Exploitative strategy: play the opponent, not the chart

Exploitative poker is not chaos — it is conditional math. You still need a baseline (chart, solver, coach line). Then you apply if-then rules tied to observed behavior: if they fold to flop cbet 72%, cbet range; if they never fold turn, stop bluffing and value bet two streets thinner.

The four exploitable axes

• Preflop: too loose → 3-bet and iso wider; too tight → steal and overlimp traps.
• Flop: overfolds → small cbet range; overcalls → big value, no air.
• Turn: panic vs barrels → double barrel value and bluffs; sticky → one-and-done.
• River: overfolds → overbluff; underfolds → thin value, zero bluffs.

Sizing as the exploit multiplier

Most pool exploits are not range changes — they are size changes. A villain who folds 55% to 33% pot but 28% to 75% pot is telling you to size up for bluffs and size down for thin value. Charts rarely teach that because charts assume mixed strategies. Live poker assumes humans pick one size when nervous.

When exploitation becomes a leak

You are exploitative wrong when you adjust on one outlier hand («he called with jack-high, he is a station») or when you level yourself into bluffing a guy who only called because he misclicked. My rule: two confirmations in the same category before the exploit goes live. First time — note. Second time — trial exploit. Third time — default until counter-evidence.

I remember iso-raising a nit into a 4-bet pot

I remember widening iso to A9o against a VPIP 14% limper because «exploit loose limp.» He limp-raised to 12bb — first time in 40 minutes. My chart said fold; my note said punish. I stacked off A9 vs AQ and wrote «do not exploit sample size 1» in the ledger. Expensive lesson, $0.80 if I had folded.

Tag three villains on your next NUTS POKER session with one-line exploits and run them for 500 hands before adding a second adjustment. Exploitation scales when it is boring, measured, and written down — not when it is a hunch dressed as genius.