Glossary

EV (expected value) is the average result of a decision over the long run: the sum of all possible outcomes weighted by their probabilities. A decision with positive EV (+EV) makes money in the long term even if a particular hand is lost; a decision with negative EV (−EV) loses money even if you got lucky today.

Example: an all-in shove that wins a 20 bb pot 55% of the time and loses 10 bb 45% of the time has EV = 0.55 × 20 − 0.45 × 10 = +6.5 bb. Comparing the EV of shoving versus folding is the foundation of push/fold strategy.

Equity is the share of the pot that "belongs" to your hand right now: its probability of winning at showdown multiplied by the size of the pot. AA versus KK preflop has about 82% equity — of a 100 bb pot, your pair of aces "owns" 82 bb.

Equity changes with every street: a flush draw on the flop holds ~36% against top pair, but by the river it becomes either ~100% or ~0%. Decisions are built on comparing equity with the price of continuing — pot odds.

Pot odds are the price of a call relative to the final pot. If the pot is 100 and your opponent bets 50, the call costs 50 with a pot of 200 after you call: 50 / 200 = 25%. That is the minimum equity at which the call breaks even: win more often and you profit, less often and you lose.

The quick formula: bet / (pot + bet + your call). A half-pot bet gives 25%, a full-pot bet 33%, a 2x-pot overbet 40%. Comparing pot odds with your hand's equity is the first filter for any call.

Implied odds are the "hidden" part of your pot odds: the money you expect to win additionally on later streets once your draw comes in. Direct pot odds may not justify the call, but deep stacks and position can make playing on profitable.

The classic example is set mining: calling a raise with a small pair is justified when stacks are at least 15–20 times the price of the call — you flop a set only ~12% of the time, but it wins a big pot. The flip side is reverse implied odds: a hand that keeps paying off while second best.

Fold equity is the part of the expected value of a bet or shove that comes from the probability that your opponent folds. The total EV of aggression is the sum of two scenarios: "everyone folds — the pot is ours" and "we get called — we play on our equity".

Fold equity is exactly what makes wide-range shoves profitable on a short stack: even a weak hand earns money if folds happen often enough. The shorter the stack and the later the position, the more weight fold equity carries in the decision — it is the core of push/fold strategy.

Outs are the cards in the deck that improve your hand to a presumed winner. A flush draw has 9 outs, an open-ended straight draw 8, a gutshot 4.

The quick conversion of outs into equity is the "rule of 4 and 2": on the flop, multiply your outs by 4 (the chance to get there by the river); on the turn, by 2. A flush draw is ≈36% on the flop, ≈18% on the turn. Count only "clean" outs: a card that completes your straight but gives your opponent a flush is more of a minus than a plus.

MDF (minimum defense frequency) is the share of your range you must defend against a bet so that bluffing "any two cards" is not profitable for your opponent. The formula: pot / (pot + bet). Against a half-pot bet MDF = 67%, against a pot-sized bet — 50%.

Fold more often than MDF and your opponent prints money with automatic bluffs; defend wider and you overpay their value range. In practice, MDF is a guideline, not a law: against passive players who almost never bluff, overfolding to their aggression is more profitable than "correct" defense.

Variance is the natural spread of results around expected value. Even a player with a consistently positive EV loses for days and weeks: a hand with 80% equity loses every fifth all-in, and runs of such "bad luck" are the norm, not an anomaly.

The practical consequences: judge yourself by the quality of your decisions, not the result of the session; keep a bankroll that can survive the downswings of your format (tournament variance is several times higher than cash); do not change strategy over a small sample. A 20–30 buy-in downswing is business as usual for a tournament regular, not a reason to break your game.

Push/fold is the short-stack mode of play (usually up to 10–15 bb) where only two decisions remain: all-in or fold. A standard open raise no longer works at that depth — after a raise and a call you are committed to the pot anyway, and the fold equity of a shove is lost.

Shoving ranges are computed mathematically (Nash equilibrium) and depend on stack, position, and the number of players behind: at 10 bb the button profitably shoves ~40% of hands, early position about 15%. This is the most "solved" part of poker: the charts give a ready answer — all that is left is drilling them to autopilot.

An open raise is the first raise in a hand, made when everyone has folded to you. Opening with a raise rather than a limp is the standard: a raise wins the blinds without a fight, builds the pot with strong hands, and seizes the initiative.

The opening range widens from early position to late: UTG at a full table opens ~15% of hands, the button 40–50%. Typical online sizing is 2–2.5 bb (bigger with deep stacks and antes). The key rule: position matters more than cards — K9s is a normal open from the button and a fold from early position.

A 3-bet is a re-raise against an open raise (the first "bet" is the blinds, the second the open, the third your re-raise). It works with two ranges: value (QQ+, AK — building the pot) and bluffs with good properties — hands like A5s that block your opponent's aces and pick up equity when called.

Typical sizing is 3x the open in position and 4x out of position. Flat-calling instead of 3-betting keeps the pot small and lets the blinds into the hand, which is why the modern game prefers a polarized 3-bet to passive calls.

A squeeze is a 3-bet made after there has been an open raise and at least one call. The pressure lands on two players at once: the opener is uncomfortable continuing against a re-raise with a player still behind, and the caller has already shown their hand was not strong enough for a 3-bet of their own.

There is more dead money in the pot than usual, so the sizing is bigger too — around 4x the open in position and up to 5x out of position. The range is polarized: premiums for value plus bluffs with blockers like A5s; hands like 87s are better off just calling — seeing a cheap flop matters more to them.

A steal is an open raise from late position (cutoff, button, small blind) made primarily to pick up the blinds and antes without a fight. With antes there is ~2.5 bb in the pot before the deal — a 2.2 bb raise that works a bit more than half the time is profitable with no cards at all.

That is why stealing ranges are wide: 45–50% of hands from the button, and even wider from the small blind against a single opponent. The defense against a steal is the re-steal (3-bet) and wide big blind defense. In mid-stack tournament play, stealing and defending the blinds are the main source of chip accumulation between big pots.

A limp is entering the pot by calling the big blind instead of raising. The open limp in no-limit holdem is almost always a mistake: it does not win the blinds, does not build a pot for a strong hand, and gives up the initiative — the hand plays with no fold equity and no clear plan for the postflop.

Exceptions exist: limping the small blind with deep stacks, limping strategies on a very short tournament stack, and over-limping behind several limpers with a small pair or suited connector for a cheap flop. But the basic rule is simple: a hand is worth either a raise or a fold — "seeing a cheap flop" costs more than it seems.

A cold call is calling someone's open raise when you have not yet put any money into the pot (unlike defending a blind). The hand falls into a "middle" range: too weak to 3-bet, too strong to fold — pairs for set mining, suited connectors, strong broadways in position.

The downsides: you let the blinds into the pot at a good price and play without initiative, and a cold-calling range is capped — no premiums, since those would have gone into a 3-bet. That is why cold calling out of position has all but disappeared from modern strategy: there, the choice usually comes down to "3-bet or fold".

A continuation bet (c-bet) is the preflop aggressor's bet on the flop, keeping the pressure on whether or not they connected with the board. It works because the flop usually misses both players: an opponent who whiffed folds to a bet a meaningful share of the time.

Frequency and sizing depend on texture: dry boards like K72 rainbow allow a small c-bet (25–33% of the pot) with almost your entire range; connected, wet boards like 987 with a flush draw demand selectivity and a bigger size. The automatic "100% c-bet on any flop" is an outdated habit that gets punished with check-raises and floats.

A check-raise is a check followed by a raise after your opponent bets. It is a dual-purpose play: with strong hands it grows the pot off someone else's bet; with drawing hands it adds fold equity to your chances of improving, turning a passive line into a semi-bluff.

The check-raise is the main punishment for automatic c-betting: an out-of-position player who never check-raises invites bets with impunity. The classic balance is to check-raise both your strongest hands (sets, two pair) and suitable semi-bluffs on the same texture, so the line does not give away the strength of your hand.

A donk bet is an out-of-position bet into the player who was the aggressor on the previous street, instead of the standard check. The name hints at a classic amateur mistake: betting "to find out where I am" telegraphs a medium-strength hand and solves your opponent's problem for them.

A purposeful donk bet exists on textures that are dramatically better for the caller's range: a 653 flop after a big blind defense against early position puts all the sets and two-pair combos squarely in the defender's range. But the baseline out of position stays the same — check your entire range, then check-raise or call.

A semi-bluff is a bet or raise with a hand that is most likely behind right now but has outs to improve: a flush draw, a straight draw, sometimes two overcards. A semi-bluff has two paths to the pot: your opponent folds right away, or the draw comes in and the hand wins at showdown.

That sum — fold equity plus the equity of improving — often makes aggression more profitable than a passive call with the same draw. The best candidates are draws with no showdown value: calling rarely wins with them, while aggression adds a way to take the pot. A pure bluff with no outs is a different and far more expensive story.

A value bet is a bet designed to get called by a weaker hand. The check question before betting: "which worse hands will pay?" If only better hands call, the bet turns into a bluff against yourself — that is what the typical overvaluing of top pair on a scary river looks like.

Thin value is a bet with a hand just above average, when your opponent's calling range is only slightly weaker than your hand; the ability to bet thin is one of the main sources of winrate at low stakes, where opponents call too wide. Sizing follows the goal: big against inelastic ranges, smaller to keep weak pairs in the pot.

A blocker is a card in your hand that leaves your opponent fewer combos of certain hands. Hold an ace and they have half as many AA combos (3 instead of 6) and a quarter fewer AK combos. Bluff selection is built on this: A5s is a standard bluff 3-bet precisely because of the ace.

Postflop, blockers pick your river bluffs: the ace of spades on a three-spade board blocks your opponent's nut flush — bluffing with that "dead" ace beats bluffing with a random hand. The flip side is unblockers: cards that block your opponent's folding range make poor bluffs.

A barrel is a continuation of aggression on the next street: the second barrel fires on the turn after a flop c-bet, the third on the river. A single bet gets called wide, so a bluff often realizes its value only through your willingness to keep firing.

Good second-barrel cards are the ones that strengthen your range and weaken the caller's: overcards to the flop (an ace or king on the turn), cards that complete your draws. Bad ones improve your opponent's typical calling range. The key to barreling is planning the whole hand while still on the flop: which turns you continue on, and which ones you give up on.

ICM (independent chip model) is a model that converts tournament chips into monetary value through the probabilities of finishing in each place. The key consequence: chips are non-linear — doubling your stack does not double your expected payout, so a chip lost is worth more than a chip won.

Hence the fundamental tournament rule: risking elimination demands a bigger edge than in cash games. An all-in call that is profitable in pure chip EV often becomes a money loser on the bubble or at a final table — especially with short stacks still behind. ICM pressure hits the medium stack hardest: the big stack shoves with impunity while the medium stack has to fold even strong hands.

The bubble is the stage of a tournament right before the money, when only a few eliminations remain until the payouts. ICM pressure peaks: busting here is the worst outcome in the tournament — zero with the prizes almost in hand.

Stack dynamics diverge to the extreme. The big stack attacks almost with impunity — only premiums can call its all-ins. The short stack chooses between a desperate shove and trying to crawl into the money. The medium stack has it worst of all: plenty to lose, and no fold equity left against the big stacks. The last player eliminated before the money is called the "bubble boy".

An ante is a forced contribution every player makes to the pot before the cards are dealt, on top of the blinds. Modern tournaments widely use the big blind ante format: the player in the big blind posts the ante for the whole table — one chip instead of nine small ones — which speeds the game up considerably.

Antes change preflop math: there is ~2.5 bb in the pot before the deal instead of 1.5 bb, so stealing becomes more profitable, opening and blind-defense ranges widen, and the default fold gets more expensive. With antes, short-stack push/fold ranges widen too — there is simply more out there to pick up.

A knockout, or bounty tournament (bounty, knockout), pays real money for busting an opponent. In progressive knockouts (PKO), half of the eliminated player's bounty is added to your own, so big stacks become ever more valuable targets.

Bounties are dead money on top of the pot, so calling ranges against stacks you cover are noticeably wider: a call that loses chips is often profitable once the reward is counted. The smaller your stack relative to the price of your bounty, the more aggressively it gets called — and the less fold equity you have. The classic beginner mistake is chasing bounties without doing the math: the reward justifies a discount on your equity, not any call.

GTO (game theory optimal) is the Nash equilibrium strategy: a set of action frequencies such that no deviation can increase your opponent's winnings. A GTO player loses to no one in the long run — but also does not extract the maximum from other people's mistakes.

The practical value of GTO is not in memorizing solver outputs but in understanding the structure: why ranges split into value and bluffs in a certain ratio, why mixed frequencies exist, how protection against exploitation is built. The baseline is GTO; deviations are deliberate exploits of specific opponent mistakes: play theory against unknowns, punish the players you have studied.

An exploit is a deviation from the equilibrium strategy to maximize profit against a specific opponent mistake. Against a player who folds to c-bets 70% of the time, the right answer is not a balanced frequency — it is betting almost any two cards.

Typical low-stakes exploits: do not bluff calling stations (they call everything), overfold against passive players (their raises always mean strength), isolate limpers wide, go for thin value. The price of exploiting is your own vulnerability: by deviating, you open yourself up to counter-exploits. So the order matters: first understand the GTO baseline, then deviate from it deliberately — not the other way around.

Position is your seat at the table relative to the button, which sets the order of action. The player "in position" acts last on every postflop street and makes decisions having already seen the opponent act; the player "out of position" is forced to act blind.

It is the most durable edge in poker: information, control of the pot size, the ability to extract value or see a free card. That is why all preflop ranges are built around position — the button opens three times as many hands as UTG, and the worst seats at the table are the blinds, playing the entire postflop first. The practical rule: the worse your position, the stronger your hand needs to be.

A range is the entire set of hands a player can have in a given spot in light of their actions. An opponent who 3-bets from early position does not "have AK" — they have a range like «QQ+, AK, A5s», and the decision is made against that whole spectrum at once.

Thinking in ranges is the dividing line between a beginner and a prepared player. The beginner guesses at two cards; the regular narrows the opponent's range action by action and computes their own hand's equity against the final spectrum. Ranges are written as a 13x13 grid — the same as push/fold and opening charts — with hand fractions down to mixed frequencies.

Polarization is a range structure with two poles: very strong hands plus bluffs, with no middle. Big bets on later streets are naturally polarized: a medium-strength hand has no reason to inflate the pot — it either checks or bets small.

The opposite is a linear (merged) range: a continuous spectrum from strong hands downward, typical of opens and thin value. Understanding whose range is polarized and whose is capped (stripped of the nuts by earlier actions) directly dictates sizing: a polarized range bets big, up to an overbet, while a capped one is forced to defend. The ratio of value to bluffs inside a polarized bet is set by your opponent's pot odds.

Tilt is an emotional state in which a player deviates from their normal strategy: widening ranges, chasing losses, calling "on principle". The classic trigger is a bad beat or a run of coolers, but players also tilt from fatigue and after a big win ("euphoric tilt", with its illusion of invincibility).

Tilt is a direct consequence of misunderstanding variance: whoever expects winning play to show a profit every single day ends up angry at the math. What works: a stop-loss in buy-ins, a pause after a rough hand, judging a session by the quality of decisions rather than the result. The most expensive part is not the bad beat itself but the next hour of play after it.

Rake is the poker room's fee. In cash games it is a percentage of each pot (usually ~5%) with a ceiling, the "cap"; it is taken only from hands that reach the flop (the "no flop, no drop" rule). In tournaments the fee is built into the buy-in: in a "100+10", the ten is the room's rake.

At low stakes, rake is your main opponent: at typical sizes it eats several big blinds per hundred hands — more than the average player's winrate. Two practical takeaways: compare rooms by effective rake with rakeback included, and account for rake in your strategy — the higher it is, the tighter the profitable ranges for entering the pot.

Rakeback is the return to the player of part of the rake they have paid. The formats vary: a fixed percentage, progressive systems of statuses and missions, cashback promotions, rake races. The real return percentage an active player gets differs several-fold between rooms.

For a regular, rakeback is not a bonus but part of the winrate: at stakes where rake is comparable to profit, getting 20–40% of the fee back can double a month's result or pull a break-even player into the black. That is why rooms are compared by effective rake — the fee minus all returns; we collect each room's terms and current percentages on our review pages.

A HUD (heads-up display) is software that shows opponent statistics right on top of the table: how many hands a player plays, how often they raise, whether they fold to c-bets. The core trio: VPIP (percentage of hands voluntarily played), PFR (preflop raise percentage), and aggression. A VPIP/PFR of 55/5 is a passive recreational; 24/20 is a regular.

A HUD feeds on your own hand histories, so it works only where the room provides them. The industry is moving toward bans: some rooms block trackers entirely, others restrict them. Where HUDs are allowed, playing without one against equipped regulars means voluntarily giving up an information edge; each room's status is listed in our reviews.

Your bankroll is money set aside strictly for poker and kept separate from living expenses. Bankroll management (BRM) is a set of rules for choosing stakes so that natural variance can never wipe out your roll.

Rough guidelines: cash games — 30–50 buy-ins (100 bb stacks) for your stake, regular MTTs — 100+ buy-ins, large-field and knockout tournaments — up to 200–300: the higher the variance of the format, the deeper the cushion. The rule works both ways — moving down in stakes during a downswing is just as mandatory as moving up after growth is allowed. Playing with money you cannot afford to lose breaks both your decisions and your enjoyment of the game.