The 2+2 Forum Archives: Varying flop bet sizes (long)

This is something I've been thinking about quite a bit, but have not had much chance to experiment with.

Most people in this forum seem to be advocates of always betting the full size of the pot, regardless of their hand strength, board texture, or number of opponents. This is the strategy which is recommended in SuperSystem. However, other noted poker pros have expressed other opinions. Ciaffone/Reuben's book adviced betting half pot on drawless boards, and closer to full pot on scarier boards, regardless of the cards they held. Ray Zee has previously written on 2+2 that he prefers to underbet the pot in most situations, as it allows him to see very cheap cards sometimes while drawing, he can sometimes pick up a pot with minimal risk, and if he gets raised he can make a read on his opponent to decide what further action to take. When watching High Stakes Poker on GSN, I've observed that most of the players do vary their bet sizes based upon board texture and number of opponents.

In these spots we have competing desires. We want to play big pots when we have big hands, and invest as little money as possible with weak hands. However, our attempts to realize this goal will ultimately backfire if our betting gives away too much information about our hand. This leads to our desire for deception and disguise. So as extreme cases, we may have one player who is too straightforward and always bets big when he has a hand and makes small bets as bluffs or blocking bets, while the player at the other extreme always bets exactly the same no matter what he has.

Neither of these players will succeed in their efforts to win big pots with their good hands and lose the least with their losers. The first player will not get action on his big hands as they are not disguised, and he will likely have a number of pots stolen from him when he shows weakness. The second player is always playing pots of the same size regardless of hand strength, which in reality is actually letting the opposition dictate the pot sizes as they are free to raise whenever they want to. The second player has very little power to utilize any pot control and will get stuck in a lot of bad spots.

So, I've been thinking about this and trying to come up with an optimal betting system. In the past I feel I've gone too far overboard in my efforts to constantly disguise my hand for fear that variations in my betting will leak too much information. After thinking about this for a while, I've come to the conclusion that as long as my betting patterns are dictated by more than simply the strength of my hand, then my opponents will not be able to read my betting accurately enough to outplay me. They are just as likely to misread my hand as they are too make an accurate read.

Some examples of what I'm talking about...

I open from the CO for 3 bb. All three players left in the hand call me. The flop is AA5 rainbow. The blinds check. How much should I bet here? If I bet half pot, then I might get called by a random pocket pair or a five, or even raised by someone who may have folded to a stronger bet. However, I am risking much less money and therefore my bluff does not have to work as often as a bigger bluff. So I could bet half pot here with complete air, laying myself good odds to pick up the pot. Or I could bet half pot with a very strong hand that wants action, and hope my feigned weakness entices loose action. However, if I have a hand like A7s or TT, I may not want that loose action as I won't know if my hand is good. So with marginal hands perhaps I bet closer to the full size of the pot, if I know that I can shut down completely if I get called. And I can also make those big bets with very big hands (55, A5s, perhaps AK) which want to build a big pot.

As another example, I raise on the button and the BB calls. The flop is Q86 with two spades. BB checks, and I think most of the time he will either call or fold when I bet. Suppose he will call a big (pot size) bet with top pair or better, or a flush draw, but if I bet half pot he will call with some pairs worse than top pair. In this spot I might bet big with sets, because sets want to play big pots. I bet big with the nut flush draw, because if I hit my flush then I want to win a big pot, and because I'd prefer my opponent fold some weak pairs which are currently beating me (but if he does call then I'm still 50/50 to win the pot, so the extra money that I've risked does not actually cost me anything... it is free pot building for me). If I have AQ I might just bet half pot, so that I can get action from hands which are drawing quite slim against me (99, etc), and to help control the pot size. With air I can sometimes bet big and sometimes bet small.

If my opponent is unwilling to try bluffing me off my TPTK type of hands then he does not have any good defense against my strategy (remember I specified that this opponent was unlikely to checkraise me), since they cannot determine whether they are ahead no matter what hand they have (unless it's top set) or how much I bet. However, I am manipulating the action so that I am getting the right amount of value for whatever hand I hold.

On the other hand, if I think my opponent is likely to checkraise my small bets (sometimes with marginal hands, draws, bluffs, etc) but won't checkraise my big bets very often, then I can bet small with my sets to induce action (and with my big draws if he'll fold somewhat often to a push). With TPTK hands I bet bigger, as this has now become a better method for controlling the pot size and for avoiding tough decisions. With air I'm still free to choose either bet size, depending on which option seems to hold the better risk:reward.

So, over the course of many hands, I will not be giving away much information with the amounts that I bet, as I may be betting my marginal hands very hard in some spots, and betting them softly in others, etc. I can constantly adapt to my opponents, and in order for them to adapt to me, they have to figure out all of the factors that go into my decision making process, and then they must accurately emulate the way that I consider all of those factors. But since most of my hands will not be shown down, they will likely either draw incorrect conclusions about my betting patterns (which I can exploit, as I can also recall which hands I've shown down), or my betting may simply appear to be somewhat random.

I've chosen only to discuss flop betting as this post is already too long, but the same principles can be applied to the turn and river as well. For example, I could employ a strategy of grossly underbetting the flop, but on the turn I either check or bet full pot. If people call too liberally on the flop, then they will get faced with a lot of turn bets that they cannot call (which could be bluffs). If they fold a lot of flops, then I'm receiving very good odds to attempt my bluffs and I can give up the times that I do get called.

In the end, the goal of my betting is always to manipulate the range of hands that my opponent is playing against me so that I am always in a position where I have an easy decision. When I have a marginal hand I do not want to end up in a spot where I think I'm behind about 70% of the time, especially if I'm facing a bet or a raise. I can either underplay my hand a bit so that I get action from more hands I beat, or overplay my hand to the point where I can safely fold to further action. If I have a hand with no showdown value, I want to either be certain that my opponent has a big enough hand that further bluffing will be pointless, or I want his range to be wide enough that I can still show a profit by firing again. If I'm drawing, I want to avoid spots where I get raised and do not have odds to chase, and I don't have enough stack left to generate folding equity with a push. (This means I have to consider what the turn betting will be like before I choose my flop bet if I'm OOP.) And drawing hands also fall into the aforementioned "no showdown value" category in some spots.

Thoughts?

The 2+2 Forums