I could do with some opinions here. I've been under the impression that making a bet to test the water when a scary flop shows up is good play for several reasons:

1) If your bet is raised it confirms your suspicions and you can fold without further loss
2) If the opponent calls then you can use your judgement on their player style to determine what they might have, and
3) The opponent might fold allowing you to take the pot down there and then.

However, on another poker forum an advanced member said that it was generally accepted that betting for info is a bad idea. What are your opinions on this?

P.S. An example would be this hand someone posted:

$2/$4 No Limit Holdem
5 players

Stacks:
UTG ($816.80)
CO ($428.00)
BTN ($389.95)
SB ($76.00)
Hero (BB) ($398.00)

Pre-flop: ($6, 5 players) Hero is BB K K
1 fold, CO raises to $14, BTN calls $14, 1 fold, Hero raises to $58, CO folds, BTN calls $44

Flop: 6 3 A ($132, 2 players)
Hero checks, BTN bets $132, Hero folds

Final Pot: $264

BTN wins $262 ( won +$72 )
CO lost -$14
Hero lost -$58

I would've bet instead of checking on the flop personally. The point is that if the other player has Ax then they are already 90% likely to win, and yet a lot of people were saying the check was right and they should have called the $132 bet

Your bets give you information but all your bets are going to be bluffs/semi-bluffs or value bets.

True, but is that a bad thing?

Well say in that hand above you bet out.

If he calls or raises then you have information - you know he has an A (most of the time but w/e).
If he folds, you know you had the best hand.

Essentially all you do is keep better hands in and fold worse, i.e. the opposite of what you want to achieve.

Oh I see.
So what would the best play have been in the hand above?

Probably c/f or c/c but it depends on the player.

I'm rubbish at poker anyway.

I cbet here most of the time, even though it's effectively a bluff... if you're going to check-fold here, then you have to be check-shoving when you have AK or else you're setting yourself up to be bluffed a lot by announcing that you have a pocket pair, and I'd rather be leading out with AK. I suppose you could check-call, but that's a pretty expensive way to find out if your hand is good.

betting for info is a limit concept, that doesn't really transfer much to no-limit (because of the higher implied odds and larger range of hands)

basically in limit a bet for info is like this:

A player raises up front, he's fairly selective and pretty vanilla, and you re-raise with AQ. You're re-raising because the player is vanilla, and will most likely only cap hands like JJ+ AK. If he calls you can put him on a medium pair or a hand like AQ/AJ/KQs; if he caps you know you're going to need flop help.

Continue this concept on the flop: the flop is AT5 and your opponent bets into you. A bad player calls trying to get to showdown with his TPGK; a good player raises to see where his opponent is at, if you get re-raised you can safely muck this very good, but obviously second best hand. If he just calls, you can assume the ace is a scare card for him, or he's slowplaying a monster.

you make info bets on pre-flop and on the flop, before the betting limit doubles. These info raises save good players a lot of money. Bad players will never understand why you would raise someone to see if you should fold.

This is why smooth calling raises is so bad in limit poker (aside from family pots), you're left guessing.

i cbet ~75 here. no need to pot it