Reopening a Week 1 folk rule#

Week 1 named “use the rule of 4 and 2 to estimate your equity on a draw” as a claim to be cashed in later. This is that week.

The rule of 4 and 2#

A fast mental-math shortcut for equity on a drawing hand:

• Two cards to come (you’re on the flop): outs × 4 ≈ equity %
• One card to come (you’re on the turn): outs × 2 ≈ equity %

No hypergeometric distribution required — just count outs and multiply.

Why a shortcut at all#

Exact equity needs a hypergeometric calculation most players can’t run at the table in real time. The rule of 4 and 2 trades precision for speed.

The question this deck answers: how much precision, exactly?

The flagship hand#

Hero: A♠K♠. Board (turn): 5♠ 9♠ 2♥ Q♦.

Two spades in hand, two on board — a nut flush draw, no pair, one card (the river) to come.

Counting the outs#

13 spades total, minus the 4 already visible (2 in hand, 2 on board):

9 outs.

Rule of 2 vs exact, one card to come#

The rule underestimates here by about 1.6 percentage points.

Rule of 4 vs exact, two cards to come#

Same 9-out draw, but caught on the flop instead of the turn:

Now the rule overestimates, by about 1.0 percentage point — and the error grows with the out count:

Outs, by draw type#

The flagship hand is one entry in a small catalog. Same counting method, different shapes:

Gutshot: 4 outs#

Hero Q♣J♦, board 9♥8♠2♣. Only a Ten completes 8-9-T-J-Q — one rank, four cards, none visible: 4 outs.

Open-ended straight draw: 8 outs#

Hero 9♦8♣, board 7♠6♥2♦. A 5 or a Ten completes the straight from either end — two ranks, eight cards: 8 outs.

Combo draw: 15 outs#

Hero J♠T♠, board 9♠8♥2♠. A flush draw (9 spade outs) and an open-ended straight draw (Q or 7, 8 cards) — but the Q♠ and 7♠ are each already counted as flush outs. Union, not sum:

9 + (8 − 2 overlap) = 15 outs, not 17.

The miscount that’s easy to make#

Back to the flagship hand: A♠K♠ on 5♠9♠2♥Q♦. It’s tempting to count outs straight off the board:

13 spades − 2 seen on the board = 11 outs.

Why that’s wrong#

Your own hole cards remove supply from the deck too, not just the board. Hero is holding two of those thirteen spades:

13 − 2 (board) − 2 (hero’s own hand) = 9 outs, not 11.

What the miscount costs#

A 4.3-percentage-point overestimate, purely from forgetting to remove your own cards from the count. This is card removal doing its damage silently, inside a step that looks like simple arithmetic.

Exact hypergeometric equity: the formula#

For outs cards among unseen remaining, in draws cards still to come:

P(at least one out) = 1 − C(unseen − outs, draws) / C(unseen, draws)

The complement — the probability of missing every out — is easier to count directly, so the formula counts that and subtracts from 1.

Plugging in the flagship hand#

outs = 9, unseen = 46 (52 minus 6 known cards), draws = 1:

P = 1 − C(37, 1) / C(46, 1) = 1 − 37/46 = 9/46 ≈ 19.6%

The pot-odds threshold#

To decide whether a call is profitable, compare your equity against the equity a call needs to break even:

required equity = call / (pot + call)

pot here already includes whatever the opponent just bet.

The flagship’s numbers#

Pot is $100 (including villain’s bet). Hero must call $50 to see the river:

required equity = 50 / (100 + 50) = 50 / 150 = 33.3%

What implied odds add#

If hero expects to win extra money on later streets when the draw hits, that expected extra money counts toward the pot in the threshold formula:

required equity = call / (pot + call + expected future winnings)

The flagship, with $60 implied#

required equity = 50 / (100 + 50 + 60) = 50 / 210 ≈ 23.8%

Still above the flagship’s 19.6% equity. Implied odds narrow the gap here — they don’t close it.

How much would it take?#

Solving for the implied winnings that make the call break even:

50 / (150 + X) = 9/46 → X ≈ $106

That’s the concrete size of the implied-odds argument this call would actually need — not “probably enough,” a number.

Reverse implied odds#

The mirror problem: sometimes hitting your draw doesn’t win the pot — a bigger flush, a redraw you didn’t see. That risk only pushes the real required equity higher, never lower. This deck’s numbers so far assume it away; a careful player doesn’t.

The hand, from the top#

Flop: 5♠9♠2♥. Hero calls a small bet holding A♠K♠. Turn: Q♦, no help to the board’s texture. Villain bets $50, making the pot $100.

An informal range read#

Villain’s sizing and the line so far look more like top pair or an overpair than a slowplayed set or two pair — a rough class, not a distribution. Week 6 formalizes how to turn a betting line like this one into an actual range.

The numbers, side by side#

Pricing the decision in EV#

EV(call) = equity × pot − (1 − equity) × call

This turns “fold” from a verdict into a dollar figure — what calling is worth on average, not just whether it clears a threshold.

The flagship’s EV#

equity = 9/46, pot = $100, call = $50:

EV(call) = (9/46 × 100) − (37/46 × 50) = −$20.65

What the folk rule would have priced it at#

Using the rule-of-2 estimate (18%) instead of the exact figure:

EV(call) = (0.18 × 100) − (0.82 × 50) = −$23.00

Same fold either way here — but the folk rule misprices the decision by about $2.35, and the error table earlier showed that gap only grows on bigger draws.