GTO vs exploitative play: what "optimal" actually means

Game-theory optimal (GTO) play is a strategy that cannot be exploited: if you played it perfectly, no opponent could adjust in any way that beats you in the long run. It's the poker equivalent of a fortress — not the most profitable strategy against any particular opponent, but the one with no doors left open.

Two things GTO is *not*:

• It's not "tight" or "aggressive". Solver output is often shockingly aggressive in some spots and shockingly cautious in others. GTO is a frequency, not a mood.
• It's not maximally profitable against weak players. Against someone who calls far too much, the GTO bluffing frequency lights money on fire. The fortress doesn't attack; it just never falls.

Exploitative play means deviating from the balanced baseline to attack a specific leak. Every leak has a mirror-image counter:

• They fold too much → bluff more and steal relentlessly — but tighten your thin value bets: the few hands that do call you are stronger.
• They call too much (the calling station) → stop bluffing entirely, value-bet hands you'd normally check.
• They never 3-bet without queens-plus → fold to their 3-bets without remorse, and open into them all day.
• They c-bet 100% of flops → check-raise and float them on every board that misses their range.

The catch: every exploit opens a door in your own fortress. If you never bluff the river against the station, an observant player can fold every bluff-catcher against you. Exploitative play is a bet that your opponent won't notice or won't adjust — against most opponents, an excellent bet.

The strongest practical framework is the one strong players actually use:

1. Learn the GTO baseline for the common spots — opening ranges, defence ranges, c-bet frequencies, jam charts. This is your default against unknowns and tough opponents.
2. Gather reads. Showdowns, sizing patterns, timing, frequencies — does this player fold their big blind too much? Do they only raise the river with the nuts?
3. Deviate in the direction of the leak, proportionally to your confidence. A small read justifies a small deviation; a glaring, repeated leak justifies a wholesale departure from theory.

Notice the order. Deviating *without* a baseline isn't exploitative play — it's just having your own leaks. You can't profitably leave a line you never knew.

You open the button, and the big blind — who has folded eight of the last ten big blinds — defends. The flop comes K72 rainbow, missing your A5s.

• The GTO line: c-bet this board at a high frequency for a small size with your whole range; ace-high is a fine candidate. Balanced, solid, unexploitable.
• The exploitative line: this player's fold-to-c-bet is enormous. Bet *every single hand* — the worst hand in your range prints money. Balance is irrelevant; they aren't paying attention to your frequencies, only their cards.

Now the same spot against a station who has called down three streets with third pair twice this hour: check your ace-high back, give up on bluffing, and when you do bet, bet big with hands that beat their calling range. Same cards, opposite lines — both correct, because correctness depends on who's across the table.

ShoveIQ grades your decisions against the verified baseline — that's the fortress, and you need it first. But the tables are also full of opponents with deliberate, recognisable leaks: stations, nits, maniacs, over-folders. The coaching that follows each hand ties your line to *who you faced*, because the lesson "this was a fine c-bet against a folder but a spew against a station" is the whole second half of poker.

Master the baseline, then earn your deviations. That's the entire game in one sentence.