A safety play in bridge is a line of play that deliberately gives up a chance at an extra trick in exchange for a better chance of making the contract. You pay a small price, usually an overtrick, to protect against the layout that would otherwise beat you. The hard part is recognizing the moment, because a safety play never feels heroic. It feels like leaving money on the table.
I wrote about the bigger strategic tension in Safety Play vs Greed: Choosing the Right Line. This one is more hands-on. Two deals, what the safe line looks like, and what it costs you.
Start by asking what you can afford to lose
Every safety play begins with the same question, asked before you touch a card from dummy: how many tricks do I actually need from this suit?
Most players skip it. They see a finesse and take it, because finesses are what you do with an ace-queen. But if the contract needs only one trick from that suit, and you already have it in your hand, the finesse is not a plan. It's a gamble on a result you don't need.
Hand one: take the sure trick
You're in 3NT, not vulnerable. West leads a heart and you hold up your king once, so West's hearts are now set up and you have no second stopper. Counting, you find eight tricks outside diamonds. In diamonds, dummy has ♦A Q 3 and you have ♦7 5 2.
The finesse is tempting. If West has the king, the queen wins and you make ten tricks. If East has it, East wins and returns a heart through your stopper, West runs the suit, and you're down.
The safety play is embarrassingly simple. Cash the ace. Nine tricks, guaranteed, no matter where the king lives.
What did it cost? Put numbers on it. Making 3NT non-vulnerable is +400; with an overtrick it's +430. Going down one is minus 50. At IMPs, the overtrick you gave up is worth 1 IMP. The disaster you avoided is worth 10. Easy trade.
Hand two: if it loses, make it lose to the right player
This one is a close cousin of the safety play, usually called an avoidance play. Same idea, though: arrange things so the bad outcome is survivable.
You're in 3NT again. West has led spades, you've held up and won the third round, and East showed out on it. So West has long spades and an entry would let them cash out. East is the safe hand; West is the danger hand.
Clubs are your source of extra tricks. Dummy has ♣K J 10 3, you have ♣A 9 5 2, and the queen is missing. You can finesse for it in either direction. Which way?
Cash your ace, then lead low toward dummy's K J 10. If West plays low, put in the ten. If this finesse loses, it loses to East, and East has no spade to return. You're still in control. Finesse the other way (king first, then low toward your A 9) and a losing finesse hands the lead to West, who promptly runs the spades.
Both directions are roughly a coin flip on the queen. Only one of them turns a wrong guess into a disaster. Pick the one that doesn't.
When a safety play is the wrong play
The same deal can call for a safety play at one table and punish it at another, depending on the scoring.
At IMPs, as hand one showed, overtricks are nearly worthless and making the contract is everything. Safety plays are almost always right there. At matchpoints, though, every extra 30 points can move you from below average to a top. If the whole room is going to make 3NT, cashing the ace for exactly nine tricks might score worse than half the field who took the finesse and made ten. IMPs vs Matchpoints goes deeper into why the same card can be right in one format and wrong in the other.
There's also a middle case. If you're in a contract nobody else will bid (a thin game, say, or a slam the field stops short of), just making it is already a great score. Protect it. Greed is for contracts everyone is in.
The part nobody tells you: you can't judge it alone
Safety plays have an awkward feature. When you cash the ace and the king was onside all along, it feels like you blew a trick. When you take the finesse and it works, it feels brilliant. Your own table will lie to you in both directions, which is the argument I made in Why "Making Your Contract" Isn't Always Enough.
Whether the safety play was right depends on what everyone else holding your cards did. Did the field go down in 3NT while you made it? Then your boring ace was the best play in the room. Did everyone make ten? Then maybe you were too careful for the format. At a kitchen table that question usually never gets answered. That's the gap Bridge@Home fills: your foursome plays pre-dealt hands and gets scored against other players who held the same cards, so you find out whether your caution cost you or saved you.
The takeaway
Before you play to trick one, count what you need from each suit, then ask which layout could beat you. If a small sacrifice removes that layout, and the scoring doesn't punish the lost overtrick, make the sacrifice. It will feel timid. Do it anyway, and then go look at how the other tables did.