need close to opening values to move.
A side benefit is that it gets more dangerous for opponents to balance.

Rainer Herrmann