Exploiting Tight and Loose Opponents

The GTO vs. exploit guide covers when you should deviate from a balanced range. This one covers what actually changes when you do - the specific hands that get added or cut, and why, for the same spot against two opposite reads on the same opponent.

The read has to be real

Every exploit range on this site starts from a tag on the players still left to act: does this seat fold more than a balanced opponent would, call more, or play close enough to GTO that it isn't worth deviating for? That tag has to come from a real, sustained sample against that specific player - a stat from a HUD, a documented pattern, a big enough sample to trust - not a read off one hand. A wrong tag makes your range worse, not better, because you've given up GTO's protection for nothing.

Which is why the tag comes third, not first. If you play low stakes and want the method rather than the mechanism - how to build the baseline, spot the mistake, and only then adjust - push/fold strategy for low stakes covers it in order.

The same spot, two opposite reads

Below is the button's shove range at 10bb effective, three ways: the GTO baseline, then the solver's best response once the big blind is tagged as folding too much, then again once it's tagged as calling too much. Small blind plays GTO behind the button in all three.

Widening against a tight big blind

A big blind that folds too often makes bluffs profitable that weren't before - a hand doesn't need real equity if it wins the pot outright often enough. Hands like Q4s, J6s, J5s, 85s, K7o, and K6o go from a clean fold at GTO to a full shove here: almost none of their value comes from what happens if called, it comes from how often this specific opponent doesn't call at all. That's 10 percentage points wider than the GTO baseline.

Tightening against a loose big blind

A big blind that calls too often takes the opposite adjustment: the hands that were only shove-worthy because of fold equity get cut back, because the fold equity that justified them is exactly what a loose opponent denies. Hands like 86s and 65s go from an automatic shove at GTO to shoving only a fraction of the time here - they still have real equity if called, so they don't disappear the way the widen-side hands appear out of nothing, but there's meaningfully less reason to get it in against someone who keeps showing up to see a flop. That's 2 percentage points tighter than the GTO baseline.

The adjustment isn't symmetric

Widening against the tight big blind moves the range further than tightening against the loose one does, even though both profiles deviate the big blind's own range by the same 20% of its GTO mass in opposite directions. That's not a coincidence: a villain who folds more hands to hero outright creates brand-new bluff-shove candidates that had no case at all under GTO, while a villain who calls more mostly just erodes the fold equity behind hero's existing marginal shoves - those hands still have real equity when called, so they get shoved less often rather than cut to zero.