How to Practice Poker Without Burning Bankroll

Table time is not practice. Practice is a rep where you commit to an answer, find out immediately whether it was right, and repeat until the right answer arrives without effort. Most players confuse the two, play 20,000 hands a year and finish that year with exactly the leaks they started with. The loop below takes about three hours a week, and almost none of it touches your bankroll.

Why volume alone teaches you nothing

A session does give you feedback. The feedback is just a liar. You can play a river spot perfectly and lose $200, or call off with the worst hand at the table and get there. Over a few hundred hands, results tell you about the deck rather than about your decisions.

The maths is unforgiving. A strong online winner earns somewhere around 5bb/100 — at NL50 that is $2.50 per hundred hands, while a normal session swings tens of buy-ins in either direction. Your win rate needs tens of thousands of hands before it says anything your ego has not already decided.

Volume is also a slow way to collect reps. You see roughly 75 hands an hour on one online table and about 30 live, and you fold most of them. A 20-minute drill puts more genuine decisions in front of you than a four-hour live session.

Practice is any rep where you commit to an answer and then find out whether it was right. Everything else is just playing.

The weekly loop

The whole routine fits in three hours, spread across the week. The order matters: drill first, review second, fix third.

1. Two 20-minute range drills. Preflop spots, answered under mild time pressure, graded instantly.
2. One 45-minute review of marked hands. Six to ten spots from your own sessions, not random hands from a forum.
3. Thirty minutes on one leak. The single most expensive pattern your review turned up, and nothing else.
4. Then play. Sessions are where you collect next week's material.

Play sessions are deliberately not part of the three hours. If you cut study to make room for volume, you have gone back to paying for lessons the deck will never give you.

Drill ranges until they are automatic

Preflop is the decision you make most often and the only part of poker you can simply memorise. Get it wrong and every street after it starts from a worse place, with a range that cannot defend the boards it will meet.

The gap between seats is enormous. The BTN opens roughly 40-45% of hands in 6-max; UTG opens roughly 15-17%. If you play those two seats within ten per cent of each other, you are either bleeding chips from early position or leaving money on the button every orbit. Grab preflop charts to drill from and check your own opens against them before you assume this is someone else's problem.

Reading a chart is not drilling it. Reading feels productive and teaches almost nothing, because recognition is not recall — you will nod along at K9s in the HJ and then misplay it at the table. Drilling means a spot appears, you answer, you are told, the next spot appears.

Review the hands you actually played

Mark hands while you play — every spot where you paused, plus every pot over about 40bb whatever the result. Do not analyse mid-session; that costs you money in the hand still in front of you. Marking takes a second and hands your future self a list worth studying.

Review with the guess-first method. State the decision, commit out loud to an answer and a reason, and only then look. If you check first, you will agree with whatever you see and learn nothing, which is why reviewing your own hand histories so often feels useful and changes nothing.

Close calls deserve a number before an opinion. Run a pot odds calculator on every marginal call you marked: facing $50 into a $100 pot you risk $50 to win $150, so you need 25% to break even. Being wrong about the price is the cheapest leak in poker to fix and one of the most common.

Fix one leak at a time

Your review will surface five problems. Pick one. Written as a rule you could follow half-asleep, a leak becomes something you can actually change this week.

• Name it as a behaviour, not a feeling. Not "I play badly in the blinds" but "I fold the BB to button opens too often".
• Write the rule. Against a 2.5x BTN open, defend every pair, every suited ace and every suited broadway before you consider folding — a simplification, and a far better one than the fold reflex it replaces.
• Pick one number to watch. Blind defence frequency, fold-to-3bet, or how often you continuation-bet after raising. One stat, measured weekly.
• Give it a fortnight. Then take the next leak, and let the fixed one run on autopilot.

Chasing five leaks at once means none of them sticks, because you cannot tell which change moved which number.

Practice modes compared

Not every hour of study buys the same thing. Decisions per hour and quality of feedback are what separate the modes:

Practice mode	Decisions per hour	Cost of an error	Feedback
Live cash session	5-10	Real money	None, and delayed
Online 6-max, one table	25-40	Real money	None, and delayed
Play-money tables	25-40	Nothing	None
Hand-history review	15-25	Nothing	Yours, or a solver's
Solver-graded drill	150+	Nothing	Instant and exact

The pattern in that table is the entire argument. The modes that cost you the most money give you the fewest decisions and the worst feedback. That is why poker practice games with feedback sit at the top of the list: the grading is the product, and the reps are free.

What the first month looks like

Week one is a baseline. Drill your opens, review ten marked hands, and write down the leak that shows up most. Expect the drill to be humbling — the seats you never think about are usually the ones costing you the most.

Weeks two and three are repetition. Same loop, same leak, one stat tracked. This is the part everyone skips because it is dull, and it is where the improvement actually happens.

Week four you re-test. Run the same drill cold, check whether your one number moved, and pick the next leak. From there the loop repeats indefinitely, which is what a structured poker study routine really is: a small number of habits you do not have to renegotiate every week.

Three hours a week, most of it away from the money. Do that for a month and you will have more useful reps than a year of grinding blind ever gave you — and a bankroll still intact enough to use them.