How to exploit calling stations: thinner value, fewer bluffs, bigger bets

A calling station is the player who doesn’t fold. He calls the flop with anything, calls the turn with a draw or a weak pair, and looks you up on the river with ace-high. Everybody knows you should “value bet him”. Fewer people know how much the maths changes, or in which direction.

In this article I use one solved flop in the Poker Academy postflop solver, once at equilibrium and once with the opponent locked as a calling station. Then I put the numbers next to two formulas: what a bluff needs, and what a value bet earns. Everything that doesn’t come from the solver or from arithmetic is marked as my heuristic.

What a calling station is, and what he isn’t

The leak is one thing: he folds less often than he should. He is not aggressive with it either, since he rarely bets or raises without a strong hand and doesn’t bluff much. That’s why he’s also called passive-loose.

You’ll usually spot him by a big gap between how often he puts money in preflop (VPIP) and how often he raises (PFR), and by how often he goes to showdown. I’m not giving threshold numbers here. They depend on the game (6-max or full ring, cash or tournament), the stakes and the sample size, and I don’t have a source I’d trust for one set of cut-offs. Treat any HUD rule of thumb as a heuristic and check it against the hands you’ve actually seen.

This article used to say that against a station you can keep barreling the turn, because “most of his strong hands would have bet the flop”. That’s backwards. A passive player calls with his strong hands instead of betting them, so his calling range is wide and uncapped, and more bluffs on the turn is the wrong adjustment.

Preflop: what calling too much looks like

Start with the baseline. This is the cash simulation I use for the whole article: Normal NL200 GG, 6-max, 100bb, no ante. HJ opens to 2.5bb and CO has to decide.

The solver flats 23 of 1326 combos here and 3-bets 124. A calling station flats far more. What follows is my heuristic, not a solver result: the extra flats are mostly the hands a regular folds (weak offsuit broadways, weak aces, small suited hands), and because he 3-bets less, his flatting range also keeps strong hands a regular would 3-bet. So open your value hands as usual, don’t widen your opening range with hands that only win by folding him out later, and expect a lot of single-raised pots against a wide, uncapped range.

Rule one: a bluff needs folds

A pure bluff (a hand with no equity that gives up when it doesn’t get a fold) risks the bet to win the pot. It breaks even when the opponent folds this often:

That number goes up fast with size. Here it is for the four bet sizes on the flop I use below, where the pot is 6.5bb (the app labels 1.6bb as 25%; exactly it is 24.6%):

Against a normal opponent you compare that with how often he folds. Against a station you compare it with how often he folds after he stops folding, which is a thing you can measure.

The spot: a calling station locked in the solver

Same simulation. HJ opens 2.5bb, CO calls, button and blinds fold. The pot is 2.5 + 2.5 + 0.5 + 1 = 6.5bb, both players have 97.5bb behind. The flop is 8♣︎8♥︎6♦︎ and HJ checks to CO, who is in position.

The Poker Academy postflop solver has node locking. You fix how one player plays at chosen nodes and the solver re-solves the rest of the tree against that fixed strategy. Instead of locking by hand, you can pick a profile. The Calling Station profile does one thing: it takes the hands a player folds that are closest to continuing (by the solver’s own EV) and turns most of them into calls. It works on the player’s first two decisions of a line and comes in three strengths:

M sits in between at 23.0%. In the L version, the 20% of the folding range that is closest to a call continues 95% of the time, the next 10% calls 85% of the time and the next 5% calls 33% of the time. The rest of that decision stays at the GTO strategy. I locked HJ as Calling Station L.

Here is HJ facing CO’s 150% pot bet, 9.7bb into 6.5bb, first at GTO and then as the station:

The station folds 21.3 percentage points less. A 150% bluff needs 59.9% folds. Compare:

At equilibrium that bluff clears its break-even by 13.9 points. Against the station it’s 7.3 points short. In big blinds, for a bluff with zero equity that gives up to any call or raise:

The same bluff against the GTO HJ earns 0.738 × 6.5 − 0.262 × 9.7 = +2.26bb. A profitable bluff became a losing one, with nothing changing except how often the opponent folds.

Two things in that chart are worth more than any single number in it:

• The bigger the bet, the less the bluff survives. The station takes away roughly the same share of folds at every size, but big bets needed more folds in the first place. The 25% bluff still clears its line (30.1% vs 19.8%), the 75% bluff barely does (46.9% vs 43.0%), the 150% bluff doesn’t.
• This is one flop and one lock. Whether a 25% stab still works against your real opponent depends on his real folding, which you only know from hands. The direction is what carries over: fewer bluffs, and the bluffs you keep should be small.

For more on how the lock works and the minimum defence frequency behind these numbers, see GTO vs exploitative poker.

Rule two: value bets get paid by worse

The other side of the same leak. Take the simplest case: river, you are last to act, and he only calls or folds (a station rarely raises). When you bet B and get called, you win B more than checking when you’re ahead and lose B more when you’re behind. If w is how often you win against the hands that call:

So a value bet only needs to win more than half the time against the calling range, not against his whole range. A constructed example, pure maths with assumed win rates: pot 20bb on the river, you bet 15bb (75% pot). Against a regular’s tight calling range you win 55% of the time when called. Against a station, the same hand wins 65% when called, because his calling range is full of worse hands.

Three adjustments come straight out of that formula:

• Thinner value. Hands that win less than 50% against a regular’s calls can win more than 50% against a station’s calls. Those hands change from checks to bets. How thin depends on his actual calling range, which is a read, not a solver number.
• Bigger value. The extra is B(2w − 1), so it grows with B. If he calls a pot-size bet about as often as a half-pot bet, the bigger bet earns more: in the example 6.0bb instead of 4.5bb per call. A player who calls anything is the one opponent where “bet bigger with value” doesn’t cost you calls.
• No sizes built to get folds. A small bet exists partly to fold out weak hands cheaply. He doesn’t fold them.

What the solver does against him

Here is CO’s best response one node earlier, after HJ checks, with HJ still locked as Calling Station L:

At equilibrium CO bets 25% pot with 65.0% of its range and checks 32.6%. Against the station the small bet almost disappears and 71.1% of the range goes all-in for 97.5bb into 6.5bb. The app’s lock-impact line puts CO 3.31bb ahead of its GTO EV at this node and HJ 3.47bb behind.

This is what the station does when the jam comes:

HJ needs 48.4% equity to call. The GTO HJ calls 10.6% of its range; the station calls 36.3%. Those extra calls are hands the solver folds because they don’t have the price, and the lock-impact line shows what they cost HJ at this node: −5.68bb.

Please don’t take the 71% jam to the table. Calling Station L is the extreme profile, and the solver answers an extreme model with an extreme strategy. My guess, not a solver result: a real station still folds his worst hands to a shove 15 times the pot (97.5 / 6.5 = 15.0), and against him most of those jams just get folds from hands you wanted to get paid by. Only the direction carries over: value gets bigger, the small bets that only work through folds shrink, and pure bluffs lose their reason to exist.

Street by street, as heuristics

None of this is solver output for your spot. It’s how I apply the two rules above against a player I’ve seen call too much.

• Flop. Bet your value and strong draws. Check more of the hands that would be c-bet bluffs against a regular, especially on boards that hit his wide flatting range. If you do stab with air, use a small size, because big bluffs need folds he won’t give.
• Turn. Keep betting for value, and size up. Don’t fire a second barrel as a bluff just because he’s weak on average: his calling range still contains all his strong hands, because he doesn’t raise them.
• River. Bet thinner for value and bigger with your strong hands. Give up more of your missed draws. The river is where B(2w − 1) matters most, because there is no more equity to realise.
• When he bets or raises. A passive player’s aggression is information. Fold more of your marginal bluff-catchers than you would against a balanced player. That’s an exploit too, and it’s the one people forget.

Out of position against a station

Out of position you can’t rely on him to bet for you, so you build the pot yourself: lead or check-raise your strong hands instead of checking and hoping. With medium hands you can check and call more lightly only when he actually bets, which a passive player does less often. Heuristic, again: the solver result above is for the in-position player, and I don’t have a locked solve for the out-of-position side of this spot.

Cash games and tournaments

The whole example is a 100bb cash game without antes. In cash games stations are common, sessions are long and you get enough hands to trust the read. In tournaments the same two formulas hold in chips, but chips aren’t money. Near the bubble and at final tables ICM makes calling off a stack more expensive for everybody, so even a station calls off less, and your big value bets and shoves lose some of their edge. Lock against what a player does at that stage, not at a cash table.

Wrapping up

• A calling station folds too little, and that is the whole leak.
• Bluffs need bet / (pot + bet) folds. The locked station on 8♣8♥6♦ folds 52.6% to a 150% bet that needs 59.9%.
• Value bets earn B(2w − 1) when called. Against a wide calling range w goes up, so bet thinner and bigger.
• Check your read in the solver: lock the profile, compare the node with GTO and look at the EV before you change your game.

Cheers! 🙂