HFAP Question re: randomizing pre-flop play

Since I usually play against the same recurring cast of characters at the local 10-20 game, I've taken to applying a little game theory and randomizing my fold/limp/raise decisions with certain hands for deception purposes.


In HFAP, S&M recommend "occasionally" raising pre-flop in early position with hands such as 67s or other suited connectors, in order to mix up your opponents who are used to seeing early raises only from big hands such as group 1 or 2. What constitutes "occasionally" in a "typical" medium-tight 10-20 game? I've been doing something along the lines of 25% with 67s, 68s, 78s, 79s, 8Ts; 33% or so with 89s, 9Ts, 9Js, TJs; I rarely play anything smaller early. In fact, I rarely even play most of these cards early -- depending on the player lineup, I'll usually fold them if I'm not raising. The question is whether playing them 25% of the time is too frequent? Not frequent enough? Anybody done sims to determine optimal frequency of fold/limp/raise decisions with connectors? (BTW, I randomize based on suits or on what cards have shown up on the previous board. (e.g. I'll pick a hand; the middle card on the flop is a diamond; the next 8dTd I get dealt early I'll open-raise; I also re-randomize every now and again.))


When I drop down to 6-12 or if my opponents are weak, I'll usually revert to straightforward play -- but I've found that I really enjoy exasperating the better players when they figure me for a busted AK draw, they raise against a ragged-seeming board, and I show down 68s for the nut Str8.


Any comments would be appreciated....

RK,


I'm only taking a stab at this based on intuition, not a real academic grounding in game theory.


My guess is that 25% percent with these hands is far to often and maybe even too difficult to randomize properly (that is not going overboard). Seems to me that 5% may still be too frequent, I'm pretty sure that I do it far less often than that even, and I don't think I can be much the worse for it, though I could be wrong and I'm anxious to see what others think. This play doesn't have to work very often to have a desired lasting effect among regulars. Besides, many players won't distinguish this raise from a raise with the same hand in later position when there are no limpers, a play which I make noticibly more often, with probably the same effect and less exposure.


JMHO,


Mike

Your making these plays 25% and 33% of the time with slightly better hand. This is far to high imo. And bordering on being the "live" one in the game. I'm not going to pretend to know a percentage that is correct, but I would guess less than 5% of the time. As Mike points out your not much worse off if you never do it. Or just once in a very long time. But you will be much worse off if you do it too much. If you feel you need to make these types of plays in the game you regularly play in, the game is probably not a good one. And you should look for a more profitable game if your playing poker to make money. If your just doing it for the competition and don't care about the money then stay in that game and do your best.

As far as HPFAP goes, it does explicitly say to make these types of raises about 1/3 of the time if you are the first in from early position. Later, it says to reraise with these hands "occassionally" from middle position; your 25-33% sounds reasonable to me. And like you say, against unobservant opponents you probably shouldn't ever need to make these raises.