I stay away from them in general
but I will play them if I see my opponents over valuing A t-k (as is often the case.
if I'm playing A2 with a board of A27, and my opponent has AK, I will take him all in much of the time. in that case, I risked alittle, and I got alot.
doesn't happen often but when it does, it gives you a good boost.
someone mentioned that they like A 2-5 more than A 6-9 because of the straight posibbilities, and I used to agree because straights are just so profitable. But I've since changed my mind. With a hand like A K-T, the only straights you can get with both cards are nut straights. The opposite applies to A2-5. When you play and hand like that, you are vulnerable to losing a large sum of money to small suited connectors like 67 or 46. many times, you'll lose a good chunk of money to someone playing a hand that you think is ridiculous like 36 (with a board of 245, you hold A3) but your mistake was that you thought your hand was invincible because, "nobody would possibly play 36" oh believe me, they're out there. (i'm one of them)
For this reason, if I ever do play A2-5, I'm weary of that possiblity, but A 6-9, thats one let thing to worry about.
if I'm under enough pressure, of course, I'll go all in with A-anything but my preffered A-Rag is A8 or A9. my theory is thus:
There are twelve cards behind A, and that is K-Q-J-T-9-8-7-6-5-4-3-2.
if i'm under alot of pressure, and I have A8, I can beat 6 other possible Aces out there (A2, A3...A7), but 5 other possible Aces will beat me (AK, AQ,...A9). At A7, 6 aces will beat me, and I can beat 5. (this applies before the board is flipped of course) it may be a small difference, but its what I use to figure out what I'd like to play.