Advanced Strategy

The rules of 42 are simple; playing it well isn't. Here's how to think about each phase of a hand once you've got the basics down.

Evaluating a hand before you bid

Before you bid, look at your hand suit by suit and ask: if I named this trump, how many of the 7 dominoes in that suit would I hold, and how safe are the rest of my tiles once it's trump?

• Trump length. Four or more dominoes in a suit is a strong trump hand; five or more can carry a hand almost by itself.
• The trump double. Holding the double of your trump suit matters a lot -- it's the one domino nobody can beat, and it lets you draw out the other trumps in play without risking anything.
• Count protection. A count domino (5-5, 6-4, 5-0, 4-1, 3-2) is a liability if it's stuck off-suit in a hand with few trumps -- you may be forced to give it up cheaply. The same count domino sitting inside a strong trump suit is an asset instead, since you control when it's played.
• Off-suit doubles. A double outside your trump suit is still useful: it's the strongest domino of its own suit, so it wins a lead there unless someone's void in that suit and trumps in.

Add it up and you get a rough sense of how many tricks and how much count you can realistically expect to win. Bid to that number, not to what you'd need to win to feel good about it -- overbidding and getting set costs you marks outright, while a modest bid you comfortably make is never a bad outcome.

Your offs carry directional risk

A domino that's neither trump nor a double isn't just "not a winner" -- it's a specific liability, because a non-double, non-trump domino always leads as its higher pip. So the risk it carries depends entirely on which suit that is. Say your trump is fours and you're holding the 3-3 double alongside two offs that both lead as fives, but you don't hold 5-5. The moment one of those has to be led or sluffed, whoever holds 5-5 can run you for the trick plus 10 count -- 11 points -- and if 5-0 is also out there, that suit alone is worth 16 points against you. Two hands with the same trump length and doubles can carry very different amounts of real risk depending on what their offs point at, and that's worth weighing before you commit to a bid or decide to take one yourself instead of leaving it for your partner.

Choosing trump

Trump length usually beats raw pip value. A suit where you hold five dominoes including the double is normally a better trump than a suit where you hold three high-value dominoes -- length gives you control of the hand for longer and lets you flush out the opponents' trumps before they can use them defensively.

Low suits -- blanks, aces, deuces -- make weaker trump for a different reason: a domino's rank within a suit is set by its other pip, so every domino paired with a blank is the worst member of whichever other suit it belongs to, every ace-paired domino is second-worst, every deuce-paired domino third-worst. Outside of that suit actually being named trump, those tiles are close to dead weight. That cuts the other way too: if you're holding a decent, playable low suit, the only way to make it worth anything is to name it trump yourself, so bid it a little more assertively than the same length in a higher suit would call for -- passing and hoping to support a partner's bid wastes the one thing that hand is actually good for.

No Trump: when to skip a suit entirely

Naming a suit isn't the only option once you win the bid -- you can declare No Trump instead, and with nothing able to trump in, the single biggest risk to a doubles-heavy hand disappears completely. Under a normal trump contract, leading a double you're proud of can still get run over by an opponent who has none of that suit and trumps in over it. Under No Trump, that can't happen -- your double is simply the best domino of its suit, full stop.

Three situations tend to call for it:

• A lot of doubles, ideally with backup. Three or four doubles across different suits is the classic shape. It's even better when you also hold the next-highest domino of one of those suits -- lead the double first, and your follow-up is now the highest survivor, extending your control into a second trick from the same suit.
• Your partner has SOH'd. If they've shown support and you're sitting on doubles of your own, No Trump lets you lead straight into them with no risk of either of you getting trumped along the way.
• Your doubles can pull count on their own. Holding the doubles of suits that also carry count -- say 5-5, 6-6, and 4-4 -- lets you lead each one in turn. Whoever's forced to follow suit (or has nothing else to sluff) may have to give up 5-0, 6-4, or 4-1 right into your winning trick. Three tricks like that can realistically be worth 30+ points on their own, and none of it was ever at risk of being trumped away.

The trade-off: your non-double tiles get no special protection under No Trump the way they might behind a long trump suit, so it's a much better fit for a hand that's genuinely carried by its doubles than for a merely-decent hand that happens to have two of them.

The bid ladder

A handful of bid values aren't arbitrary -- they're breakpoints in how much you can afford to lose. A bid of N leaves you a margin of 42 - N points before you're set, and that margin lines up with real thresholds:

Bidding one of these clean values is a direct claim on your own hand. Everything else -- 32, 33, 34, 37, 38, and so on -- doesn't line up with a breakpoint, which is exactly what makes those values useful for a different purpose: signaling. That's the whole subject of the next section.

Deciding whether to bid at all

Passing is a valid, and often correct, decision. A hand with no suit longer than three dominoes and no double to anchor it rarely makes a good bid -- you're much better off passing and letting someone else take the risk than forcing a marginal bid and handing the other team marks when it fails.

There are really two kinds of bid: the one you want to take and play yourself, and the one that tells your partner how strong your hand is (see SOH below). Before you bid to take it, your hand should clear two bars: at least three trumps in the suit you'd name, and at least four dominoes that can win a trick. Count winners honestly:

• Trumps that end up on top. Every trump ranked above yours that's still out costs you one of your lower trumps to draw it. With 3s trump and 3-4, 3-5, 3-6 in hand, leading the 3-4 loses to the 3-3 -- but now the 3-6 and 3-5 are the two best trumps left. That's two winners from three trumps, not three.
• Off doubles, and the unbroken run directly under them in that suit. Holding 6-6 and 6-5 is two winners; 6-6 and 6-3 is just one, since the 6-5 and 6-4 still sit between them.

Strength of Hand (SOH) bidding

Bidding by more than you strictly need to can be a way of inviting your partner to keep going, rather than a claim that you want to win the bid yourself. If you're holding good support -- off-suit doubles that are nearly always the strongest of their suit, or count dominoes you can protect instead of being forced to give away -- that support is only worth anything once someone wins the bid and names trump. A hand with three or four doubles, or somewhere around 20 points of count, or a mix of both, is often better spent backing a partner's bid than winning a modest bid outright yourself. Raising a little further than the bidding strictly requires is one way to let your partner know that pushing higher isn't as risky as it looks, without committing to anything yourself -- you can still end up passing.

This works any time your partner hasn't bid yet -- including when you're the very first to bid. There's no rule that says SOH only applies to raises; an opening bid that doesn't look like a plain claim is just as surprising, and just as much an invitation, as an odd raise later in the round.

This only works as a nudge, not a decoder ring, and it shouldn't be read as one. There's no fixed number that reliably means a specific hand -- what actually carries the signal is the surprise of the raise itself. Beating the previous bid by exactly 1 is completely unremarkable; it's what everyone does just to stay in the bidding, and it says nothing at all. So the floor for an SOH bid is always the reference bid plus at least 2 -- if the previous bid is 31 and you want to signal support, 33 is the minimum; 32 is just the cheapest possible raise and reads as nothing. The same logic applies when opening: 30 is the unremarkable baseline, so 32 is the floor for an opening signal. How far past that floor you go is a rough, situational read on how much support you have, not an exact count of your dominoes -- 32 might mean a little count and a lead or two, while 34 says you're sitting on three or four strong doubles. Don't go much past that, though: your partner has to bid over your signal to back it, so a jump to 38 leaves them taking the hand at 39 or more, which means losing four points or fewer. Over thousands of simulated hands, contracts reached that way made well under half the time, while signals kept in the 32-34 range (backed at 35 or 36) did much better. And a raise should never land on one of the ladder's clean values (31, 35, 36) -- if it would, push one further, since a bid that matches a real claim doesn't read as a signal anymore.

Responding to a partner's SOH bid

Hearing the signal is only half of it -- what you do with it matters just as much. If your partner has bid an SOH value and it's your turn, the default is to back them up, not to fall back on judging the bid purely by your own hand. They didn't put that number out there to stand on its own; they put it out there because they're counting on you. Passing should be rare, and it's only really justified when you have nothing to offer at all -- no suit better than two trumps anywhere in your hand. Two trumps is never enough to bid on, signal or not, since it can't sustain control of a hand no matter how good those two tiles are individually -- but anything better than that, you owe it to the bid to at least take it up, even if your own hand alone wouldn't have gotten you there.

How far you push depends on what you're holding. If the signal was in the strong zone (37-41) and you've also got a near-certain sweep in your own hand, that's corroborating evidence on top of your partner's word -- go to marks. Be honest about "near-certain" here: a failed marks bid costs double, and hands that look like they'll sweep get caught for one trick far more often than they seem to. Otherwise, overbid as far as your hand reasonably supports and trust the partnership: even taking it at the bare legal minimum beats leaving your partner's signal to die on the table and get set immediately, which is the single worst outcome an SOH exchange can produce.

Leading tricks

As the bid winner (or whenever the lead falls to you), your first few plays set the tone for the hand:

• Lead trump early if you're long in it. Drawing out the opponents' trumps while you still have plenty left removes their ability to ambush your later count dominoes.
• Lead a double if you're short in trump. An off-suit double is the strongest domino of its own suit, so it wins unless someone's void there and trumps in -- and either way, it buys you information about who's out of that suit.
• Don't lead count blind. Leading a count domino into a suit you don't control is how you hand the other team easy points -- lead your low, safe dominoes first and save count for when you know it's protected.
• Draw out the blocker first. Play a domino that's worth nothing to pull out the one that would beat your count domino, so your count becomes the best one left. With 4s trump, never lead the 6-4 without the 4-4 -- it's worth 10 and walks straight into the double. Lead the 4-2 instead: whoever holds the 4-4 has to spend it on a 1-point trick, and now your 6-4 is the boss of the suit. The same goes for every suit, trump or not.
• Walk the last one. Once every higher domino of a suit has been played, the one you're holding can't be beaten in that suit -- lead it whenever you need the lead or a trick.

Following and discarding

When you can't win a trick, how you lose it matters:

• If your partner is already winning the trick and nothing left to act can beat it (you're last to play, or everyone still to act is already known out of both the suit and trump), load it up with count instead -- that trick is banked, so it's the safest place to spend it. If there's still a real chance it gets taken away, throw your least useful domino instead and hold your count for a safer moment.
• If you need to win the trick, play the cheapest domino that actually wins it. Save your strongest trumps for when you need them, not for tricks you could have taken more economically.
• Never throw count on a trick you know is lost. If the domino on top can't be beaten by you or anyone on your side still to play -- you can tell from what's been played and who's shown out of what -- keep your count and throw something worthless. A count domino you hold onto might still land on one of your side's tricks later; one you throw on a lost trick is gone for certain.
• On defense, push points. The bidder can only be set by points, so whenever your side might still win the trick, get your count onto it. Hoarding count on defense mostly helps the bidder.
• Cover your count. Keep a worthless domino of the same suit next to each count domino (a 4-2 beside the 4-1, say), so that when the bidder leads to pull it you can follow with the cover and choose when the count goes in yourself.
• The 5-5 is a winner, not just points. It wins its own trick once trumps are gone, so don't rush to hand it to your partner. Give it up when its 10 points make the bid, or when you know your partner will take the trick.

Track what's been played

Good 42 players keep a running count of which trumps and which count dominoes are still out. Once you know every trump has been played, your remaining trump-suit dominoes are guaranteed winners; once you know both ten-count dominoes are gone, you can stop worrying about protecting against them. This kind of bookkeeping is what separates a hand played on instinct from one played with real control.

Track who as well as what. A player who can't follow a suit has just told the whole table they're out of it, and the bidding tells you more: whoever named trump almost certainly holds three or more of them, and a partner who signaled support holds doubles. The computer players keep exactly this ledger and reason from nothing else -- they never see a hand they shouldn't.

Marks bids: know the risk

A marks bid asks you to win every trick in the hand -- one lost trick and the whole bid fails, at a cost of that many marks instead of a benefit. It's rarely correct with anything less than an overwhelming trump suit (five-plus dominoes including the double) and a hand with almost nothing exposed off-suit. Treat it as a rare, high-conviction play, not a routine escalation.