How to Bluff in Poker
A bluff works if your opponent folds often enough to cover what you are risking. Risking a half-pot bet to win the pot needs a fold 33.3% of the time; a pot-sized bet needs 50%. That is the whole test — the number comes from the price, not from how your opponent looks. Pick hands that can still improve, size so the required fold rate is one a real player hits, and stop bluffing people who do not fold.
The only question a bluff asks
A bluff is an offer: fold and I take the pot, call and I almost certainly lose what I bet. It shows a profit when the folds come often enough to pay for the calls. That break-even point depends on one thing — what you risk against what is already out there — and you can work it out before you act:
- •Required fold % = bet ÷ (pot + bet). Risk 50 into a pot of 100 and you need 50 ÷ 150 = 33.3%.
- •Nothing about the opponent enters that formula. It is the price of the bluff, and it is fixed the moment you choose a size.
- •Everything else — reads, image, history — is your estimate of whether this particular opponent clears that bar.
How often your bluff has to work, by size
Computed from the same function behind our fold equity calculator, for a pot of 100. The left column is what you bet; the middle is how often they must fold for the bluff to break even; the right is the mirror image — how much of their range they have to keep in to stop you profiting automatically:
- •25% pot — needs a fold 20.0% of the time. They must defend 80.0%.
- •33% pot — needs 24.8%. They must defend 75.2%.
- •50% pot — needs 33.3%. They must defend 66.7%.
- •66% pot — needs 39.8%. They must defend 60.2%.
- •75% pot — needs 42.9%. They must defend 57.1%.
- •Pot — needs 50.0%. They must defend 50.0%.
- •Two times pot — needs 66.7%. They must defend 33.3%.
The number on the right has a name
That defence figure is minimum defence frequency, and it is not a separate idea — it is the same arithmetic read from the other chair. If a half-pot bluff needs folds a third of the time, then folding more than two thirds hands you the pot for free. This is why small bluffs are hard to beat and large ones are easy to beat: a 25% stab only needs one fold in five, but a double-pot shove needs two in three.
- •Minimum defence frequency says how wide they must continue, not that they will. Most players fold far more than it allows, which is precisely what makes small bluffs profitable.
- •It assumes your bet could be a strong hand. Against someone who has seen you bluff four times in an orbit, the model stops describing them.
- •Work the number for any spot with the MDF calculator.
Which hands to bluff with
The best bluffing hands are not your worst hands. A hand with no future has only one way to win; a hand that can improve wins twice — when they fold, and when they call and you hit. Pick in roughly this order:
- •Draws you would be happy to see called — a flush draw bluffing the turn still wins about a fifth of the pots where it gets looked up. This is a semi-bluff, and it is the cheapest bluff in poker.
- •Hands that block their calls — holding one of the cards that makes their strongest hand means fewer of those hands exist. The ace of the flush suit is the classic example.
- •Hands with no showdown value — a busted draw that cannot win a checkdown costs nothing to turn into a bluff. A weak pair that might win sometimes does.
- •Not your absolute trash on the river with no blockers. There is usually a better candidate in your range, and using the worst one means bluffing too often.
Choosing a size
Bigger bluffs win more when they work and need to work more often. Smaller ones are the reverse. Two practical consequences fall out of the table above:
- •Bluff small when you need volume. A 33% stab only has to work a quarter of the time, and against a player who folds half their range it is close to free money.
- •Bluff big when you are representing something specific and the story only makes sense at that size — but understand you have moved the bar from 25% to 50% or beyond.
- •Use the same sizes you use for value. A size you only ever show up with when bluffing is a tell, and no amount of fold equity survives being read.
When not to bluff
The arithmetic above assumes a fold is possible. Several common situations break that assumption, and no bet size rescues them:
- •Against a player who does not fold. The formula needs a fold rate; against a calling station the rate is near zero and every bluff is a donation.
- •Into several opponents. Each extra player has to fold, so the chance of getting through collapses. A bluff that needs 33% from one player needs something closer to 58% from each of two.
- •When they are priced in. Someone drawing to a hand that beats you with correct odds to call is not folding, however the bet is framed.
- •On a board that misses your story. If nothing you would have bet earlier connects with this card, the bet says so.
Related tools
Frequently asked
How often does a bluff need to work?
It depends entirely on size. Bet ÷ (pot + bet) gives the break-even fold rate: 20% for a quarter-pot bet, 33.3% for a half-pot, 50% for a pot-sized one and 66.7% at twice pot. Below that rate the bluff loses money, above it the bluff makes money, and nothing about the opponent changes the figure — only whether they clear it.
What is the difference between a bluff and a semi-bluff?
A pure bluff wins only when they fold. A semi-bluff also wins when they call and you improve — betting a flush draw, for instance. That second way of winning means a semi-bluff needs fewer folds to break even, which is why draws make the best bluffing hands and busted hands with no blockers make the worst.
Is it better to bluff big or small?
Small, more often than players expect. A 33% pot bet needs a fold less than a quarter of the time; a pot-sized bet needs half. Most opponents fold well above the smaller bar and well below the larger one. Size big when the hand you are representing would be bet big, not to apply pressure for its own sake.
What is minimum defence frequency?
The share of their range an opponent has to continue with so that bluffing them automatically is not profitable. It is the same arithmetic as the break-even fold rate, seen from the defending side: a half-pot bet needs folds 33.3% of the time, so the defender must continue 66.7%. Real players defend far less than this, which is where most bluffing profit comes from.
Should I bluff more against good players or bad players?
Against good players, in general — they are capable of folding. Weak players call too much, and the break-even calculation needs folds to exist at all. The usual mistake is bluffing the person who never folds because they are 'bad', which has it exactly backwards.
Can you bluff multiple opponents?
Rarely well. Every player has to fold, so the required rate compounds: a bluff needing one opponent to fold 33% of the time needs roughly 58% from each of two for the same break-even. Multiway bluffs want much stronger reasons than heads-up ones — a real draw, or a board only you can credibly have.
Terms used in this article
Keep reading
New to poker? Take the free 15-minute crash course →