A Better Hand Can Be Quartered More Often

October 4, 2026 by LLOSAT

You flop the nut low in a crowded Omaha hi-lo pot. That feels reassuring, until you remember that another player can have the same low. Should you judge your hand mainly by how likely it is to get quartered?

No. A better hand can collect a quarter of the pot more often and still collect more of the pot overall. Our new Omaha8 simulator gives a useful example. A backup low card saves some hands that would otherwise lose the low half. Some of those saved hands finish with a shared low, so the quartering count rises along with the hand's strength.

This article adds a numerical comparison to our discussions of quartering and scooping. The question is about the cards: what changes when your nut low has backup, and what changes when it also has a nut-flush draw?

Same nut low, three different hands

Six players see this flop:

Compare these three hands. Each currently makes the nut low, 8-7-4-2-A, using A2 from the hand and all three flop cards.

1. Bare A2: A♠ 2♥ 9♦ K♣

You have the best possible low right now, but little help for high. There is no backup low card and no flush draw: holding the ace of spades alone is not enough in Omaha.

2. Backup A23: A♠ 2♥ 3♦ K♣

Replace the nine with a three. If an ace arrives, you can use 2-3 for low; if a two arrives, you can use A-3. That gives you another way to make a strong low when a card duplicates one of your original low ranks.

3. Backup plus a nut-flush draw: A♠ 2♠ 3♦ K♣

Now change the two of hearts to the two of spades. The low ranks stay the same, and you have two spades in your hand to go with two on the flop. Another spade can make an ace-high flush. A paired board can still produce a full house or four of a kind that beats it.

What happened at showdown

We ran 100,000 deals for each hand against five opponents. Every opponent received a uniformly sampled legal four-card hand, and everyone reached showdown. These are random-card opponents, not measured low-limit player ranges. There was no betting, folding or rake.

For this controlled comparison, the opponents and turn-river cards were identical across the three hands. The two cards exchanged out of each hand were reserved as dead cards. That isolates the comparison, but changes the available deck; we also checked each hand with an ordinary remaining deck below.

Swipe the tables sideways to compare all three hands.

Average pot share is the average fraction awarded at showdown. A half-pot finish counts as 50%, a quarter as 25%, and a scoop as 100%. It is often called showdown equity. It is not the chance of winning outright, and it is not betting profit.

Adding the three increased average share by 7.89 percentage points, with an approximate paired 95% interval of 7.73 to 8.05 points. Yet quarter-pot finishes increased by 2.82 points, with an interval of 2.62 to 3.02 points. The stronger hand was quartered more often because its low survived more often.

A quarter can be a rescued result

Suppose an ace comes on the turn. With bare A2 you can still make the original eight-low using the flop, but someone holding 2-3 can now make a better seven-low using the board ace, four and seven. Your low has been counterfeited: it still qualifies, but it is no longer the best possible low.

With A23, you can switch to your own 2-3 and make that seven-low too. If another player also has it, you may share the low half. Taking a quarter in that situation can be much better than losing the low half entirely.

Backup is not a guarantee. Both an ace and a two arriving can leave A23 vulnerable, and a low can still be shared among several players. Count useful backup ranks, then read the new board again.

Here is a fully known quarter-pot example. You hold A♣ 2♦ J♣ T♦, your opponent holds A♦ 2♣ K♥ K♦, and the board is:

Both players make the nut low, 8-4-3-2-A. Your opponent also makes four kings, using two kings from the hand and two from the board. You receive one quarter; the opponent receives three quarters. The simulator verifies those exact fractions. A nut low does not promise half the pot.

High potential changes what a shared low costs you

The last hand had the same low ranks as the second. In the matched deals, their low awards were identical. Its extra value came from high: scoop frequency rose from 1.03% to 22.96%, and average share increased by another 14.01 percentage points. The paired 95% interval for that share increase was 13.94 to 14.08 points.

Winning high while sharing low two ways gives you three quarters instead of one quarter. Winning both halves outright gives you the whole pot. That is why a draw to the nut high is valuable alongside low protection in a multiway pot.

Omaha requires exactly two hole cards and three board cards for each hand. Your high and low can use different pairs. These rules, including the eight-or-better low qualifier, are described in the PokerStars Omaha hi-lo rules.

Did the reserved cards create the conclusion?

We repeated the comparison with no dead cards, using 50,000 deals per hand and a separate seed. Opponents and runouts were sampled from each hand's ordinary remaining deck; these runs were not paired deal for deal.

The precise percentages changed, but the lesson held: backup improved average share while increasing quarter-pot finishes, and the nut-flush draw added substantial scoop potential. The reserved-card numbers should not be treated as ordinary-deck estimates.

Use the lesson at a low-limit table

Look past the words "nut low." Ask which board cards can take away its nut status, whether another low pair in your hand helps, and what you can make for high. A23 with a nut-flush draw deserves a different plan from bare A2 with weak high cards.

Do not let fear of a quarter erase the value of backup. A count of quartered hands leaves out the hands where the weaker holding loses its low altogether. Average share, scoop potential and the strength of both halves give a fuller picture.

Keep the price and the actual players in view. Real opponents choose which hands to play and which draws to continue with. Players who favor A2 can share your low more often than uniformly dealt opponents do. Later bets, raises, folds and rake change the money you win or lose. These showdown percentages cannot tell you whether to call or raise.

Being quartered is not automatically a losing hand either. Your return depends on how much you contributed and how much everyone else put in. In a multiway pot, a quarter can cover your investment; in other circumstances it cannot. Make that money decision from the actual pot and betting action.

Study details and downloads

Poker Simulator 0.15.0, ruleset omaha8-uniform-showdown-v1. Flop: 4s 7s 8d. Hero plus five uniformly dealt opponents; all six reach showdown in one common pot. No betting, rake, folding, opponent ranges, all-ins or side pots. The study comprises 450,000 sampled showdowns, plus a fully known split-pot fixture.

The controlled comparison used five independent seeds, 2026100411 through 2026100415, with 20,000 trials per hand per seed. All three runs exclude As, 2h, 9d, Kc, 3d and 2s from the unknown-card pool: four are held by the hero and the other two are dead. The live deck, unknown slots and random streams therefore match. All 20 saved traces per seed were checked for identical opponent cards and boards. The no-dead-card check used seed 2026100421 and 50,000 trials per hand.

Equity intervals use pooled sample standard deviations and a normal 95% approximation. Difference intervals use paired seed-block means with a Student t interval and four degrees of freedom. Outcome rates in the tables are descriptive. All intervals measure Monte Carlo sampling noise under these fixed assumptions; they do not measure how well random opponents represent your table.

Download the results table (CSV), study summary and paired comparisons (JSON), or complete reproduction package (ZIP). The package contains the exact source inputs, full simulator JSON/CSV reports, audit traces, seeds, commands and a Python script for rerunning the study with engine 0.15.0.