Full table is definitely better for solid play - I think the more passive play on there also helps, as you're seeing flops more cheaply, so the variance is less. While maniacs may be profitable, I'm figuring they may be more beneficial at full ring games, where you can use them to get money out of other players who won't respect their bets.
On 6-max tables, I'm still trying to work out what the best strategy is, I guess. My starting hands were basically based on Hutchison point-count rules, plus playing a few hands which would either fit or fold from the Big Blind since the effective cost was halved. When someone else raised, I'd question whether my hand had backup as well. My starting hands were the same ones I'd been pretty consistently profitable at on Stars... I suspect the problem is that the players I've been against on Full Tilt have generally been more aggressive.
I'm getting the impression that the Hutchison method doesn't really work so well against really aggressive players (particularly short-handed), as it favours A2-type hands which will often flop just a low draw with back door high draws. You then have to determine whether you're getting the right odds
on the combined draws against looser players who have every chance of having the stronger made high hand (through on average having less low cards).
The question is what hands it improves the value of. I'm guessing that A2+Pair and A3+Pair hands might benefit. While suited cards will occasionally flop a straight, that's pretty uncommon (1%?), so again you're looking at paying for draws, just hopefully you'll have more draws to use, and even the 1 suited card will give you a back door draw to marginally improve your odds. I'd therefore not necessarily worry about that too much. A23 is almost certainly improved in value, because other players are more likely to play A2 hands, and you'll sometimes be able to cap the pot with a redraw to counterfeit their lows. AA2, AA3 hands, particularly suited or double-suited I'd figure have the best potential, as they may remain the best on the flop (occasionally), have reasonable high potential anyway, and have at least a half-decent shot at nut low.
I know about the potentially significant number of outs you can get from wrap straight draws, but I'm not sure how significant they are in terms of starting hand selection - if they're connecting low cards or the like then they'll either not be looking at nut low most of the time, or will generally be drawing to the nut straight. That leaves 4 paint cards, which haven't fared too well for me either - I guess there I've possibly played a little loose, and should follow the Hutchison high-only selections in which any 4 high cards has to also be either double suited, suited with a pair, or double-paired, so you're looking at either 2 good suits for potential flushes along with the potential for wrap straight draws, 1 suit and a shot at trips, or 2 shots at potentially nut trips with some lesser straight potential.
In any case, do people agree on it being not just playing tighter, but placing slightly different emphasis on hand types that's important, and if so which hands would you look to play short-handed against maniacs?
For the record, this time I opted for the full-ring $0.5/1 table and came away with a modest profit. I actually made one of my drawing hands this time, and mostly made the most of made hands. Maybe I'll revert to the full ring games... I'd figured that 6-max might be more profitable due to the maniac players who'll happily pump the pot with nothing, but at the moment it's looking like that isn't the case, or at least that variance is much, much higher there.