Continuation bet always?

The other day I played some limit hold'em. But I found I didn't enjoy it very much for some reason. Time for a change I thought.

So probably against my better judgment I returned to $5 NL, where I'd gotten kicked around before, where I vowed I wouldn't play again. But that was four months ago.

Not a lot to say for far about it. I mined a couple of sets and I'm actually up 1 buyin since my brief return. Nothing special but it beats being down yet another buyin.

One thing that I'm thinking about is continuation betting the flop heads up. I feel that I'm sometimes just throwing money away continuation betting on flops I miss if the other player, no matter how bad, has probably caught some piece of the flop, or knows that I probably missed.

So I've decided to mix things up a bit and randomize my continuation betting. This is just for heads up flops. In a 3 or more way flop it is much simpler and I would generally only cbet with a strong made hand.

For heads up I'm trying the following guideline for continuation betting the flop.

1. continuation bet with a hand 75% of the time. check 25% of the time
2. continuation bet without a hand 50% of the time. check 50% of the time

I need to define with a hand. I have a hand if I flop an overpair or top pair. Also a pocket pair with just 1 overcard on the board, or flopping middle pair [like AQ on a KQ3 flop].

Otherwise I don't have a hand. That would be Ax hands that missed the flop or hands like 99-TT with two or more overcards on the flop.

Harrrington said to use your watch to randomize your play, but online I took the opportunity to play with Python and wrote a small program to do it.

import random
import sys

print "Press enter for a number. Press 'x' or 'q' to quit."

while True:
 instr = sys.stdin.readline()
 firstchar = instr[0]
 if firstchar == 'x' or firstchar == 'q':
     break;

 randnum = random.random()
 print "The next number is ", randnum

I think that randomizing the flop continuation bet has some benefits. If you cbet 100% of the time then this is the case.

1. I cbet with a hand.
2. I cbet without a hand.

Since most flops miss most players cbetting 100% of the time can be exploitable both by skilled observant players who can read preflop ranges against the texture of the flop. As well bad players will call with weak made hands that outflopped you and you are just paying them off.

By randomizing the flop cbet you present your opponents a more difficult proposition.

1. I cbet with a hand.
2. I check with a hand.
3. I cbet without a hand.
4. I check without a hand.

With extra deception you are harder to read and exploit.

There are some consequences of course. For example by giving a free card on the flop you could lose on the turn. That's true but even if someone has as many as 15 outs then you are still good on the turn around 70% of the time.

If opponent has 2-6 outs then sometimes you'll lose to the free card when you could have won on the flop, but around 90% of the time you'll still be ahead on the turn, now with only 1 card to come.

Also with this approach it would be possible to preflop raise with AQ-AK, then just check fold the flop if you miss. These are tradeoffs you have to make.

I know this is $5 NL so many might think randomizing is silly. Perhaps it is, but at some point you have to start to think about it so this would be as good a place to start as any.

One thing I like about it is that it should create more involved postflop play. Hands may not be as well defined and there should be more turn and river play. That at least can make for more interesting and challenging situations. That would make it more enjoyable to play and I am just a recreational player.

Advanced players seem to feel that most profit and loss in poker occurs postflop so creating more postflop situations should be good for my game.