Opponents with identities, not difficulty sliders
Ingrid waits all night for aces. Priya came to see cards, not to fold. Aisha throws
chips at every flop. A hundred players rotate through the table, and they all fall
into eight recognisable types. You are never asked to remember a person — you are asked to
recognise a pattern, which is the skill that transfers to a room full of strangers.
A coach that grades the read, not just the odds
Because the trainer knows each opponent’s true range, it can compute exactly what
the most profitable decision is against that
player — then tell you whether you found it. Not what a balanced strategy would do.
What actually wins against the person in front of you.
They adjust to you
The table tracks how often you fold. Fold too much and the thinking players start
betting at you with nothing — a shark inside an orbit, a calling station never.
How fast someone adapts is itself a tell you learn to read.
Quirks the statistics cannot show you
Two players can have identical numbers and still play nothing alike. One fires a
second barrel every time; one gives up on the turn. One saves every bluff for the
river; another has never bluffed a river in their life. One bets small with a merged
range and one bets big with a polarised one — a sizing tell you can learn to
read. None of it appears on a stats readout. You have to watch.
Two opponents built to beat the readout
The Trap looks passive on any HUD, because he never bets his good hands
— he checks them and raises you when you bet for him. The Gear Shifter
changes style every few hands on purpose, so the session average describes a player
who is not sitting there any more. Neither can be identified from statistics. That is
the point.
The live game: straddles, limpers, four-way pots
The parts of poker other trainers ignore — and the parts you
will actually face at $1/$2. Isolation sizing over limpers, why a straddle halves
everyone’s stack, and why one pair is a bluff-catcher when four people saw the flop.
It asks for your read before it tells you
Facing a bet, the app stops and asks one question first: against the hands
he would be betting here, are you ahead, marginal or behind? Then the verdict
leads with whether the read was right, before it says a word about the action. Guessing
wrong is not wasted — being asked before being told is what makes the answer
stick.
Show me his range
After any graded hand, one tap opens the whole table the coach used: all 169 starting
hands, shaded by how often this player bets each one, with the arithmetic
underneath. Not a model of a typical opponent — the actual weights the bot rolled
against. No other trainer can show you this, because no other trainer wrote the
opponent.
Nothing is tested before it is taught
Every chapter opens with a briefing that teaches exactly what its levels are about to
ask — the seats and why position pays, how to count outs, what a range even is,
what to do about each type — and ends by telling you what you are about to be
tested on.
Two trainers for the fundamentals
Opening ranges are poker’s basic strategy, so they get their own practice mode
that keeps its own memory and never touches your stats — including a map of the
hands you personally keep getting wrong. And a read trainer that starts you
with one opponent to name instead of five, then shows which types you are mixing up.
A written debrief at the end of every session
Not a scoreboard. It opens with the best decision you made, then what it cost you,
then the single most expensive call of the night with a button to replay it. It grades
decisions and never chips — and on a short session it refuses to put a letter on
you at all, because a handful of hands cannot support one.
A leak report that refuses to guess
It names your three biggest leaks and what each costs. Postflop costs are measured
from real expected value. Preflop costs are labelled estimates. Read leaks get no dollar
figure at all, because any number there would be invented. It will not judge you on
results, and it says so.