Poker strategy, from pot odds to GTO

What poker strategy actually is

Every decision at a poker table is the same shape: you are being asked to put in a certain amount to win a certain amount, and you have to judge how often that trade comes out ahead. A strategy is a consistent way of answering that, not a list of plays to memorise.

Three things decide the answer. Your equity, meaning your share of the pot if the hand were played to the end from here. Your position, meaning whether you act before or after your opponent for the rest of the hand. And your range against theirs, meaning the whole set of hands each of you could hold, rather than the one you happen to be looking at.

Beginners lose money to very ordinary errors, not to subtle ones: playing too many hands, calling to see one more card at a bad price, and betting an amount that has nothing to do with the pot. Fix those three and you are past most opponents at low stakes without learning a single advanced concept. The rest of this page is written so you can read it in that order, and the mechanics it assumes are on the poker rules page.

Starting hands and ranges

A range is the set of hands a player could have, given everything they have done so far. Thinking in ranges is the single biggest step up from beginner play, because you stop asking "does he have aces", which is unanswerable, and start asking "of all the hands he would play this way, how many beat me", which is a question with a number attached.

Your own range starts before the flop, and it should be much tighter than instinct suggests and much wider on the button than under the gun. In a six-handed game a reasonable opening range runs from roughly the strongest 15% of hands in early position to around 40% on the button. Big pairs, big aces and suited connectors first, offsuit trouble hands last.

• Premium: A-A through J-J, A-K. Raise from anywhere, every time.
• Strong: T-T down to 8-8, A-Q, A-J suited, K-Q suited. Open from most seats; let them go against heavy resistance from an early raiser.
• Speculative: small pairs, suited connectors like 7-6 suited, suited aces. They want cheap flops and several opponents, because they either connect hard or fold easily.
• Trouble: offsuit weak aces, K-J offsuit, two face cards that keep making the second-best hand. These look playable and cost the most.

Ranges only work if they are consistent. If you raise your good hands and limp your speculative ones, your range is face-up to anyone paying attention and no amount of later cleverness repairs it. Play the same hands the same way from the same seat, and your opponents have to guess.

Position is worth more than cards

Acting last after the flop is a structural advantage that never goes away, because you make every decision knowing what your opponent did first. It lets you control the size of the pot, bluff at a lower cost, and get value from hands you would otherwise have to check.

The practical translation is unglamorous: play more hands on the button and the cutoff, far fewer from the blinds and under the gun. The same two cards can be a comfortable raise in one seat and a clear fold two seats earlier, and that is not a close call. It is the largest single edge available to a player who has learned nothing else.

Out of position, keep pots smaller and lines simpler. In position, you can call wider, bluff more, and value bet thinner, because the last word is always yours. The Texas Hold'em page has the seat-by-seat table if the names are unfamiliar.

Pot odds, equity and how to calculate them fast

Pot odds are the price you are being offered. The arithmetic is one division: what you must call, divided by what the pot will contain after you call it. That figure is the share of the time you need to win for the call to break even.

If the pot holds RM100 and your opponent bets RM50, you are calling RM50 into a pot that will be RM200. Fifty divided by two hundred is 25%, so you need to win at least a quarter of the time. Below that, folding makes money. Above it, calling does.

Then you need the other half: how often you actually win. Count your outs, the cards that make your hand, then use the rule of four and two. With two cards still to come, multiply outs by four for a rough percentage. With only the river left, multiply by two. Nine outs for a flush draw on the flop gives about 36% by the river, and the true figure is close to 35%.

Put the two together and most calling decisions answer themselves. A flush draw facing a half-pot bet on the flop needs 25% and has about 35%, so calling is correct. The same draw facing a pot-sized bet on the turn needs 33% and has about 19%, so it is a fold unless something else is paying for it.

That something else is implied odds: the money you expect to win later if you hit. Chasing a gutshot at a bad immediate price can still be right against an opponent who will pay you off for their whole stack. The mirror image is reverse implied odds: hands like a weak flush draw that cost you more when you hit than they win, because the times you make it, someone has made a better one. Both are estimates, not calculations, and both are routinely used as an excuse for a call that was simply bad.

Odds calculators and equity tools are worth using away from the table to check your instincts against the real numbers; the poker tools page covers what they do and where they mislead.

Bet sizing: what the number is supposed to mean

Bet an amount related to the pot and to what you are trying to achieve, never to how much you like your cards. Sizing that tracks hand strength is the most readable habit in poker, and opponents pick it up faster than you would like.

Two numbers explain most sizing decisions. As a bluffer, a bet of B into a pot of P needs to work B ÷ (B + P) of the time to break even, so a third-pot bluff needs to succeed 25% of the time and a pot-sized bluff needs 50%. As a defender, the minimum share of your range you must continue with to stop a bluffer profiting is P ÷ (P + B), about 67% against a half-pot bet, 50% against a pot-sized one.

Read those together and the logic of small bets appears. A small bet risks less, needs to work less often, and lets you bet a wider range of hands, which is why modern play uses a lot of third-pot bets on boards that favour the bettor. Large bets are for polarised situations: strong hands and genuine bluffs, on boards where your opponent has many hands they hate to fold and cannot comfortably call.

Whatever you choose, use the same sizes with strong hands and with bluffs. A player who bets big with the goods and small with air is not running a strategy; they are narrating their cards.

Bluffing that actually works

A bluff is a bet that only wins if your opponent folds. It follows that the first question is never "is my hand weak" but "can this opponent fold, and does the story I am telling make sense from their side of the table".

• Bluff with equity where possible. A semi-bluff (betting a flush draw or an open-ender) wins immediately when they fold and still has outs when they do not. This is where most profitable bluffing lives.
• Bluff the boards that favour your range. If you raised before the flop and the board comes A-K-4, you credibly hold more of those cards than a caller does. On 7-6-5 you do not, and betting there is a story nobody believes.
• Bluff fewer opponents. Against one player you need one fold. Against three you need three, and the arithmetic collapses quickly.
• Do not bluff a player who does not fold. At low stakes this is most of the table, and it is why bluffing is a smaller part of winning poker than television suggests.
• Give up cleanly. One bet that fails is a bet. Three bets fired because you have already invested two is how a small loss becomes a stack.

The balance point is that your bluffs and your value bets should look identical. Choose the hands you bluff with in advance: the ones with the least showdown value and the most outs. Do not decide in the moment that this is a good spot because you missed.

GTO and exploitative play

GTO stands for game theory optimal, and it is widely misunderstood as "the best possible play". It is not. A GTO strategy is one that cannot be beaten by any counter-strategy. It breaks even against perfect opposition and wins against imperfect opposition by exactly as much as their mistakes are worth. It is a floor, not a ceiling.

Its practical content is balance. You bet a mixture of strong hands and bluffs in the same spots and at the same sizes, in ratios that leave your opponent genuinely indifferent, since no adjustment they make gains them anything. Solvers compute those ratios, and the useful output for a human player is not a memorised chart but a handful of principles: bet more often on boards that favour you, use smaller sizes with wider ranges, defend enough that bluffs do not print money, and do not let your sizing reveal your hand.

Exploitative play is the opposite instinct: find the specific mistake in front of you and attack it. Against a player who folds too much, bluff more than any balanced strategy would. Against a player who calls everything, stop bluffing and value bet thinner than feels comfortable. Exploitative play wins more against imperfect opponents. It also opens you up to being counter-exploited, which is the price of it.

For nearly everyone, the honest answer is a mix: use GTO principles as a default so you are hard to read, and deviate deliberately when someone shows you a leak worth attacking. Studying solver output before you have fixed your preflop ranges is studying the roof before the walls.

Cash game strategy

In a cash game the chips are money, the blinds never rise, and you can leave at any time. That changes the goal from surviving to maximising every individual decision. You are never playing for a finishing position, only for expected value on this hand.

• Buy in full. Deep stacks are what let position and skill matter. A short stack turns every hand into a preflop shove-or-fold decision, which is easier to play against.
• Choose the table. Game selection is a larger edge than any technical adjustment. A table with several loose passive players is worth more than a tough table two stakes lower.
• Rake changes the maths. The house takes a percentage of most pots, so marginal hands that break even in theory lose money in practice. It is why playing too many hands is punished harder online than a solver would suggest.
• Stop-loss beats stop-win. Leaving because you are tired, tilted, or outclassed is a decision. Leaving because you are up is not — the table does not know or care.

Tournament play pulls in a different direction, because rising blinds and a prize ladder mean chips change value as the event goes on. That is a subject of its own on the tournaments page, including the survival premium near the money that makes correct play tighter than raw pot odds would suggest.

Bankroll management

Poker has enough variance that a good player can lose for weeks and a bad player can win for months. Bankroll management is the practice of making sure that ordinary variance cannot end your ability to keep playing, and it is the least exciting and most load-bearing part of any strategy.

Common guidance among serious players is to keep somewhere around 20 to 30 buy-ins for a cash game and 100 or more for tournaments, where the swings are far larger because most entries return nothing. Those are conventions, not laws. The right number depends on how big your edge is, how much the swings bother you, and whether losing the roll would matter to your life.

Two rules do more work than the numbers. Keep the poker money separate from money that has a job, so that a losing month is a losing month and not a problem. And move down in stakes when the roll shrinks, before the game forces you to. Playing a level you cannot comfortably lose at changes your decisions in ways you will not notice until they have cost you.

Online there is a third rule, and it is administrative rather than strategic. The roll is not in your pocket but in an account somebody else runs, so moving down in stakes only helps if the money can also move out when you decide it should. The money side of an online poker account is worth understanding before the bankroll rules can do any work.

Nothing here makes results predictable. There is no bet-sizing scheme, progression or staking plan that turns a losing strategy into a winning one, and any source promising a guaranteed return is describing something that does not exist.

Reading opponents, online and live

Physical tells are the most overrated subject in poker and the most entertaining, in that order. What actually generates reads is behaviour that repeats: the seats a player enters pots from, the sizes they use, how they respond to aggression, and whether their bets change shape when they are strong.

Online, the reliable signals are structural. Bet sizing that varies with hand strength. Timing — an instant call is often a routine decision, a long pause before a call is often a marginal one, and an instant large raise is rarely a bluff. And volume: a player at eight tables is following a default strategy and is not reading you at all, which makes them predictable and hard to bluff.

Live, the useful tells are also the boring ones: how quickly chips are counted out, whether a player looks at their cards when the flop comes suited, and above all the difference between how someone acts in a hand they have decided about and one they are still deciding. Treat any single read as a small adjustment to a decision you were already making on the maths. A read is a tiebreaker, not a reason.

The vocabulary for all of this, from bet sizes and lines to ranges and board textures, is on the poker terminology page. The A-to-Z glossary covers the terms that come up across every table game.

How to study, and what to expect

Volume without review mostly repeats the same mistakes faster. The players who improve are the ones who play a session, mark the three hands they were unsure about, and work out afterwards what the alternative lines were worth.

A workable order: fix your preflop ranges first, because they are the decisions you make most often and the easiest to get objectively right. Then work on bet sizing and pot odds, which are arithmetic and therefore learnable. Then board texture and range reading. Solver work and GTO study come after all of that, not before. They answer questions you cannot yet ask.

And be honest about the timeline. The rules take an evening. Beating the lowest stakes takes most people months of deliberate work, and there is no strategy, system or tool that shortens it or makes a session's outcome certain. Poker in Malaysia is restricted under the Common Gaming Houses Act 1953 and the Betting Act 1953, with an additional prohibition for Muslims under Sharia law; nothing here says real-money play is permitted where you live. If you do play for stakes, decide the amount before you sit down and stop at it. 18+.