How it works

Real hands. GTO baseline. Ranked exploits.

SharkIQ measures actual decision frequencies from real hands in specific online rooms, compares each spot class to GTO baseline, and ranks the gaps by EV. Here's the short version, then you can try one spot.

  1. 1

    Real hands, not theory

    We aggregate decision frequencies and sizings from real hands in a specific online room — Suprema NL10–NL50 evening Brazil first. Every spot class — preflop opens, flop c-bets, river over-bets — gets measured from actual decisions players made at that pool.

  2. 2

    GTO baseline is the reference

    We hold those measured frequencies up against the game-theoretic optimum, sourced from solver dumps (PioSolver, Monker, GTOWizard exports). The signed gap — pool minus baseline — is the divergence. Every divergence carries its sample size and a confidence label; if data is thin, we say so.

  3. 3

    Adjustments are the counter-play

    Operators translate every significant divergence into a ranked exploit recommendation. Pool over-folds turn vs polar sizings? Barrel turn bigger and more often. Pool under-3-bets from the blinds? Open wider on the button. Every adjustment cites the divergences that justify it — no floating advice.

  4. 4

    Sessions drill the muscle memory

    A session is 10 scenarios from the pool, each exercising one ranked adjustment. You pick an action, we grade against the exploit pick, and the feedback ties your decision back to the divergence that motivated it. ~7 minutes per session.

Try it

One real spot, graded live.

Demo spot · no sign-in required

Pick how you'd play this hand

Real scenario from the Suprema pool. We'll grade you against the exploit pick.

SPOT · PREFLOP

You're on the BTN vs MP.

PREFLOP
preflop
POT2.5bb
BTN · YOU
100bb
SB
100bb
BB
100bb
UTG
100bb
MP
100bb
limp1bb
CO
100bb
K
J
Broadway offsuit
EFF 100bb

← YOUR ACTION →

Glossary

The jargon, translated.

Every term that has a dotted underline anywhere in the product opens a popover with the short version. Here's the full list.

Concepts(11)

GTO
GTO is the equilibrium strategy. Against another perfect opponent it can't lose; against a flawed real-world pool it leaves money on the table because it doesn't exploit mistakes. SharkIQ measures real pool behavior and shows you exactly where to deviate from GTO to capture that pool's leaks.
Pool
A specific player population — Suprema NL10–NL50 evening Brazil — defined by room, stake, time slot, and ecology. SharkIQ measures one pool at a time from real hand histories.
Divergence
Signed gap between this pool's measured frequency (HH aggregate or operator-tagged sample) and the GTO baseline for the same spot class. Carries a sample size and a confidence label.
Exploit
A counter-strategy that capitalizes on a measured divergence — gains EV that strict GTO play leaves on the table.
Adjustment
Operator-curated exploit recommendation tied to ≥1 supporting divergence. Ranked by estimated bb/100 gain. No floating advice — every adjustment cites the divergences that justify it.
Baseline
The GTO-derived reference frequency for a spot class. Sourced from solver dumps (PioSolver, Monker, GTOWizard). The divergence is computed by subtracting baseline from pool measurement.
Range
The full distribution of hands a player could have at a decision point — not a single hand.
Fold equity
EV captured from villain folding before showdown. Bigger bets = more fold equity, fewer call-backs.
Capped range
A range without the strongest hands — usually because villain didn't raise earlier when they had them.
Polarized range
A range of only very strong + very bluff hands, no middling value. Big sizes signal polarization.
Value bet
Betting expecting to be called by worse hands — pure EV from villain's wrong calls.

Metrics(3)

bb/100
bb/100 normalizes win rate across stakes. A 5 bb/100 win rate at NL25 means $1.25 per 100 hands; at NL100 it's $5. Pros over very long samples typically run 2–8 bb/100.
EV gap
Expected-value difference between two plays. Positive = your pick was better, negative = you left money on the table.
Sample size
Number of hands a metric is based on. Below ~500 hands, variance dominates — treat numbers with caution.

Positions(6)

UTG
Under the Gun — earliest seat. First to act preflop, plays tightest.
MP
Middle Position — second to act preflop.
CO
Cutoff — one seat before the button. Wide opening range, great for steal attempts.
BTN
Button — last to act postflop. The widest range, highest EV seat.
SB
Small blind — forced 0.5bb post. Acts last preflop, first postflop. The hardest spot.
BB
Big blind — forced 1bb post. Wide defending range; closes the preflop action.

Actions(10)

Limp
Just calling 1bb preflop instead of raising. Usually weak — pool tell.
Iso-raise
An isolation raise (iso) targets a single weak opponent (usually a limper) and raises bigger than a standard open. Goal: get heads-up vs the weak hand and deny blinds easy multi-way overcalls.
Open-raise
First raise in. Opening size sets the pot for the rest of the hand.
3-bet
Reraise over the opener's preflop raise — pressure + value combo.
4-bet
Reraise over a 3-bet — usually polarized to AA/KK and bluffs with blockers.
C-bet
Continuation bet — the preflop aggressor bets the flop as a follow-through.
Double barrel
Following a flop c-bet with another bet on the turn. Pressures cards villain hasn't improved on.
Triple barrel
Betting flop, turn, AND river — the full pressure line. High variance, big EV vs over-folders.
Jam
All-in. Removes remaining decisions — pure equity vs villain's calling range.
Check-raise
Check first to look weak, then raise when opponent bets. Strong line — usually value-heavy.

Ready for the full ten-spot session?

Free to start. Sign in with Google or email.

Want to see the underlying data? Browse the pool catalog.