Heads-up poker's most profitable edge in 2026 isn't a solver line or a GTO approximation—it's a HUD stat that tells you when a heads-up opponent is bluffing the river at a frequency that makes calling with third pair correct. The two-seat fish, the player who sits alone at a table waiting for action and then plays every hand because the blinds are cheap, leaves a river aggression pattern that a properly configured HUD exposes in about 40 hands. You don't need a read. You need the right columns turned on.

Why heads-up stats lie at six-max and tell the truth at two

Most players build their HUD for six-max and never rebuild it for heads-up. That's the mistake. At a full table, a 55% river bet frequency means nothing because the sample is spread across nine opponents and the pot odds are different on every street. Heads-up, the same 55% is a fingerprint.

The stats that matter in a two-seat game are narrower than the ones that matter at six-max:

  • VPIP and PFR gap. A fish with 78% VPIP and 12% PFR is limping or calling almost everything and raising only premiums. That gap is your entire preflop strategy.
  • River bet frequency. This is the stat the title is about. Sample size matters—under 30 river opportunities, ignore it.
  • Fold to river bet. The number that tells you whether a thin value bet prints.
  • WTSD (went to showdown). Fish who call down 45% of the time are not folding rivers. Adjust accordingly.

A HUD that shows you VPIP and PFR but hides river aggression is a HUD built for the wrong game.

The 40-hand threshold and why it's not arbitrary

Forty hands is roughly the point where a heads-up opponent's river bet frequency has a confidence interval tight enough to act on. At 20 river opportunities, a player betting 60% of rivers could be running hot on a 30% true frequency. At 40, that gap narrows enough that the number means something.

The practical version: set your HUD to color-code river bet frequency only after 35 river opportunities, and treat anything above 65% as a bluff-heavy profile. Against that profile, calling with any pair that beats a busted draw is correct at standard pot odds. Against a 40% river bettor, it isn't.

This is where most recreational HUD users go wrong. They see a red number and call. The number has to be paired with a sample size, or it's noise dressed as information.

What the fish actually does on the river

The two-seat fish has a predictable leak: they bet the river when checked to, regardless of their hand, because heads-up feels like a duel and checking feels like losing. Their river bet frequency runs high, their fold-to-river-bet runs low, and their showdown range is wider than their betting range should be.

That combination—high river aggression, low fold-to-bet, wide showdown—means two adjustments:

  1. Call wider on the river against their bets, because their betting range is wider than their value range.
  2. Value bet thinner on the river, because they call with hands that shouldn't call.

Neither adjustment requires a solver. Both require a HUD column you probably don't have turned on.

The part nobody puts in the affiliate review

HUDs are legal in most US-regulated poker rooms and prohibited in others. Check the terms of the specific site before you build a layout—getting your account frozen over a stat overlay is a bad way to learn the rules. And if you're playing heads-up for hours chasing a fish, set a session limit. The edge is real; the variance isn't.

The open question is whether the fish adapts. A two-seat regular who notices you calling their river bets three times in a row will either tighten up—killing your edge—or double down, which is the outcome you want. Which one happens depends on whether they're tracking their own stats. Most aren't. That's the whole game.