Pusoy
The Filipino thirteen-card game set out properly — rules, scoring, naturals, and the numbers behind arranging a hand.
The Filipino thirteen-card game set out properly — rules, scoring, naturals, and the numbers behind arranging a hand.
If you searched for how to play Pusoy and half your results were about Pusoy Dos, that is not your mistake. Google itself cannot reliably separate them: at the time of writing, the top result for how to play pusoy is a Reddit thread posted in a baseball forum, and most of the page-one results underneath it describe the wrong game entirely.
They are two different games that happen to be dealt from the same 52-card deck, thirteen cards each. If it is the shedding game you are after, our Pusoy Dos guide covers its ranking and suit order in full. The card-game reference pagat.com puts it in one line: “In the Phillipines it is known as Pusoy, again not to be confused with Pusoy Dos, which is Big Two.”

| Pusoy | Pusoy Dos | |
|---|---|---|
| What it really is | Chinese Poker (Cantonese Sap Sam Cheung, “13 cards”) | Big Two |
| What you do | Split your 13 cards into three poker hands and compare them | Shed all your cards first by beating the last play |
| Highest card | Ace | The two — which is where the name dos comes from |
| Turns | None. Everyone sets in private, then all hands are exposed at once | Played in turns around the table |
| Where the skill sits | One arrangement decision, taken once, with no information about anyone else | Timing, card counting and knowing when to take control of the pile |
Everything below is about Pusoy.
Four players, one standard 52-card pack, thirteen cards each — the whole deck is dealt, and there are no wild cards. Before the deal, the table agrees what one unit is worth, because Pusoy is scored in units rather than in a pot.
You then split your thirteen cards into three hands, which are placed face down in front of you:
The one hard rule is that back must beat middle, and middle must beat front. Standard poker rankings apply to the two five-card hands. An arrangement that breaks the order is a misset, and a misset is not a clever trick — it is simply not a legal hand.

When everyone is ready, all hands are exposed and every player compares each of their three rows against the corresponding row of every other player. You win a unit for each row you beat and lose one for each row that beats you. Tied rows pay nothing.
That comparison structure is the whole game, and it produces the result beginners find hardest to accept: you can hold the single best five-card hand at the table and still lose money. Pagat’s worked example ends exactly that way — the player holding aces full at the back finishes the deal down a unit, because the other two rows were too thin.
The front hand has only three cards, so only three hand types can exist there: three of a kind, one pair, and high card. And here is the rule that catches almost everyone who arrives from ordinary poker:
Straights and flushes do not count in the front hand. Three cards in sequence are worth nothing. Three cards of one suit are worth nothing. Pagat states it plainly — “there is no value in having a front hand with three consecutive cards or three cards of the same suit” — and the Filipino guides that get this right, such as tongitshub.com, agree: no straights or flushes count in the front.
The practical consequence is worth spelling out. Putting 7♥-8♥-9♥ in the front feels like you have protected the row. You have not. It is a nine-high hand, and it loses to any pair on the table. Those three hearts are almost always doing more work inside a five-card flush at the back or the middle.
Every Pusoy guide tells you to put your strongest cards at the back. That advice is trivially correct and almost useless, because the back row very nearly sets itself. We wanted to know what that instinct actually costs, so we computed it.
Method. Thirteen cards can be split into a five-card back, a five-card middle and a three-card front in exactly 72,072 ways. We dealt 20,000 random hands, evaluated all 72,072 splits of every one of them, discarded the arrangements that break the back ≥ middle ≥ front order, and compared what remained. The front hand was scored with straights and flushes excluded, as the rules require. The headline figures were re-run on three separate seeds and agreed to within a percentage point.
Result 1 — you have less freedom than it looks. Of those 72,072 possible splits, a median of only 22,188 are legal. Roughly two arrangements in three are illegal before you start thinking.
Result 2 — and this is the number that should change how you set. In 78.4% of deals, making your back hand as strong as it can possibly be leaves you with a weaker middle than you could otherwise have held. In 46.2% of deals it costs you an entire hand category in the middle — most often a two pair collapsing to a single pair.

Put next to each other, the two middle-row distributions make the cost visible. The left column is the best middle hand you could hold if you set for it; the right column is what you are left with once the back row has taken first pick.
| Middle hand | Best you could set | What is left after maximising the back |
|---|---|---|
| Full house or better | 3.1% | 2.6% |
| Flush | 11.2% | 6.7% |
| Straight | 19.1% | 10.8% |
| Three of a kind | 8.3% | 2.6% |
| Two pair | 50.8% | 33.6% |
| One pair | 7.5% | 33.8% |
| High card | 0% | 10.0% |
Read the bottom two rows together. Set for the middle and you are almost never stuck with a bare pair or worse — it happens 7.5% of the time. Set for the back and you are stuck with a bare pair or worse 43.8% of the time. That is the trade you are making, every hand, usually without noticing.
None of this argues for a weak back row. It argues that the back row should be strong enough rather than as strong as possible — because a full house at the back beats exactly the same opponents that a flush at the back beats, most of the time, while the two cards you freed up may be the difference between winning the middle and losing it.
The same 20,000 deals give a sense of what you should expect to be able to make in each position. These are the strongest hands available for that row on its own — they cannot all be achieved at the same time, which is precisely the problem.
| Row | Most common best-available hand | Worth knowing |
|---|---|---|
| Back | Full house 44.6%, flush 33.9%, straight 11.1% | A back row weaker than two pair is essentially impossible — it happened in none of the 20,000 deals. |
| Middle | Two pair 50.8%, straight 19.1%, flush 11.2% | Two pair is the middle-row par score. A straight in the middle is a strong hand, not an ordinary one. |
| Front | One pair 94.2%, high card 4.8%, trips 1.0% | Nearly everyone can put a pair in front, so a pair wins nothing by itself — the rank of the pair is what decides it. |
That last line is the quiet one. Because 94% of players can field a front pair, the front row is usually decided by whether yours is a pair of tens or a pair of fours — a difference of one card that most beginners give away without a thought.
The simplest system is one unit per row won, settled between every pair of players. Most tables then add extra payments for strong hands in particular positions. A typical schedule:
| Hand and position | Pays |
|---|---|
| Three of a kind in the front | 3 units instead of 1 |
| Full house in the middle | 2 units instead of 1 |
| Four of a kind in the back or middle | 4 units instead of 1 |
| Straight flush or royal flush in the back or middle | 5 units instead of 1 |
Two details about that table decide real money, and both need agreeing before the cards come out.
First: in the standard schedule these bonuses only pay on rows you actually win. A straight flush in the middle that loses to a better straight flush pays you nothing at all. There is a widely played variant in which the extra payments count whether the row wins or loses — and under that variant the correct way to set a hand changes, because a monster row becomes worth something even when it is beaten.
Second: the front three of a kind is the best-priced hand in the game. It pays triple, and our deals say you can only make one about 1% of the time. If trips are your only realistic route to a front-row win, that is usually where they belong — provided the remaining ten cards still make a legal set.
Common additions include a bonus of 1 or 3 units for a sweep (winning all three rows against one opponent), and the overall point, where the net winner of each pairing takes one extra unit. The overall point matters more than it sounds: it rewards winning two rows comfortably over winning one row spectacularly, which pushes correct play further towards balance.
Some thirteen-card holdings win automatically, without being set into three rows, provided you declare them before the hands are exposed. Every guide lists them. Almost none says how often they turn up, so we worked it out — the first three exactly, from the combinatorics, and three straights from a two-million-hand simulation because overlapping sequences make the exact count awkward.
| Natural | What it is | Typically pays | How often you get one |
|---|---|---|---|
| Three flushes | Back and middle both flushes, front three of one suit | 3 units | 1 in 112 |
| Six pairs | Six pairs and one odd card | 3 units | 1 in 283 |
| Three straights | Back and middle both straights, front three in sequence | 3 units | about 1 in 310 |
| Complete straight (the Dragon) | One card of every rank, ace through king, suits irrelevant | 13 units | 1 in 9,462 |
Line those numbers up against the payouts and something jumps out that we have not seen stated anywhere else: the three 3-unit naturals are nothing like equally rare. Three flushes arrives roughly once every 112 hands. Three straights is close to three times rarer, at about one in 310. They pay the same. If your table uses the flat 3-unit schedule, it is paying the easiest natural in the game exactly what it pays one you may go a year without seeing — which is worth knowing before you agree to it, and is presumably why some schedules, such as the Malayan one pagat records, pay three straights more than three flushes.
One subtlety that separates experienced players from beginners: you are not obliged to declare a natural. A three-flush hand declared wins a flat 3 units. The same thirteen cards set normally will often win all three rows anyway, and can pick up bonus payments on top — two flushes in the back and middle are strong rows. Declaring is the safe play; setting is frequently the bigger one. The Dragon is the exception. Take the 13 units.
A misset — sometimes called a fouled hand — is an arrangement where the required order does not hold, for example a middle hand that beats the back. Pagat’s rules do not attach a price to it, because they treat an illegal arrangement as simply not permitted. In practice, at a table where the cards are already down, the near-universal treatment is that the offending player loses all three rows to every opponent, and some published schedules add a fixed penalty on top of that. Settle which one your table uses before you deal, not afterwards.
The good news from the simulation is that a misset is always avoidable: across 20,000 random deals, every single hand had at least one legal arrangement, and the typical hand had more than twenty thousand of them. A misset is never bad luck. It is always a mistake, and usually a rushed one.
Surrender is offered at some tables. Before the hands are exposed, a player with a hopeless holding may fold and pay each opponent as though they had lost two rows out of three — one unit each, or two under the overall-point variant. Casino schedules can be harsher; at the Hawaiian Gardens casino version pagat documents, surrender costs 3 units to each opponent. Surrendering also gives up any chance of a bonus payment, so it is a damage-control move, not a strategy.
If you learned Pusoy in the Philippines, the scoring above may not be what you grew up with. Pagat records a version played there in which a deal has at most one winner: to win, you must beat every other player on at least two of the three rows, and if nobody manages that, nobody wins the deal.
That single change transforms the game. Under standard scoring you can profit from a good hand by picking off one weak opponent. Under single-winner scoring, a row that is merely decent against three different players is worth nothing at all, and consistency across all three rows becomes the only thing that pays. If you are playing this version, the balance argument from our numbers above stops being a marginal edge and becomes the whole strategy.
Other regional schedules differ too — the Singapore version pays a straight flush in the middle 14 units, against 5 in the common schedule, and Vietnamese Xập Xám doubles or triples payments for sweeping every opponent. There is no single official rulebook for Pusoy. Agree the schedule out loud before the first deal, because the correct way to set a hand genuinely changes depending on which one you are playing.
If you enjoy Pusoy, Tongits is the other Filipino card game worth learning — a three-player rummy variant that is, if anything, more widely played, and our Tongits strategy guide gives it the same numerical treatment we have given the setting problem here. For the table games you will actually find in a casino lobby, start with our online casino games guide.
Away from the card table, the Filipino games we have covered are mostly ones where the price is fixed rather than played for. The color game and Sakla are the two perya stalls whose house edge can be proved exactly, Jueteng is the illegal numbers game whose odds are set by the very statute that bans it, and sabong is the cockpit betting sport whose online form has been suspended nationwide by executive order.
Pusoy is overwhelmingly a social game — played at home, at family gatherings and in dedicated Filipino card-game apps, usually for small stakes or none at all. It is not a standard online-casino table game the way blackjack or baccarat are, and you will not find it in most casino lobbies. Chinese Poker does appear in a handful of specialist card rooms, and the open-face variant has its own following, but if you are searching for a place to play Pusoy specifically, a card-game app is the realistic answer rather than a casino.
Rules and gambling law vary considerably from one country to another, so check what applies where you are before playing for money. If you do play for stakes, agree them in advance, treat them as the cost of an evening rather than an investment, and stop when you said you would. Support is available at GambleAware.org. Play only if you are of legal age where you live.
After going through several successful experiences with several industry pioneers, Fadi Hayek has managed to excel in the world of content for online gaming by providing unique content for several projects that allowed him to obtain the necessary experience to launch his career. Last but not least, Fadi also holds several certificates, including but not limited to content marketing and iGaming.