The Bicycle Guide to The Theory of Poker

1. 1Classify the ante before you sit. Add up every forced bet, including blinds, and divide by the average future bet. 5% or less is small, 15% or more is large, and anything between is average. The source's calibration points: $100 is average in a $1,000-$2,000 game, and 50 cents is average in a $5-$10 game.
2. 2Set your starting requirements from that class. A lower ante means tighter play and a higher ante means looser play. In a large-ante game, expect the looser requirements to carry into later rounds, so you bet marginal hands for value that you would throw away in a small-ante game.
3. 3Price each decision off the pot as it stands now. In an eight-handed $10-$20 game with $1 antes, the pot is $8, so a $10 bet lays $10 to win $8, and a caller gets $18-to-$10. Ignore which chips were yours. Your own ante and earlier calls are not part of the calculation. The one exception is the hold 'em blind: play it a little looser because the pot odds are better, not because the money is yours.
4. 4Measure what it costs to come in, relative to the betting limits, in the specific game. Check three things: the blind-call price against the flop bet, the price of a raise or a call of a raise against the flop bet, and the amount an early-position limper can be raised.
5. 5Use the structure comparison to choose hand types. When entry is expensive, play only hands that don't depend on high implied odds. Raise the value of ace-king and big pairs and cut the value of hands like 6-7 suited and baby pairs.
6. 6Scan the players before you commit to the table. You need one or two bad players, or five or six mediocre ones, to have a profitable game. If nobody looks weaker than you, find another game.
7. 7In high-ante pots, raise good hands rather than slowplay them, because a cheap call is correct for opponents getting 8-to-1 or 10-to-1 on the first round. In low-ante games, slowplaying big hands to draw in worse hands becomes more reasonable.
8. 8In high-ante games, attempt ante steals, especially against tight players, because they carry positive expectation. In multi-way pots, shift value toward four-flushes and open-end straights and away from mediocre pairs.

The source compares a $10-$20 hold 'em game with the Golden Nugget $15-$30 game. In the $10-$20 game, calling the $5 blind puts in half of the $10 flop bet. In the $15-$30 game, calling the $10 blind puts in two-thirds of the $15 flop bet. Raising, or calling a raise, costs $10 in the $10-$20 game, which equals the flop bet. In the $15-$30 game it costs $25, almost twice the flop bet. An early-position blind caller in the $10-$20 game risks a raise of only the initial bet. In the $15-$30 game the same caller risks a raise of another $15, one-and-a-half times the initial bet. Entry is therefore much more expensive relative to the limits, so you cannot count on high implied odds. You play tightly: ace-king and big pairs gain value, and 6-7 suited and baby pairs lose it. The source says a player who understood this had an edge over players who were great in $10-$20 but kept playing the same way in $15-$30.

• If the ante is 15% or more of the average future bet, loosen starting requirements and carry that looseness into later rounds. If it is 5% or less, tighten up.
• If you hold a good hand and the ante is large, raise. If the ante is small and the hand is big, slowplaying to build the pot becomes reasonable.
• If the structure makes entry cost large relative to the flop bet, drop hands that need implied odds (baby pairs, small suited connectors) and upgrade ace-king and big pairs.
• If you are in the hold 'em blind, play a little looser because of the better pot odds. Do not play looser because you have already invested money.
• If a tight player is alone in a large-ante pot, try to steal the antes, since the play has positive expectation. If no one at the table is worse than you, leave.

• Carrying one game's hand values into a costlier structure. The tell is that you play 6-7 suited and baby pairs in a $15-$30-style game and keep losing when raised.
• Treating your own ante or earlier calls as yours. The tell is that you call a bet 'because I already put in two bets,' or play looser while dealing in a home game.
• Playing too tight in a high-ante game. The tell is that antes eat you up while observant opponents give you no action on your big hands, and you get beaten when you do get action.
• Slowplaying big hands in a large-ante pot. The tell is that opponents get 8-to-1 or 10-to-1 on the first round, hit a perfect card, and outdraw you.
• Ignoring who is at the table and focusing only on structure. The tell is that you sit down in a game of top players, with no sucker, because the stakes looked fine.

• ▸Total forced bets (antes + blinds): add every forced bet in the initial pot.
• ▸Average future bet: the typical bet size in this game.
• ▸Ante ratio and class: divide forced bets by the average future bet. Mark small (5% or less), average, or large (15% or more).
• ▸Blind-call cost vs. flop bet: write the fraction, as in $10 into a $15 bet.
• ▸Raise cost vs. flop bet: price of raising or calling a raise, as a multiple of the flop bet.
• ▸Early-limper re-raise risk: the amount you can be raised after limping, as a multiple of the initial bet.
• ▸Hand types to upgrade/downgrade: name the hands that gain value and those that lose it.
• ▸Player assessment: list the bad or mediocre players you can identify, and note whether there is a sucker.
• ▸Stay/leave and strategy notes: record your decision, the starting-requirement shift, and your raise-or-slowplay and steal plan.

1. 1Classify each opponent before choosing any deceptive play. Use the observable evidence the source points to: their play and exposed cards (the 'Reading Hands' material) and the type of game. Amateurs who want a lot of action and hustlers who prey on them favor exotic games, and the source ties the style labels loose-passive, loose-aggressive and tight-passive to who sits in the seat. Write one line per opponent: weak or unpredictable, average, or tough and deceptive.
2. 2Ask one question of every opponent: will this person make a mistake that my play can induce? The source's standard is that you give opponents every opportunity to make a mistake, because that mistake is your gain even when they get lucky and win the hand. Deception is a tool for creating mistakes, not for showing off.
3. 3Against weak opponents, check whether deception is needed at all. Weak players already lose to sound fundamentals: they chase and call too much, they rely on big hands and lucky draws. Tricks and ploys 'may sometimes work' in weaker games, but fundamentals are what you need. Keep your hand disguise to the minimum that still gets them to call.
4. 4Against tough, deceptive opponents, drop gimmicks and play from sound precepts. The source says a hustler with an arsenal of tricks who is not also a good player will not get the money against tough competition. Expect your tells, ploys and predictable raises to be read. Rely on correct odds, position and hand reading, and treat any deception as a calculated play with a stated reason.
5. 5Before raising, decide what the raise does to this particular opponent. Raising with the best hand is right only when opponents are getting good enough odds to overcall, or when you think an opponent will call a double bet though he shouldn't even call a single bet. If you know this opponent will fold to a raise but would misplay by calling if he knew your cards, just call.
6. 6Before bluffing or raising as a bluff, estimate two things about the opponent: the chance he holds the hand that beats you, and the chance he will throw away a hand that is merely good (the source's example is two pair). The source states plainly that this estimate requires the ability to read hands and to read players. Compare it with the size of the pot. If you cannot make the estimate for this opponent, do not bluff him.
7. 7Judge each decision by its logic, never by its result. The source's warning: you cannot judge a play by how it worked out in one instance. If a weak opponent draws out on you, the play was still right. If a bluff worked on someone who rarely folds, it was still wrong.
8. 8Update your classification after every showdown and every fold. Treat 'weak' as a hypothesis the table keeps testing. Because a good player naturally takes more bad beats than a poor one, a lucky draw by an opponent does not reclassify him as skilled.

The source's last-card example shows how opponent skill drives the choice. You were drawing to a flush, the pot is bet before you, and a player acts behind you. If you make the flush and raise, the player behind you may fold, and the original bettor may fold if he holds only two pair. If he has a full house and reraises, calling his reraise costs you two units and folding costs one. Now suppose you miss the flush and the first player bets. The source says you might raise against certain opponents, because if the bettor has only two pair he too may fold. Whether that bluff has positive expectation depends on the pot size and your estimate that the bettor does not hold a full house and will throw away two pair. Against a weak, predictable player who calls with two pair, the raise has no support and you check or fold. Against a player whose reads you trust, and who folds two pair to a raise, you can make the play. The cards are the same in both cases. Only your read of the opponent changes the decision. A related reminder comes from Bobby Baldwin in 1981. He lost the world championship lead when opponents hit two-outer queens (2 among 44 unseen cards, a 21-to-1 shot) twice in a few hours. His point is that a good player gets all his money in with the best hand and so is drawn out on more often. You do not change your play because a weaker opponent got lucky.

• If the opponent is weak and calls too much, then bet your good hands for value and skip elaborate disguise. The mistake you want comes from his over-calling, not from your trickery.
• If the opponent is tough and deceptive, then rely on correct odds, position and reading rather than ploys, and bluff only when you can defend a specific estimate of how often he folds.
• If you hold the best hand and an opponent would fold to a raise but misplay by calling a bet, then just call and give him the chance to make the mistake.
• If you hold the best hand and opponents are getting odds good enough to overcall, or one will call a double bet he shouldn't call even as a single bet, then raise.
• If you cannot judge whether an opponent will throw away a good-but-not-great hand, then do not bluff him. The estimate requires a read on the player.

• Using stock tricks against everyone, such as gesturing to fold and then raising. These work on weak players and fail against tough ones. The tell is that your ploys win small pots from weak players and lose big ones to strong players.
• Misjudging a weak player as predictable and bluffing him. If he will not fold a mediocre hand, your raise costs money. The tell is that your bluffs get called by hands you assumed he would fold.
• Judging a play by one result. You abandon a correct raise or call because a weaker opponent drew out, or you repeat a bad bluff because it once worked. The tell is that your explanation for a play starts with 'it worked' or 'he got lucky'.
• Raising to drive out a worse hand when the opponent would have called and made a mistake. Your raise cuts down his odds and removes the mistake you were paying for. The tell is that opponents fold hands you wanted to keep in.
• Treating a player as an expert because he beats you. A good player takes more bad beats than anyone, so his getting drawn out on proves nothing about skill. The tell is that you reclassify opponents after single hands instead of after patterns.

• ▸Opponent / seat: Name or seat number and game type (limit, pot-limit, no-limit; stud or draw).
• ▸Skill class: Weak/unpredictable, average, or tough/deceptive, plus the evidence you saw.
• ▸Style tag: Loose-passive, loose-aggressive, or tight-passive, with one observed hand.
• ▸Mistake he makes: Where this player errs, for example calling when he should fold or overcalling a raise. If none observed, write 'unknown'.
• ▸Fold-to-raise read: Will he throw away a good-but-not-great hand (two pair)? Yes, no, or unknown, plus evidence.
• ▸Pot size and odds he is getting: Are his odds good enough to overcall my raise?
• ▸Planned play: Bet for value, just call, raise, bluff or fold, with the one-line reason tied to his mistake.
• ▸Deception budget: None, light or full, and why (tricks only against weak players; sound play against tough ones).
• ▸Post-hand review: Was the logic right regardless of result? Do I change his classification, and what is the evidence?

1. 1Define the metric first: hourly rate = expected win per hour, which equals the sum of the mathematical expectations of your individual plays minus the rake (or seat charge). Track it as expectation per hour, not as results from a session.
2. 2Find the source of the edge before you sit. Your edge does not come from holding better cards. It comes from situations where your opponents would play incorrectly if they had your hand and you had theirs. List the specific mistakes you see and name each one.
3. 3Price each mistake. Estimate how much one instance costs the player on average, then multiply by how often they make it per hour. Lowball example: three players call $10 and draw two cards, losing about $2 per play, 8 plays an hour each, so 3 × $2 × 8 = $48 an hour in total losses.
4. 4Split the pool. Divide the total hourly loss by the number of roughly equal players who can collect it. In the example, $48 shared among four equal players is $12 an hour each. Your rate is your share of the bad players' losses.
5. 5Deduct the house take. In a public card room, subtract the rake or hourly seat charge. The source cites Las Vegas rakes of 10% of each pot up to $4 in smaller seven-card stud games, and 5% up to $3 in larger stud games, hold 'em and most others. Net the rake out before deciding whether the game is worth playing.
6. 6Make the play that maximizes positive expectation or minimizes negative expectation on every decision, because your long-run win is the sum of those individual expectations. Judge each decision by its expectation, not by whether the hand won.
7. 7Decide whether to play at all. Don't sit in a game with an insufficient projected rate. Stay only if you expect weaker players to arrive or good players to tilt when losing. Quit if those good players jump off winners.
8. 8Adjust deliberately when variance matters. Against weak opponents with a limited amount to lose, or on a short bankroll, don't push small edges: skip maximum raises as a small favorite and fold marginal calls. You give up some hourly rate but protect the win. It takes a few more hours.
9. 9Review against the projection. If your estimate is correct, eventual winnings approximate your projected hourly rate multiplied by total hours played. Treat that as a working estimate based on judgment and experience, and revise it as you learn.

Sklansky's lowball example. You see three players each call $10 and then draw two cards, a very bad play. Each loses about $2 per call on average. Each does this eight times an hour, so each loses about $16 an hour and the three together lose about $48 an hour. Four other players are roughly equal in skill, including you. Those four split the $48, so your projected rate is $12 an hour. That is before the rake or seat charge, which you must subtract. Sklansky notes you can't be this precise in most games, since other variables affect the number. If you play 100 hours, the projection says you win about 100 × $12 = $1,200 net of whatever you deduct. Individual sessions will deviate from that, which is why you judge yourself on the rate and not on the night.

• If the rate you project after rake is insufficient, then don't sit, unless you expect weaker players to arrive soon or good players to start playing badly while losing.
• If a play has a higher expectation, then make it, even if it risks losing this pot. A good fold still earns or saves a specific amount.
• If your opponent would be correct to call or fold given what you actually hold, then you gain nothing from that play. Look for the line that lets them make the mistake. Sklansky just called $200 with the nuts to let a two-pair opponent call at 7-to-2 on a 10-to-1 shot, rather than raise him out.
• If opponents are weak but have little to lose, or your bankroll is short, then play less than optimally: fold marginal calls and skip maximum raises as a small favorite. You trade some rate for a safer win.
• If you are in a game with a low projection but dropping out would hurt your reputation for only playing when you have much the best of it, then staying may be worth it for political reasons.

• Judging by results. You feel anxious after a losing night or elated after a big score. The tell is that your mood tracks the session outcome instead of the quality of decisions. Treat poker as earning an hourly rate, not gambling.
• Ignoring the rake. You estimate your share of opponents' losses but never subtract the rake or seat charge. The tell is a projection that looks fine but a real win rate that falls short.
• Mislocating the edge. You believe you win because you hold better cards. The tell is that you can't name the specific opponent mistakes producing your profit. The edge is in situations where opponents would misplay if the hands were reversed.
• Raising someone out of their mistake. You maximize this pot's win instead of giving the opponent a chance to err. The tell is that you fold out hands that would have called incorrectly. Sklansky's call with the nuts that let his opponent call at 7-to-2 on a 10-to-1 shot is the model.
• Pushing small edges in the wrong spot. You raise maximum as a small favorite against weak players with limited money, or on a short bankroll. The tell is that you get lucky-bust or watch weaker players get lucky and quit big winners.

• ▸Game and stakes: name the game and limit you are assessing.
• ▸Opponent mistake #1: describe the specific error (e.g., calling $10 and drawing two cards in lowball).
• ▸Average cost per instance: estimate how much the mistake loses on average each time (e.g., about $2).
• ▸Instances per player per hour: how often each player makes it (e.g., 8).
• ▸Number of players making it: how many opponents commit the error (e.g., 3).
• ▸Total hourly loss pool: cost × instances × players, summed over all mistakes (e.g., $48).
• ▸Number of roughly equal players sharing the pool: include yourself (e.g., 4).
• ▸Your gross share per hour: pool ÷ sharing players (e.g., $12).
• ▸Rake or seat charge per hour: subtract what the house takes, using the actual rake structure.
• ▸Net projected hourly rate and decision: gross share minus rake. Sit, skip, or stay for a stated reason (weaker players arriving, tilting good players, reputation).

1. 1Replace the question. When a hand gives you trouble, stop asking "What do you do in this situation?" Ask "What do I consider in this situation before determining what to do?" Write the considerations down. Rules of thumb (fold this, call that, raise the other) stall you at the beginner stage.
2. 2State the objective before each decision: the object of the game is to make money. Judge every option by what it earns against these opponents, not by how it looks, how it feels, or whether it wins this one hand. Jack Straus's attitude is the model: he would bust his own grandmother if she were in a pot with him.
3. 3Run the structural considerations in the order the book builds them. First, the ante structure and the form of the game (limit, pot-limit, no-limit, and whether the pot has two winners, as in high-low split). Then your expectation in the hand. Then pot odds and effective odds. Then implied and reverse implied odds. Treat the math as one input. The source calls it only a small part of poker logic and far less important than the concepts.
4. 4Add the opponent-dependent considerations the source lists under deception: the ability of your opponents, the size of the pot, your bet size, and the number of opponents in the pot. Ask what giving your hand away costs, and whether this table is good enough to punish a predictable bet.
5. 5Add the positional and tactical considerations: your position relative to the others, whether a bet wins the big pot right away, whether you are giving or getting a free card, and whether the semi-bluff, check-raise or slowplay applies. These principles are essentially the same in limit five-card draw and no-limit hold 'em, so name the principle, not the game.
6. 6Check the mechanics that change the economics. In no-limit you can bet all the money in front of you, but you cannot raise a player with less money out of the pot. A short stack calls with what he has, a side pot forms for the others, and an all-in player wins only the money he called in the main pot. Know which pot you are playing for before you size a bet.
7. 7Strip out any tricks. For each plan, ask whether it still makes money against tough competition with no ploy attached. Gesturing as if to fold, then raising after the third man calls, can work in weak games, but a player whose edge is gimmicks does not get the money from strong opponents.
8. 8Judge the decision by whether you got the best of it, not by the result. Experts are at war with luck. They put their money in with the best of it and leave lucky draws to weaker opponents. Over the long run everyone gets the same proportion of good and bad cards, so the edge comes from decisions.

Take the trick the source itself uses: gesturing as though to fold your hand, then raising after the third man in the pot has called. A rule-memorizer files it under "things that work." A player with conceptual understanding runs the considerations instead. Against a weak table the trick may win a pot, which is the source's own concession. Against tough competition it adds nothing, because a super hustler with an arsenal of tricks who is not also a good player will not get the money. So the question becomes what the raise accomplishes in concept terms. Does it win a big pot right away? Does it drive out opponents? Does it get a free card? Is your hand a favorite or second-best? Is your position good? Those are the considerations that earn money, and the gesture is decoration around them. The same method handles the side-pot mechanics. With $500 in front of one player and $50,000 in front of another, the big stack can bet his whole $50,000 but cannot raise the short stack out. The short stack calls his $500 and a side pot forms for the remaining players. If the short stack wins, he collects only the main pot he called into, and the best remaining hand takes the side pot. A player who understands this sizes his bet knowing the short stack's equity is capped at the main pot. The source gives no numeric hand to compute here, so the example shows the method and not a figure.

• If you catch yourself asking "what do I do with this hand?", then switch to "what do I need to consider here?" and list ante structure, odds, opponents, pot size, position and number of opponents before acting.
• If a play depends on the opponent being fooled by a gimmick (a fake fold, needling, angering him) and not on the strength of your hand, your position or the pot odds, then drop it against tough players. It earns money only in weaker games.
• If the game is not a standard limit game, then adjust the concepts before applying them. The source says they relate particularly to limit, relate to pot-limit and no-limit only when properly adjusted, and do not always carry over to high-low split games with two winners.
• If a decision comes down to a mathematical calculation, then compute it, but give the concept (deception, position, free card, semi-bluff) equal or greater weight. The source ranks math as a small part of poker logic.
• If your expectation in the pot is positive, then take the action even when you may lose the hand. Experts accept losing individual hands and measure themselves by getting the best of it.

• Hunting for a rule per situation. Tell: you ask other players or books "what do you do with X in Y?" and your play breaks down the moment the situation differs slightly.
• Treating tricks and ploys as the essence of poker. Tell: you rely on fake folds, needling, or angering opponents, and your results hold up only in soft games and never in bigger ones.
• Over-weighting the math. Tell: you can quote odds precisely but cannot explain your bet size, whether you wanted to give your hand away, or how position changed the play.
• Carrying concepts across game types without adjusting. Tell: you apply a limit-game principle unchanged in no-limit, or in a split-pot game where there are two winners.
• Judging play by the last result. Tell: you call a decision good because you won the hand or bad because a lucky draw beat you, and you stop checking whether you had the best of it.

• ▸Game form and limit structure: note the game (draw, stud, hold 'em), limit/pot-limit/no-limit, and whether the pot has two winners, so you know how far the concepts apply.
• ▸Ante structure: note whether antes are large or small and how that changes the value of winning the pot now.
• ▸Objective and expectation: state in one line how each option makes or loses money against these opponents.
• ▸Pot odds and effective/implied odds: record the numbers, whether more cards are to come, and any reverse implied odds risk.
• ▸Opponents: note their ability and how many are in the pot. Ask what giving your hand away would cost.
• ▸Pot size and bet size: note the size of the pot, whether you want to win it right away, and what your bet size tells the others.
• ▸Free card / semi-bluff / slowplay / check-raise: name which principle applies, if any, and why.
• ▸Position: note your position relative to the other players and the bettor.
• ▸Stack and side-pot check: note effective stacks, who is short, and what main and side pots would form.
• ▸Decision and review: record your action and later check whether you had the best of it, not just whether you won.

1. 1Define the bet's expectation before you act: expectation is the average amount a bet wins or loses. Compute it as (probability of winning × amount won) minus (probability of losing × amount lost). Zero means a waste of time, positive means play, negative means pass.
2. 2Convert both sides into odds. Count the cards that win versus the cards that lose to get true odds (the 1-to-5 number guess is 4-to-1 against). Then compare them to the price the pot offers, which is the pot size to the cost of your call. You have the best of it only when the price paid exceeds the true odds.
3. 3With all cards out, call only if your chance of winning is better than your pot odds. Otherwise fold.
4. 4With cards to come, start from pot odds, then discount them. Add the future bets you will have to call, and reduce the apparent odds to effective odds. Do not use the raw pot ratio on early and middle rounds.
5. 5Check the four inputs for a drawing call: (1) your chances of improving, net of needed cards already out against you and including extra outs; (2) your chances of winning if you improve; (3) the odds on this round, allowing for a raise behind you; (4) your expected extra profit on the last round if you hit.
6. 6Add implied odds only when a hit produces a monster that figures to get action. Implied odds are your total expected win when the card hits divided by the present cost of calling. Fold when you miss rather than paying further bets.
7. 7Run a reverse-implied-odds check on every marginal call. If you are unsure where you stand, have little chance of improving, a call commits you to calling to the end, and your opponent can back off anytime, treat your odds as much worse than the pot suggests and often fold now.
8. 8Judge the decision by expectation, not by the result. Winning or losing a single bet does not change whether it had positive expectation. Do not let your session result alter your play.

Sklansky's small-pair example in $10-$20 hold 'em: you hold a small pair, there is $50 in the pot, and it costs $10 to call. Raw pot odds are 5-to-1. It is about 8-to-1 against flopping a third card to make trips, so 5-to-1 alone does not justify the call. But when you do flop a set, you should average about $100 further profit. Your implied odds are therefore about 150-to-10, or 15-to-1, which comfortably exceeds 8-to-1. The call has positive expectation. When the flop misses, you normally throw the hand away rather than call a bet. The same logic runs the other way for the opponent in draw poker: with $80 in the pot and your $10 bet, he gets 9-to-1 on a flush draw that is only about a 5-to-1 underdog. His call is correct, so you are rooting for him to fold.

• If your chance of winning is better than your pot odds and no cards are to come, then call. If it is not, then fold.
• If more cards are to come and later bets reduce your pot odds, then compare your chances against effective odds, not the raw pot odds.
• If a hit would give you a hand that wins a lot of extra money, then a call at less than the true-odds price is correct, as with a small pair at 5-to-1 for an 8-to-1 shot. Fold when the card misses.
• If you are unsure where you stand, can barely improve, are committed to later bets, and your opponent can back off, then fold, since reverse implied odds make the call far worse than it looks.
• If a bet is a likely second-best hand with a chance of drawing dead or being outdrawn after you hit, then fold most close plays because they swing into losing plays.

• Judging a play by its outcome: you regret a correct call that missed or praise a bad call that hit. The tell is that you say a play was wrong because it lost.
• Using raw pot odds on early rounds with more cards to come: you call because the pot is large and ignore the future bets you must also call. The tell is that you never compute effective odds.
• Inflating implied odds to justify weak draws: you assume you will be paid off even when your hit is not a monster. The tell is that you cannot name the hand you expect to win big with.
• Ignoring reverse implied odds: you call with a hand that looks good now but loses big when you are beaten and committed to the end. The tell is that you call 'just to see' and then keep calling.
• Letting your session result steer decisions: you play worse when stuck or leave a good game as a small winner. The tell is that your play changes with your session total rather than with whether you are a favorite in the game.

• ▸Pot size and cost to call: write the money in the pot and the amount you must put in now.
• ▸Raw pot odds: divide pot by cost to call (e.g., $50 to $10 = 5-to-1).
• ▸True odds against your hand: count winning cards versus losing cards (e.g., about 8-to-1 for flopping trips).
• ▸Future bets you must call: list likely bets on later rounds, including a possible raise behind you.
• ▸Effective odds: adjust the raw pot odds downward for those future bets.
• ▸Expected extra win if you hit: estimate the further profit from a hit (e.g., about $100).
• ▸Implied odds: expected total win when you hit divided by present cost (e.g., 150-to-10 = 15-to-1).
• ▸Reverse implied odds check: answer yes/no on unsure where you stand, little chance to improve, committed to the end, opponent can back off.
• ▸Decision: call or fold only when your odds beat the true odds after the adjustments above.

1. 1Start with the cost of transparency. If you play only one narrow set of hands, or always bet strong hands and check weak ones, a good opponent learns exactly what you hold and never errs, while you keep guessing at his hand. Name the specific information you are giving away (the source's example is playing only A,2,3 in razz) before you decide how to disguise it.
2. 2Grade your opponents first, because ability is the main criterion. The tougher they are, the more you play a hand other than optimally. Against weak or mediocre players, play straightforwardly: raise when you think you have the best hand, and do not add extra bets with weaker hands.
3. 3Disguise in both directions. With a strong hand, skip the last raise or check to slowplay. With a weaker hand, add an extra bet to look stronger. The source's example is a reraise with a three-flush on third street in seven-card stud, so that pairing on board later makes opponents fold, afraid of trips or two pair.
4. 4Scale deception to pot size. As the pot grows, good players fold no more than bad ones and bluff you less when you show weakness. Past a certain size, stop disguising and put in every bet you can.
5. 5Compare current bet sizes with future bet sizes. If early bets are much smaller than later ones, do not make a small raise with a big hand, because it puts opponents on guard. Check or slowplay it to get paid later. If bets rise sharply, consider putting extra action in with a weaker hand on the cheap round to create an impression that pays off when bets are expensive. In pot-limit and no-limit, where pots and bets can grow enormously between rounds, you can disguise more freely and give more free cards.
6. 6Count opponents. The more players in the pot, the less deception gains. You cannot fold everyone out with a weak hand, you lose bets when a strong hand misses a raise, and letting many in cheaply raises the chance of being outdrawn. Disguise mainly when the pot is heads-up.
7. 7Use the strength-showing semi-bluff when the opponent's call is correct on pot odds. If he is getting good enough odds to call knowing what you hold, you want him to fold, so show more strength than you have. Checking behind to induce an end-round bluff is a separate play, and it is worth it when the free card is cheap.
8. 8When inducing bluffs, you do not need to push opponents past correct bluffing frequency into being favorites to be bluffing. You only need them to bluff significantly more than the correct frequency. Induce bluffs from players who already bluff too much and from rare bluffers if you can get them to bluff more than their chance of having made the hand. Stop bluffs from players who bluff too little.

The source's real example is the 1978 no-limit hold 'em world championship at the Horseshoe. Addington held $275,000 and Baldwin about $145,000. Baldwin bet before the flop and Addington called. Baldwin bet $30,000 on the flop, which could mean a draw or a pair of queens, and Addington called, so he clearly held a good hand. On fourth street the ace of diamonds, a scary card, came with $92,000 in the pot. Baldwin bet $95,000, leaving himself almost broke. That is a bet sized to represent strength, and it cost him everything if called. Addington deliberated, decided Baldwin was not bluffing, and folded. Baldwin turned over worthless hole cards, an hour later won all the chips, and became champion. Deception worked here because the pot was large relative to the stacks and the story told by the betting was consistent: a bet before the flop, a bet on the flop, and a bet on the scare card.

• If your opponents are weak or mediocre, then play straightforwardly: raise your best hands and skip extra bets with weak hands. If they are tough, then mix in deception.
• If the pot is already large, as with pocket aces facing a big reraise in a no-limit pot, then drop all disguise and put in every bet you can.
• If early bets are small and later bets are much larger, then check or slowplay big hands early. A small early raise only warns opponents.
• If many opponents are in the pot, then reduce deception. If it is heads-up, disguise more.
• If an opponent is getting correct odds to call but is afraid of a stronger hand, then show strength (e.g., bet after pairing on board) to make him fold a hand he should call with. If he bluffs too much, then induce the bluff. If he bluffs too little, then stop it.

• Playing a tight, narrow range or a fixed pattern, so a good opponent reads you exactly. Tell: strong opponents consistently fold to you or play perfectly against you.
• Spending extra bets on deception against weak players who will not fold anyway. Tell: your semi-bluff raises get called by worse hands and you lose bets without gaining folds.
• Slowplaying or disguising in a huge pot or multi-way, where the benefit is small and outdraws are likely. Tell: you give free cards and get outdrawn by one of several callers.
• Letting someone who bluffs far too much slip away from his bluffs by betting into him, which stops bluffs you should induce. Tell: you bet and he folds hands he would otherwise have bluffed with.
• Giving a free card in a spot where the opponent can improve and take the pot. Tell: after you check behind to induce a bluff, he hits an ace or three-of-a-kind and beats you.

• ▸Opponent ability (weak / average / tough): grade the people in the pot.
• ▸Pot size now vs. later bets: note whether the pot is already large relative to stacks or limits.
• ▸Bet-size ratio, early to later: record how much bets will rise next round and whether this is limit or no-limit.
• ▸Number of opponents: heads-up, or how many; more players means less deception.
• ▸My hand strength vs. what I want shown: write the hand you hold and the hand you want opponents to think you have.
• ▸Planned play (slowplay, check-raise, semi-bluff, bluff, or straightforward): state it and its extra cost.
• ▸Opponent's bluff frequency (too much / about right / too little): decide whether to induce or stop bluffs.
• ▸Free-card risk: list what improves the opponent and the odds (e.g., 21-to-1 for a small pair to make trips).
• ▸Verdict (deceive / play straight) and reason: one line tying the fields above together.

1. 1State the goal before you read anything: per the Fundamental Theorem, the closer you come to knowing an opponent's hand, the fewer mistakes you make. Every read ends in a decision, either play as if you saw their cards or pick a line that makes them misplay.
2. 2Write down the candidate holdings in a specific, named form, such as 'full house', 'two pair' or 'straight draw'. Do not settle for 'strong' or 'weak'. The source frames the key read as a binary: does the initial bettor have a full house, or only two pair that he may throw away?
3. 3Cross out holdings using exposed cards and the action so far. Chapter 22 treats these as the two raw inputs: the opponent's play and the cards you can see. A holding that contradicts either goes off the list.
4. 4Put a number on the read, because the decision depends on it. The source says your estimate of the chance that the bettor does not have a full house and will throw away two pair, set against the size of the pot, decides whether a raise has positive expectation.
5. 5Test the read against the price. Compare your estimated chance of winning with the odds the pot offers. If you figure your chances are better than 5-to-1 (the source's example), calling is correct. If the read says you are worse than the price, fold. Folding correctly still earns money.
6. 6In a multi-way pot, read each opponent separately and then ask how their holdings interact. Assign each player a chance of winning, as in 30% / 50% / 20%. Then ask whether one opponent's raise would push out another player to your benefit.
7. 7Add the psychological layer last. Ask what your opponent thinks you hold, and whether his correct play based on that belief helps you. Sometimes you want him to read you accurately, as when you want him to know you have only aces so he raises and drives out a straight draw. Sometimes you want him to read you wrong.
8. 8Update the read each betting round. A raise, a reraise or a fold from someone behind you changes the candidate list. Re-run steps 3 to 5 before you act.

The source gives a multi-way spot. You hold a pair of aces and bet. Opponent A has two pair and Opponent B has a straight draw. Before the bet your chances are 30%, A's are 50% and B's are 20%. You read A as holding a made hand better than yours and B as drawing. If A raises and B folds, A rises to 60% but you rise to 40%, so the raise helps you at B's expense. Your edge is getting A to raise, and that depends on what A thinks you hold. You want him to believe you have only aces, not aces up, so he raises with the best hand and drives the straight draw out. You are getting good enough odds to call that raise, and B can no longer make a straight. The read here has two levels: what each opponent holds, and what A believes you hold.

• If you can name only vague holdings ('he's strong'), then you cannot yet compute expectation. Narrow to specific hands before putting chips in.
• If your estimated chance of winning is better than the pot odds you are offered (better than 5-to-1 in the source's example), then call. Otherwise fold, even if you will often lose the pot.
• If a raise only works when the bettor throws away two pair, then raise only when your estimate of that fold, combined with the pot size, gives positive expectation, or at least less negative expectation than folding.
• If one opponent's raise would drive out another player who threatens you, then in a multi-way pot you may want the raiser to read you correctly. Let him know your hand is weaker than it might look.
• If your opponent would play correctly given what you actually hold, then you gain nothing from the hand. Choose lines that make him play wrongly.

• Reading only your own hand strength and ignoring the opponent's range. The tell is that you cannot name what you are beating or losing to.
• Treating a read as a certainty. The source's own estimate, 'does he have a full house or only two pair', is a probability. The tell is that you reraise into a full house and pay two units instead of one.
• Ignoring the price when you have a read. A correct read still needs pot odds or expectation to justify the call. The tell is calling because 'he's probably bluffing' without comparing the odds to your chance.
• Reading each multi-way opponent as if he were alone. The tell is that you fail to notice a raise by A that removes B, or a bet that lets a drawing hand come cheaply.
• Assuming opponents read your hand the way you intend. The tell is acting as if they always know what you hold, when sometimes you want them to be wrong and sometimes you want them to be right.

• ▸Opponent / position: Name each opponent being read and where he sits relative to you.
• ▸Action so far: List each bet, raise, call or check in order.
• ▸Exposed cards: Note all visible cards that eliminate or support holdings.
• ▸Candidate holdings: Name 2-4 specific hands, such as full house or two pair.
• ▸Estimated chance of each: Put a rough percentage next to each candidate.
• ▸What he thinks I hold: Write the hand he is probably putting you on.
• ▸Pot odds vs. my win chance: Compare the price offered with your estimated chance of winning.
• ▸Chosen action and why: Call, fold, bet or raise, with the reasoning in one line.
• ▸Result and read check: After showdown, note whether the read was right and which input misled you.

1. 1Classify each opponent first. Is he a relatively tight player who seems to win too many pots without being called (a possible pot-stealer)? Is he an aggressive player who may already bluff slightly more than optimal? Or does he bluff less than he should? The classification sets the direction of everything that follows.
2. 2Pick one direction per opponent: stop or induce. Against a player who bluffs less than correct, or a tight player who looks like he steals, stop the bluffs and push him toward fearing to bluff at all. Against one who bluffs a little more than correct, induce and make him bluff even more. You are pushing him away from optimum bluffing frequency either way.
3. 3To stop a bluff, represent strength. In the source's hold 'em example, you raise with two small pair after he bets first. If he only calls with a pair of kings or a four-flush, he will not bluff the end. Strength shown earlier in the hand makes his later bluff look unprofitable to him.
4. 4To induce a bluff, represent weakness. In draw poker, open with two aces or two queens, then deliberately take three cards so the aggressive opponent sees one pair only. In hold 'em, check behind on fourth street with a high pair in the hole so he reads you as weak.
5. 5Price the cost of the induction before you try it. Checking behind gives a free card. An opponent with an ace beats you if any ace falls on the end. If he holds a small pair, the odds are 21-to-1 against him making three-of-a-kind. Run this check before you commit.
6. 6Commit to the follow-through before you act. If you tried to induce and he bets, call. If you tried to stop and he bets, fold. Decide this at the moment you choose the technique, not after he bets.
7. 7Tie it to the opponent's actual ability. Check whether the opponent bluffs roughly correctly, because against a near-correct bluffer you lose to the bluff about two times out of three on that last bet. Your best outcome is an opponent who never bluffs or bluffs far too much, so push him toward one of those extremes.

The source offers a draw poker case. You are the dealer and open with two aces, or even two queens. An aggressive player who first checked now calls. He draws one card, and you are sure he is on a come hand (a draw). You want him to bluff, so you go out of your way to take three cards, making it clear you start with only one pair. If he bets after the draw, you call. Even if you raised his bluffing tendency only slightly, you gained, because you now win more often when you call that last bet. The hold 'em version works the same way. With a high pair in the hole, you check behind your opponent's check on fourth street to look weak. This gives a free card. An ace gives him the best hand on the end, but a small pair improves to three-of-a-kind at 21-to-1 against. The stop-bluff mirror: against a relatively tight opponent who bets first, you raise with two small pair. If he only calls with kings or a four-flush, he will not bluff you out on the end.

• If the opponent looks tight but wins too many pots uncalled, then stop his bluffs by representing strength.
• If the opponent is aggressive and bluffs a bit more than correct, then induce by representing weakness and make him bluff more.
• If you tried to induce and he bets, then call every time. Folding only helps him bluff you out of more pots.
• If you tried to stop a bluff and he bets, then fold. If the hand is worth a call, you should not have tried to stop the bluff, since you removed the bluffs you wanted to catch.
• If he reraises, or calls and then bets the end, after you showed strength to stop a bluff, then usually throw your hand away. He would be afraid to bluff against that much strength.

• Inducing and then folding. The tell: you play weak, he bets, and you surrender, so he bluffs you out of more pots than he otherwise would.
• Stopping a bluff and then calling. The tell: you raise or show strength, then call his bet with a marginal hand. You have shrunk his betting range to hands you do not beat.
• Applying the wrong direction to the opponent. The tell: you induce against a player who bluffs too little, or stop against one who bluffs too much, so you push the opponent toward correct play.
• Ignoring the free card cost of weak-looking plays. The tell: you check behind with a high pair and an ace falls on the end, because you never weighed the 21-to-1 odds against his small pair improving.
• Running a technique against a near-correct bluffer without a plan. The tell: your calls on the last bet keep losing about two times out of three.

• ▸Opponent: name or seat.
• ▸Observed bluff tendency: too much, about correct, or too little, with the evidence (pots won uncalled, aggression).
• ▸Direction chosen: induce or stop, and why this follows from the tendency.
• ▸Technique: how you will represent weakness (check behind, take three cards) or strength (raise with two small pair).
• ▸Cost check: free card given? List his outs and odds (for example, a small pair improving at 21-to-1).
• ▸Pre-committed response: call if inducing, fold if stopping.
• ▸Result: did he bet, and was he bluffing?
• ▸Adjustment: keep, reverse, or drop the plan for this opponent.