The River: Value and Bluff Catching

The one street where nothing is coming

Every number in this camp so far has had a "yet" attached to it. On the river there is no yet. Your hand is finished, his hand is finished, and the only question left is whether the money should move.

That is why this is the one street where our rule layer will tell you how much nut potential a multiway pot demands. River K♥8♥3♥2♠7♦, against 3 players, opponents restricted to hands that would actually put money in, 40000 deals per row:

• A♥Q♥2♣5♦ — the nuts: wins 100%.
• Q♥T♥2♣5♦ — ranked 11: 83.3%.
• 6♥4♥2♣5♦ — ranked 41: 46.3%.

Our line is 60%. Above it a hand can contest a multiway pot; below it the hand is paying rather than being paid. The first two are above, the small flush is not.

On the river in a multiway pot, ask for a share of pots won, not a place in a ranking. Ranked 11 and ranked 41 are thirty places apart and worth 83.3% against 46.3%; the ranking is evenly spaced and the value is not.

River only, and multiway only. Heads up this quantity has nothing to say — every one of those hands plays normally against a single opponent.

The same threshold on an earlier street is simply wrong, which is why our rule layer refuses to apply it there. On a flop or turn there are cards to come and everybody's share is lower for reasons that have nothing to do with hand quality — a genuine monster measured 46.3%-style on the flop would be told to fold. We calibrated 60% on complete five-card boards and we use it nowhere else.

Bluff catching is a price, not a read

He bets the river. You have a hand that beats a bluff and loses to everything else. This is the most common decision in poker and the most common way to answer it is to guess what he has. Do the cheap thing first.

What the price asks. Half pot needs 25.0%, three-quarters needs 30.0%, a full pot needs 33.3%. That is how often your bluff catch has to be good — not how often he is bluffing, how often you win.

What a pure bluff needs from you. A bet that can never win at showdown risks its size to take the pot. Against a quarter-pot bet it needs you to fold 20.0% of the time, so you can keep 80.0% of your hands. Against a full pot it needs 50.0%.

Those two paragraphs answer different questions and together they bracket the spot: the first says how good your hand must be, the second says how much you are allowed to throw away before folding is the leak.

Answer the price first and take the fold it hands you. Inside 3 points of the line, treat the decision as genuinely close instead of manufacturing a read.

River, heads up, facing one bet with a hand that has showdown value and no way to improve — there is nothing left to improve into.

The second calculation is the one that gets misused. It describes a bet with no equity at all, and almost nothing on a river is quite that hopeless in this game — four cards on a coordinated board usually retain something. Treating 50.0% as a target to hit rather than as an outer bound turns a piece of arithmetic into a strategy nobody checked.

Both columns are arithmetic done on this page — the break-even fold rate for a bet with zero equity, and one minus it. No solver was involved, and neither number describes what anyone should do with a whole range.

Blockers close the hand, they do not open it

Same river K♥8♥3♥2♠7♦. You have answered the price and you are still on the fence. Now — and only now — the cards you are not using get a vote.

The best possible hand here is one named pair of hearts. A♥Q♥2♣5♦ holds it, so it removes 100% of the ways an opponent could have it. 6♥4♥2♣5♦ removes none: its removal is 19.7 points in the other direction, because taking cards out of the deck concentrates what is left.

And here is the part that matters for the decision. Our rule layer still declines to call the first hand a blocker worth acting on, because that best hand was rare to begin with and our floor is 1 point of absolute removal on top of 30% relative. Blocking all of something rare is still blocking almost nothing.

Use blockers to break a tie, in this order: price, then how the hand does against what is betting, then removal. Never start with removal — it is the smallest of the three effects and the easiest to talk yourself into.

River only. On earlier streets the best possible hand is still moving, so a card that blocks it now may block nothing by showdown.

Removal cannot rescue a hand that fails the second question. The ranked-41 flush wins 46.3% against three players and the ranked-11 flush wins 83.3%, and no amount of blocking closes a gap that size. Blockers choose between hands that were already close; they do not promote one that was not.

Removal figures are complete enumerations over every opponent holding on this board (the postflop rule spec), and the two thresholds are the ones this camp has used since the blocker lesson.

What to take away

• Price first, always. Facing one bet you never need more than 33.3%, and inside 3 points of the line nobody knows.
• Multiway on the river has a line, and it is the only street that does. Ours is 60%: ranked eleven clears it at 83.3%, the small flush does not at 46.3%.
• Know what a pure bluff needs. Against a quarter pot you can keep 80.0% of your hands — use it to catch yourself over-folding, not as a target.
• Blockers come last. Removing 100% of a rare hand is not a reason to bet, and it never rescues a hand that lost question two.

Every figure here is an enumeration or a seeded sample from our own engine, and every line is one of the rules of this course, not an equilibrium solution. Next, and last in this camp: the pots where all of this happens with almost no room.