Chapter 7: Exploit

The Double-Layer Structure of GTO and Exploit

Reverse the order, and it always falls apart

Layer 1: a GTO-leaning baseline

• The default when you do not know who your opponent is
• The goal is not to be exploited (not to lose)
• Without it, you have no standard to measure deviations against

Layer 2: Exploit adjustments

• Move the baseline only after observing your opponent’s deviation
• The goal is to maximize winnings
• Every adjustment you make makes you that much easier to exploit too

Baseline → observation → adjustment. Never break this order.

Why the order matters

If you adjust because you feel “this person looks likely to fold” without a baseline, that is not an adjustment; it is just a mood.

Only people who have a baseline can measure deviations.

On the other hand, never moving a step from the baseline also costs you. The population has clear biases, and not taking them is the same as throwing away profit.

Player Types

VPIP / PFR are estimates. Look at concrete actions rather than labels

Type	VPIP / PFR (est.)	The biggest trait
Nit	12 / 10	Plays only strong hands. Does not bluff
TAG	22 / 18	A standard winning player. Few big holes
LAG	28 / 23	Plays wide and aggressively. Bluffs often
Calling Station	45 / 8	Does not fold. Bluffs do not get through
Maniac	50 / 35	Attacks indiscriminately. Very large variance
Passive Fish	40 / 6	Plays but does not raise. A bet means strength

A label is a label for a hypothesis, not evidence. Use it as a set with the observation standards in the section below, “How Many Times Before You Can Strengthen a Hypothesis.”

Direction of Adjustment by Type

Which way, and how far, to shift

Calling Station	Bluff less, and go for thin value more thickly
Nit	Bluff more, and fold hard against strong hands
Overbluffer	Bluff-catch more. Pick off with even weak pairs
Underbluffer	Stop every marginal call. You can fold
Passive (bets rarely)	Their bets are strong. Be even more careful against raises
Aggressive (high frequency)	Check-raise and bluff-catch more
Maniac	Tighten your range, wait, and hit back with strong hands

Adjust one direction at a time. Simultaneous adjustments like “bluff less while also calling less” usually contradict each other.

How Many Times Before You Can Strengthen a Hypothesis

If you do not decide this in advance, you change your strategy on assumptions

Times observed	How to treat it	Action
1 time	The seed of a hypothesis. Likely chance	Change nothing
2 – 3 times	A candidate tendency. Only in the same spot	Adjust a little
5 times or more	You can treat it as a tendency	Adjust clearly
10 times or more	It becomes grounds for changing strategy	Move the baseline

The exception: the rarer the spot, the more the information is worth

• A river check-raise is an action with a frequency of only 2 – 5% (est.) to begin with
• So seeing it just two times already gives you information: “this person is the type who does it”
• Conversely, a high-frequency action like a C-bet cannot be called a tendency until you see it 10 times

The standard for judging: how rare the action is in the population average. The rarer it is, the fewer observations it takes to mean something.

Typical Leaks of the Population

Usable as starting values when you do not know who your opponent is

Not enough river bluffs	More marked the bigger the bet. You can fold
BB defense is too narrow	It is worth raising your open frequency
Folding too much to 3bets	3bet bluffs become more valuable
Too many C-bets	Check-raises and floats work
Giving up too much on the turn	You can widen your flop calls
Cannot fold weak pairs	You can go for thin value more thickly

These are tendencies; they may not apply to the opponent in front of you. Overwrite them with observation.

What You May Read from Bet Sizes

Misread it, and it gets turned against you

What you may read

• Whether the person changes size by situation or keeps it fixed
• Whether they tend to bet big only with strong hands
• What they showed in the same spot in the past

What you must not read

• Concluding hand strength from a single size
• Assuming the relation between size and strength is the same for everyone

Timing: online

• An instant call = often a limited hand (neither strong nor weak)
• A raise after a long think = often leans strong
• An instant bet = decided in advance = betting with the range

Timing: live

• The size of a movement is not very reliable (there is acting)
• Preparatory movements before a bet and how chips are held are more stable

A tell is not evidence; it is just a piece of information. Do not use it to overturn the conclusion of your range analysis, but as material to tip a borderline decision to one side.

The Risks of Exploiting

Every adjustment has a cost

It gets turned against you

If you bluff less, your opponent can fold safely. Your adjustments get read by your opponent too.

The sample is too small

Changing your strategy after three observations means distorting your strategy to fit chance.

Forgetting to change back

You keep the adjustment even after the opponent changes. When the table changes, go back to the baseline.

As an estimate, it is safe to adjust a lot when the opponent is clearly weak (Station, Nit), and not to move from the baseline when the opponent seems ordinary. Fine adjustments against intermediate players cause the most accidents.