4 Thinking in Ranges, Not Hands

The single biggest leap from recreational to competitive poker is a shift in how you think about your opponent’s cards. Beginners ask, “What does he have?” They pick one hand, get emotionally attached to it, and play the rest of the pot as if that guess were a fact. Strong players ask a better question: “What is the full set of hands he could be playing this way, and how does each part of that set interact with this board and these bets?”

By the end of this chapter you will be able to assign an opponent a preflop range, narrow it street by street as the hand unfolds, name its shape, and choose your size and action from that range rather than from a guess at a single hand.

That set is called a range, and learning to think in ranges is the intermediate-to-advanced unlock. It is the conceptual engine underneath everything else in this book: solver outputs, exploitative adjustments, hand reading, and pot-odds decisions all assume you are reasoning about ranges. This chapter builds that engine from the ground up.

4.1 What a Range Actually Is

A range is the complete collection of hands a player can hold given everything you know about how the hand has played out. Rather than a single guess, it is a distribution: a list of possible holdings, ideally weighted by how likely each one is.

Consider a simple spot. A tight, competent player opens to 2.5bb from under the gun (UTG) at a 9-handed table. You do not know their two cards. But you know a great deal about the population of hands a sensible player opens from early position. That range looks something like:

All pairs 66+ (a tighter player trims to 99+, a nit to 77+)
AT, AJ, AQ, AK suited (the full run of suited aces from ATs up)
AQ, AK offsuit, plus AJ offsuit at the loose end
KQ, KJ, QJ, JT suited, and KQ offsuit
Occasionally a suited connector like T9s as a “mix”

Written as a range, that is roughly the top 8–12% of all starting hands. You will never know which specific combination they hold, but you know the neighborhood, and the neighborhood is enough to make excellent decisions.

Key idea

A range is a probability distribution over hands, not a guess at one hand. Your job is never to “be right” about the exact two cards; it is to make the most profitable decision against the whole distribution. You can be “wrong” about their specific hand on every single street and still win money for years, because you were betting against the average of the range.

Notation we will use

Throughout the book we describe ranges in standard shorthand:

Pairs: 22+ means all pairs twos and up; 99+ means nines and up.
Suited / offsuit: AKs is ace-king suited, AKo is offsuit, AK is both.
Plus on a kicker: ATs+ means ace-ten suited through ace-king suited (ATs, AJs, AQs, AKs).
Connectors: JTs is suited jack-ten; T9s+ means the suited connectors from T9s upward.
Combos: the number of specific two-card combinations. There are 16 combos of any unpaired hand (12 offsuit, 4 suited), 6 combos of any pocket pair, and 4 combos of any suited hand. We will use combos lightly here and treat them in full in a later chapter.

4.2 How to Assign a Range

Assigning a range is detective work in two phases: set the preflop starting range, then narrow it street by street as the opponent acts.

Step 1: Anchor on position and player type

Position is the first and strongest filter. The earlier a player opens, the tighter and stronger their range, because more opponents remain to act behind them. A reasonable set of opening ranges for a typical competent online or live-pro player looks like this:

Position	Approx. open %	Flavor of range
UTG (full ring)	8–12%	Tight, high-card and pair heavy
MP / Lojack	12–16%	Adds more broadways, suited aces
Cutoff (CO)	22–28%	Wide; many suited gappers, smaller pairs
Button (BTN)	40–50%	Very wide; any two “playable” cards
Small blind (SB)	25–40%	Linear and wide in a raise-or-fold strategy; more polarized only with a limp range (strong hands raise, mediums limp)

These are typical ranges for thinking opponents and should be treated as starting estimates, not laws. A loose-aggressive recreational player might open 60% from the cutoff; a nitty live regular might open 15% from the button. Your job is to adjust the anchor to the human in front of you, which is where player typing (covered in the psychology chapters) feeds directly into range work.

Step 2: Apply every subsequent action as a filter

Each action an opponent takes is information that removes hands from their range, adds hands, or shifts the weights. Think of it as passing the range through a sieve on every street.

Calling a 3-bet removes the very top (which would 4-bet) and the very bottom (which would fold), leaving a middling, “condensed” band.
Donk-betting (leading into the previous street’s aggressor instead of checking to them) on a dry board is rare for strong players and usually signals either a specific made hand or a specific bluff, depending on the player.
Checking back the flop caps a range: it removes the strongest value hands that would have bet, which matters on later streets.
Betting big (an overbet, larger than the pot) usually polarizes the range, signaling “very strong or bluffing” rather than “medium strength.”

We will work a full example of this narrowing process at the end of the chapter.

Common mistake

The “sticky single read.” A player calls your flop bet, the turn brings a flush, and you blurt to yourself, “He has the flush.” Maybe, but the flush is only a few combos of a range that also contains top pair, sets, straight draws that just got there, and pure air giving up. Collapsing the range to one scary hand makes you fold winners and pay off the nuts. Name the whole range first, then ask how much of it beats you.

4.3 Why Ranges Beat Single-Hand Thinking

Three concrete reasons this matters at the table.

1. It makes your decisions robust to being wrong about specifics. If you decide “he has AK” and bluff, you are right or ruined based on one guess. If you decide “his range here is 60% missed broadways and 40% strong pairs, so a bluff folds out the 60%,” you are making a bet that profits on average even though any individual hand might be the 40%.

2. It protects you from results-oriented tilt. When you reason about ranges, a lost pot where the opponent turns over the one hand that beat you is simply the tail of a distribution you already accounted for. You priced it in. That connects range work to the psychological game: a bad beat becomes expected variance rather than personal failure.

3. It is the only language solvers speak. Modern study tools (GTO solvers like PioSOLVER, GTO+, or simplified trainers) do not solve for “what he has.” They solve for the equilibrium strategy of one range against another range. If you want to use that body of knowledge, you must think the way the tools think.

Key idea

You are not playing your hand against his hand. You are playing your range against his range, and then selecting the best action for the specific hand you happen to be holding within your range. Both halves matter: the range frames the decision; the specific hand picks the action.

4.4 Range vs. Range Equity

Once you can name two ranges, you can ask the central question of board analysis: whose range does this board favor? This is range vs. range equity: the average equity of one entire range against another, and just as importantly, how that equity is distributed across each range.

Suppose the cutoff opens and the big blind calls. Roughly:

CO opening range: ~25% of hands, with lots of broadways, suited aces, pairs, and suited connectors.
BB calling range: wider and weaker, with many suited gappers, offsuit broadways, small pairs, and hands too weak to 3-bet but good enough to defend at a discount.

Now run two different flops:

Flop A: A♠K♥4♦ rainbow (three different suits, so no flush draw). This board hammers the preflop raiser. The CO’s range is full of Ax and Kx and the big broadway combos (AK, AQ, KQ), while the BB defended a range that holds far fewer top-pair-or-better combos. Run the two ranges through an equity calculator (Equilab, Flopzilla, or GTO Wizard) and the CO comes out around 57–60% against the BB’s defending range. The CO holds far more of the nutted hands here: a range advantage, and specifically a nut advantage. The CO can therefore bet a high frequency, often for a small size, because their range is stronger almost everywhere.

Flop B: 7♥6♥5♠. Now the nut advantage flips to the BB. This board connects far better with the big blind’s suited connectors and gappers (87, 65, 98, 43) than with the cutoff’s ace-high and broadway-heavy holdings. The CO still has overpairs and overcards that keep the overall equities much closer to even, but the BB’s range hits far more two-pair, straights, and strong draws, so the nut hands now sit disproportionately on the big blind’s side, even though the CO had the preflop initiative and perhaps a hair more raw equity. This is exactly why a thinking big blind can profitably check-raise low connected boards: the math of the ranges supports it.

Do not memorize these two boards. Internalize the process: name both ranges, then ask which one contains more of the hands that thrive on this texture.

Drill

Take five flops you played this week. For each, write the opponent’s likely preflop range in shorthand, then label the flop with one word: does it favor you, favor them, or is it neutral? Do not look at any cards you actually saw. You are training the range-mapping reflex, not grading the result.

4.5 Range Morphology: The Four Shapes

Ranges have shape. Two ranges with identical average equity can demand opposite strategies because their hands are distributed differently. Four shapes come up constantly, and naming them quickly is a real edge.

1. Linear (depolarized)

A linear (or depolarized) range is “the top X% of hands, in order of strength”: strong hands, good hands, and decent hands, with nothing intentionally weak mixed in. A standard early-position open is linear: best pairs, best aces, best broadways, sloping smoothly down to the worst hand that still makes the cut. Linear ranges want to bet for value and protection, usually at moderate sizes, because most of the range is ahead. A closely related shape is the merged range, which keeps that value-heavy core but deliberately adds some thinner value-bets and a few bluffs, a middle ground between a pure linear range and a fully polarized one.

2. Polarized

A polarized range is built from two extremes with the middle removed: very strong hands (value) and weak hands you are bluffing with, but none of the medium-strength hands. A big turn or river overbet is the classic polarized action, the “I have the nuts or I have nothing” bet. Polarized ranges justify large bet sizes, because the value half wants maximum money in and the bluff half does not care how much it bets (it is folding to a raise either way). When you see a huge bet, your first instinct should be that the range is probably polar.

3. Condensed (capped)

A condensed range is the opposite of polarized, concentrated in the medium band: decent made hands, but few or no nut hands and few pure bluffs. The classic example is a preflop flat-call of an open-raise. The monsters 3-bet and the trash folds, so the cold-caller’s range is full of middling pairs and suited broadways. Because a condensed range is capped, holding almost no nutted combinations, it is vulnerable to large bets and overbets. You cannot comfortably call a pot-sized river bet with a range whose best hand is second pair.

The two terms differ subtly. A range is capped when it has very few or no nut-level hands, a “ceiling” on its strength; it is condensed when its hands cluster tightly around medium strength. Capped ranges are usually condensed too, and either way polarized aggression exploits them.

4. Uncapped

An uncapped range still contains the nuts and near-nuts at full weight. The preflop 3-bettor, the player who raised the flop, the player who could always have a set or top two pair: these ranges are uncapped, and that alone lets them apply pressure with big bets, because the opponent can never be sure the nuts is gone.

Key idea

Match your aggression to morphology. Bet big against capped ranges: they cannot punish you. Bet big with polarized ranges: your value wants it and your bluffs do not mind. Bet small or check with condensed ranges of your own, since you have no nuts to protect and no reason to bloat a pot you might lose. The shape of both ranges tells you the size to use.

4.6 A Worked Example: Assigning and Narrowing a Range

Let us put the whole method together. 100bb effective, 6-max cash game online. The opponent is a solid, competent regular, neither a maniac nor a nit.

Preflop. The cutoff opens to 2.5bb. We are on the button with 9♠9♣ and 3-bet to 8bb. The cutoff calls.

Assigning the CO’s range. The CO opened ~25% and then called a 3-bet rather than 4-betting or folding. That action is a powerful filter:

The very top (AA, KK, QQ, AK) would mostly 4-bet against a button 3-bettor, so we remove or heavily discount it.
The trash he opened with, weak suited gappers and offsuit junk, folds to the 3-bet.
What remains is a condensed, mostly capped calling range: pairs roughly 22–JJ (some might 4-bet as a bluff, most just call), suited broadways like AQs, AJs, KQs, KJs, QJs, and the better suited connectors T9s, JTs.

Already we have learned something structural: his range is capped. He will rarely have an overpair to most flops and almost never the absolute nuts preflop.

Flop. The board comes Q♠8♥3♦ rainbow. Pot is ~17bb.

Range vs. range. This is a dry, disconnected board. Does it favor him? He has Qx (AQs, KQs, QJs, plus a few sets of queens, eights, and threes), so he holds some strong hands. But our 3-betting range has the overpairs KK and AA, top set with QQ, and strong top pair (AK, AQ, KQ). A note on terminology: on a Queen-high board, TT and JJ are underpairs, not overpairs; only KK and AA clear the top card. On balance the board is close to neutral or slightly ours: his best realistic hands are top pair good kicker and the occasional set, no overpairs.

We hold 9♠9♣, an underpair to the queen. We are ahead of his many missed broadways (AJs, KJs, JTs, ATs) and behind his Qx and bigger pairs. Classic medium strength: too good to turn into a bluff, not good enough to bloat the pot.

We bet small, 5bb (~30% pot). The size suits the texture: on a dry board our whole range can bet cheaply, we deny equity to overcards, and we keep his weak hands in.

He calls.

Narrowing again. The call removes most pure air (he would fold AJ-high with no draw) and leaves: Qx (top pair), pairs that beat us (TT, JJ occasionally), bottom set 88 and middle set 33 as rare strong hands, small pairs peeling to spike a set, and floats with backdoor or gutshot equity (AJs, KJs, ATs, JTs). His range is now medium and mostly capped: the sets are a thin slice, the bulk top-pair-and-worse.

Turn. The board pairs low: Q♠8♥3♦3♣. Pot is ~27bb.

This card matters more than it first looks: any 88 or 33 he peeled the flop with has just made a full house, and a stray QQ would now be the top full house. For the bulk of his range, though, the trey changes nothing. It completes no draw and improves no float, so our 99 still beats his missed broadways and loses to his Qx. His range is still capped at top pair the large majority of the time, and a capped range struggles against large bets. It is no longer bulletproof, though: that thin sliver of 88/33/QQ is exactly what the paired board just upgraded into full houses.

This is where range thinking pays off. Against that mostly capped range we can still apply polarizing pressure on later streets: bet large with our value (overpairs, trips) and our bluffs (busted draws), because big bets profit on balance when his best hand is usually top pair. The occasional full house is a built-in cost of that line, not a reason to abandon it. With 9♠9♣ specifically we are neither value nor a natural bluff; we are a bluff-catcher. The disciplined line is to check and realize our equity, planning to call a reasonable river bet from his bluffs and fold to his obvious value (including those rare full houses). We are not married to the specific nines; we play the role 99 occupies within our range.

The takeaway. Notice what we never did: we never said “he has AQ.” We tracked the whole distribution from a 25% open, through a 3-bet call that condensed and capped it, through a flop call that stripped the pure air, to a turn whose ceiling told us how to size and whether to bluff-catch. Each action was a sieve, and the decisions followed from the shape of what fell through.

Common mistake

Forgetting your own range. Range thinking is symmetric: the same opponent is mapping you. If you only ever 3-bet premiums from the button, your 3-betting range is uncapped but narrow and face-up, and a good regular will fold everything but his best hands. Construct your ranges so that your strong lines also contain enough bluffs (and your checks contain enough traps) that a thinking opponent cannot read you the way you are reading him. We develop this balancing act fully in the GTO chapters.

4.7 Putting It to Work

You do not need to compute combinatorics in real time to benefit from this chapter. You need a habit. On every meaningful decision, run the loop:

Anchor. What does this player open/defend from this position? Start with a percentage and a flavor.
Filter. For each action since, ask: what does this remove, add, or re-weight? Narrow the range street by street.
Shape. Name the morphology (linear, polarized, condensed, capped, uncapped) for both ranges.
Compare. Whose range does this board favor, and where are the nut hands?
Act. Choose the bet size from the shapes, and the specific action from where your actual hand sits inside your range.

Do this consciously for a few thousand hands and it becomes automatic, the mental model that quietly powers every other skill in this book. You will stop asking “What does he have?” and start asking the only question that pays: “What is his range, and what is the best thing my hand can do against all of it?”

Drill

Reconstruct three hands from your last session purely as ranges. For each: write the opponent’s preflop range in shorthand, then cross out hands at every street as their actions filter the range, and finally label the shape of what remains. If you cannot name the shape of their river range, you did not narrow hard enough, so go back and find which action you ignored.

4.8 Summary

A range is the full distribution of hands an opponent can hold, not a guess at one hand. You play your range against theirs, then choose the action for the specific hand you happen to hold.
Assign a range in two phases: anchor on position and player type, then filter it with every action, street by street.
Range vs. range equity asks whose range a board favors and where the nut hands sit. That answer drives how often and how large you bet.
The four shapes guide sizing: bet big against capped ranges and with polarized ones; bet small or check with condensed ranges of your own; uncapped ranges can pressure freely.
The payoff is a repeatable loop: anchor, filter, shape, compare, act.