I'm going to call an extremely loose raiser anyone raising more than 20% of all hands. I'm not entirely sure about optimal counter-strategies here, but a few initial thoughts.

First, it seems to me like this player is coming pretty close to an artificial increase in the blinds. If it's a raise to $20 in a 2/5 game with available stack depth of $500 (this player usually has less, even if he buys in for more, for obvious reasons--the real issue is how to get your own piece of it), then you have 25 times the bet as possible winnings.

The player has in a sense made a 7/20 game (or 7/25, etc.) out of a 2/5 game, but his "blind" is no longer a purely random hand--it's at least a little bit better than that.

As I see it, a low M game of this sort has a lot of similarities to short-stack tournament situations--except that there are rebuys. But it devalues sets and devalues suited connectors. Basically, one is getting into a kind of "Neanderthal poker" where the brute force of big cards is all that you can really play.

Also, while there's not real time limit in a cash game, there is usually an implicit one against this player, because he's going to bust out before too terribly long. So, you do want to play enough hands to get close to your "optimal" piece of this player.

As initial idea: I think it's pretty safe re-raising this player with QQ-AA or AKs (say he's right at 20% and still has a stack), if he's playing even more hands perhaps re-raising JJ-AA and AK--basically, this guy has limped in a 7/20 game. So, I'm just thinking one can pretend it's a 7/20 with shorter stacks.

If he's short-stacked and you're unlikely to run into problems with the rest of the table, I even think it's worth CONSIDERING putting him in with AJ or 88 (I'll gladly move in on such hands short-stacked in a tournament under certain conditions).

I don't think it's really worth playing little pairs for set value, though.

In position, one may well want to play any two big cards T or better--certainly J or better.

Usually, the table it doesn't take terribly long for the rest of the table not to be respecting this raise at all. What it seems to me to do is just take all the subtleties out of the game due to the lacking stack depth. And that's probably good for this raiser, because he can't play very well anyway.

Anyhow, those are my initial thoughts on dealing with these jokers. Feedback certainly appreciated. I think there's a lot of variation possible here--definitely a big difference between a 20% raiser, a 30% raiser, and a 40% raiser. And I'm still uncertain as to getting close to an "optimal" counter-strategy.