Combo Counting Under Time Pressure

Assumptions: All examples use a 6-max online cash game at 100 big blinds effective with no rake, unless a different stack depth or format is stated in the example.

Everything in this module so far was homework-speed: full ranges, complete censuses, tables. At the table you get a time bank and an opponent who notices when you use it. This final lesson introduces no new theory — every technique here is one you've already used — but it rebuilds them as reflexes. The goal is concrete: base counts recalled instantly, any one-card removal adjustment done in under five seconds, and a defensible value-to-bluff estimate inside fifteen. Speed comes from three things: memorized anchors, batching rules that skip re-derivation, and knowing what not to count when the clock is running.

Tier one: the anchors

These must be recall, not calculation. If any of them takes a beat of thought, drill it until it doesn't:

• Pair: 6. Suited: 4. Offsuit: 12. Unpaired total: 16.
• The pair ladder by visible cards of that rank: 6 → 3 → 1 → 0 (zero, one, two, three visible).
• Unpaired hands: each visible card of either rank deletes a quarter of the full 16 — one card visible leaves 12, one of each rank leaves 9.
• A suited combo dies only when one of its two exact cards is visible: 4 → 3.

That last anchor kills a persistent illusion worth five seconds of every counting attempt you'll ever make: suits don't "interfere" across ranks. The 4♦ on board does nothing to T♦9♦ — no card is shared. Only the T♦ or the 9♦ themselves would kill that combo.

Tier two: five-second adjustments

The drill format: see the situation, say the number, check it. Every answer below was verified by enumeration; your job is to produce it from the anchors without re-deriving.

"How many JJ on J♥7♠2♦?" One jack visible → pair ladder says 3. 3 combos.

"How many AK does he have when I hold A♣K♦?" One card of each rank gone: 3 × 3 = 9 combos (2 suited — hearts and spades survive — plus 7 offsuit).

"How many T9s on T♠9♠4♦?" Two of its exact cards visible (T♠ and 9♠ kill the spade combo); the 4♦ shares a suit with T♦9♦ but is neither of its cards, so it kills nothing. 4 − 1 = 3 combos: T♥9♥, T♦9♦, T♣9♣. If you said 2, you let the diamond on board bluff you — re-read anchor four.

"How many AQ on Q♠7♥2♦?" One queen gone: 4 × 3 = 12.

"How many KQ on K♥Q♠2♦?" One of each: 3 × 3 = 9.

"How many A5s when the board shows A♥5♣2♦?" Both ranks hit once, and this is the suited case: the heart combo (A♥5♥) and the club combo (A♣5♣) each lost a card → 2 combos (spades, diamonds).

"How many 77 when one seven is visible anywhere?" Ladder: 3. It does not matter whether the seven is on the board or in your hand — visible is visible.

"How many 98s on 9♥8♥2♣?" The heart combo is dead: 3.

"How many AJ when I hold the A♦ and the board has the J♥?" 3 × 3 = 9.

"How many 66 on 6♠5♦4♣?" One six visible: 3 — and note in passing they're all sets, on a board where straights are live. Counting and texture-reading happen in the same glance.

Ten reps like these per session — invent them from whatever board you just folded to — and the adjustments stop costing attention you need for the actual decision.

Tier three: batching rules

Batching is recognizing that whole classes adjust identically, so you compute the class once:

• Sets on an unpaired board are always 3 combos per matching pocket pair. Three board ranks, one card of each gone, ladder says 3, every time. "How many set combos can he have?" on any unpaired flop is just 3 × (number of his pocket pairs that hit). No enumeration.
• Every unpaired hand using one board rank drops to 12 (or 9 if you also block a card). All top-pair classes shrink by the same quarter at once.
• Every suited hand whose exact card is on the board drops to 3. On a two-heart board, every heart-suited combo in his range is simply gone as a flush-draw candidate count of one-per-label — which is why the flush-draw census in the draws lesson was "about one combo per live suited label," computable as fast as you can count labels.
• Offsuit broadways are the heavy freight: 12 each. When estimating range sizes under pressure, count the offsuit labels first and multiply by 12; the suited stragglers are rounding.

The big-three shortcut

When the clock is genuinely running — a river jam, eight seconds left — do not attempt the full census. Count three things, in this order:

1. Sets (3 per matching pair on an unpaired board — instant from batching)
2. Top pair (the dominant unpaired value class: 12, or 9 with your blocker)
3. The single most obvious busted or live draw class (usually 15-16 combos if offsuit versions exist, 2-4 if suited only)

Everything else — the second gutshot class, the 2-combo two-pairs — moves the answer by points, not tens of points, exactly as the sensitivity analysis in the bluff-catching lesson showed. The big three give you the right call/fold answer in the large majority of close spots, and the spots they'd get wrong were too close to punish either action.

Watch it run at speed, twice.

CO

A♣♣A♣

K♦♦K♦

SBunknown

1. 1.CO raises to $2.50
2. 2.SB 3-bets to $11
3. 3.CO 4-bets to $24
4. 4.SB jams $100 total
5. 5.CO uses the count before acting

The preflop version is pure anchor work. You hold one ace and one king, so: AA → 3, KK → 3, AK → 9, QQ untouched → 6. Total 21, with the chop class (AK, 9 combos) the single biggest — which is why AK plays so much better against tight jamming ranges than its raw 39% equity suggests: 41% of the time the result is a tie, and the equity simulation confirms the structure (39% equity, but only 18% outright wins against 41% ties). You produced all of that from four memorized adjustments.

BTN

Q♠♠Q♠

Q♣♣Q♣

BBunknown

Flop

J♣♣J♣

Turn

Walk the river count once at full speed. Price: pot-size bet → 33% bar (you memorized the bar table in the bluff-catching lesson: pot = 33%, two-thirds = 29%, half = 25%). Sets: the board is unpaired, so batching applies — 88, 44, 22 at 3 each = 9 (JJ never arrives; the site's big blind 3-bets it preflop). Top pair that plays this way: KJ at 3 × 3 = 9. The screaming draw class: T9 was open-ended on the flop, bricked twice, and neither a ten nor a nine is visible → 16 combos, the full fresh-deck count. Estimate: 16 bluffs over 34 total = 47%, miles over the bar. Call. Total time: about ten seconds, and the exact census would have only sharpened a decision the shortcut already made correctly.

Building the reflex

Three drills, five minutes a day, in increasing order of realism:

1. Flashcard reps. Deal yourself a hand and a flop (or use any hand history). Ask the tier-two questions out loud: combos of the top pocket pair, combos of top-pair-top-kicker, combos of the obvious suited connector. Target: under five seconds each, including the removal adjustment.
2. Batch sweeps. Take one board and one position's range from the site charts and produce only the class totals: "sets: 9, top pair: about 30, flush draws: about 25." No per-combo enumeration allowed — batching rules only. Target: under thirty seconds for the sweep.
3. Big-three gauntlets. Replay your own river decisions from a session and re-make each one with the shortcut: bar, sets, top pair, draw class, fraction, action. Target: fifteen seconds per decision. Then check the closest ones at full precision afterward and see how rarely the answer changes.

Two traps that slow the count

The first trap is trying to name exact suits before the class total matters. If the board is K♣8♦2♠ and you need AK, say 12 first: three kings remain on the board's rank? No — the king is visible, so A K is 4 aces by 3 kings, twelve. Only after the twelve changes the decision should you ask which of those are suited. Most in-game mistakes happen because players reverse that order: they stare at A♥K♥, A♦K♦, A♠K♠, lose three seconds, and forget the much larger offsuit block of the range.

The second trap is counting blockers twice. If you hold A♣K♦, AK is already down to 9. Do not then say "and I block the suited versions too" as if that removes more combos unless you actually enumerate which suited combos died. A♣K♣ and A♦K♦ are dead, but A♥K♥ and A♠K♠ survive; that is exactly why the 9 total splits into 2 suited and 7 offsuit. The anchor math already included the removal. Use suit detail to explain the split, not to discount the same hand a second time.

When you're unsure, fall back to this sequence: identify the class, apply visible-card removal once, then decide whether suits matter. Pair first, unpaired total second, suited subset third. That order keeps your count fast and prevents the two most common errors: over-discounting blockers and letting irrelevant same-suit board cards scare you away from live suited combos.

The honest endpoint of this module isn't that you'll run a 62-combo census in real time — it's that you'll never again need to. The censuses you did slowly taught you what the answers look like; the anchors and batches reproduce them fast; and the big three catches the money decisions. When a number matters at the table now, you can have it before the clock does.