Bluffing is an essential part of poker, but it's one that a lot of people don't understand on a deep level. We categorize bluffs two ways: the first is a semibluff, the second a true bluff.
Semibluffs are bets that you make when you don't have the best hand but the next card could make you the best hand. A flush draw would be a good example of a semibluff. You don't have a hand yet, but you have a really good shot at catching the flush. A pure bluff is a bet you make when you have no chance of winning the hand and you expect that the other player will fold a certain percentage of the time, so you are making a bet in order to make them fold.
Remember, there are three ways to win poker hands. The first is to win the hand at showdown. The second is to get it all in before showdown and win the hand by showdown. And the third is to win the hand by stealing the pot. Bluffing accounts for the second two ways to win the pot, so it's essentially more important than having good hands. If you put in a strong semibluff and get it all in, not only can you win the pot by having your opponent fold on the spot, but you can also catch your hand and win by showdown.
Now let's break down each type of bluff in a little more detail. The semibluff gets its value from the combined equity of your opponents folding and you making the best hand. A bluff gets all of its equity from the percentage of times your opponent folds. This means that when you bet, whether a bluff or a semibluff, you have to be bluffing into an opponent that will fold often enough that will make your bet profitable.
If the pot is $100 and you bet $50, then I have to win that pot at least 33% of the time. I'm putting in $50 and the pot will be $150 when I take it back. If I'm winning this more than 33% of the time, I can bet there as a bluff. With a semibluff, the equity that you have for winning the pot with your draw is added to the percentage of time the opponent will fold. If I expect my opponent to fold 20% of the time but I also have a 20% chance to win the pot, then I can bet $50 into a $100 pot, even though the opponent is folding less than 33% of the time, because I have some equity with my draw. In fact, I only need my opponent to fold about 15% of the time to make this bet profitable, because 20% of the time I'll make my hand anyway and win. Because we are betting less than the size of the pot, we aren't risking so much to convince him to fold.
Let's take a look at a hand that showcases the value of the semibluff. I have on an board. That gives me a total of 12 outs, so I'm going to win the pot outright 48% of the time. I very rarely need my opponents to fold to make this a profitable bet. Nonetheless, because my equity is less than 50%, I do prefer to have them fold rather than to see a turn and a river card. In this case, I might bet something like 75% of the pot to encourage my opponents to fold but also to build the pot in case I do turn a straight or a flush.
As your equity decreases with a semibluff, for example, if I have a straight draw with just eight outs or two overcards with just six outs, the chance that the opponent folds needs to increase, or I have to be able to bet less without them folding less often. If I have two overcards for a 24% chance to win the pot, and I bet half pot, I need them to fold an additional 9% of the time for this to be profitable. Since our opponents always fold 10% of the time, this is a good bet. If I was to bet full pot, I would need them to fold 26% of the time, which is likely not the case, but might be profitable, depending on whether betting more induces the opponent to fold more hands. One read that can help you determine this is whether or not an opponent folds significantly more hands to full pot bets than half pot bets.
For the more advanced players out there, the way that we calculate the profitability of a bet is by taking our opponents' range and dividing it up into the percent of the range he calls with and the percent of the range he folds with. This approach utilizes combinatorics and is best performed away from the table due to the complicated calculations, but doing these calculations will help you intuit in the future what percent of his range he is folding.
The 10% Rule
Any time my opponents bet, I always assume that they're bluffing at least 10% of the time. Likewise, when I bluff, I assume my opponents will fold 10% of the time. Against aggressive opponents, I might bump this number to 15% or 20%. This is simply because you can't ever be certain what your opponents have. To say that you're 100% confident in what your opponents have is usually a mistake. The 10% rule provides a reasonable range to help you see the errors of that mentality and truly consider the equity value of your hand. If you've got a profitable bet if the opponent folds 10% of the time, you're going to want to make that bet. This rule definitely matters for semibluffs. If I've got a flush draw and a backdoor straight draw, I'm 41% to win the pot. If I bet the size of the pot, I need to win the pot 50% of the time. In this case, I'll always make that bet.
Poker myths about bluffing
Poker is not all about big bets and bluffs. There are a lot of TV programs out there that show all the big bluffs that happen at a poker table. I often only bluff once or twice an hour at a poker game, because the situation usually doesn't dictate the spot to be a good bluff. I very rarely make pure bluffs without good reads. A lot of players make bluffs a central part of their game and they end up going broke because eventually players will start calling you. The secret to make your bluffs profitable is to make semibluffs more often but to make your pure bluffs very, very rare. The only times when a pure bluff enters into my consideration is, say, on the river when I have no way to win the pot and I know this but I believe a scare card has come on the turn or river and I believe my opponent will fold a high percentage of his hands.
The continuation bet bluff
A special case of bluff is the continuation bet bluff. A continuation bet (or C-Bet) happens when you raise before the flop and then bet on the flop. If you flop a hand, then you're betting for value. But, if you miss the flop, then you're continuation bet bluffing. This is a special kind of bluff that I make very often. Because you've raised preflop, you are showing strength going into the hand. Once they've seen the flop, the opponent will often decide not to continue. Only about 35% of hands actually make a pair or better on the flop, which means about 65% of the time your opponent will fold to your flop bet. This allows you to make a generous profit (remember, with a half-pot bet, they only need to fold 35% of the time). Some players will often call flop bets to see a turn. These players can make your bluffs unprofitable.
What will really hurt your percentages in a continuation bet bluffing spot is having two players in the pot with you. When more than one person calls before the flop, there's a higher chance that someone caught a hand. If I'm betting half pot and need to win 33% of the time, but I've got two players who both fold 60% of the time, then my opponents both fold only ~35% of the time or so. That's pretty break-even. When you're betting into two or more players, a continuation bet can be unprofitable. I rarely C-Bet as a bluff into three or more players.