Board 21, somewhere past ten o'clock. You're declaring 4♠ and you've drawn two rounds of trumps. Or was it three? You can't quite remember whether the jack has shown up, so you play another one just in case. Nothing lost. Two boards later, the same fog costs you a contract.
Most bridge players can tell you about the brilliant hand they misplayed. Far fewer can tell you about the six ordinary hands they played slightly worse after the coffee ran out, because those never feel like mistakes. They feel like the cards.
What fatigue actually looks like at the table
Tired bridge is rarely dramatic. Nobody revokes on purpose or bids 7NT on a whim. The damage comes in small, quiet leaks. You stop counting the hand and start "sort of" knowing where the cards are. You play to trick one before making a plan, because the plan felt obvious. You take the finesse without checking whether the drop was better, or you pitch the card partner needed to see.
Each of those costs maybe a trick, and only some of the time. That's exactly why they matter. At matchpoints a single overtrick can separate a top from a middling score, and the pairs who win events are rarely the ones with the flashiest plays. They're the ones who don't leak. (There's a whole post on why consistency wins duplicate events, and fatigue is the biggest enemy of consistency I know of.)
Why you usually can't tell it's happening
Your brain is a lousy judge of its own sharpness. By the late boards you feel fine. You might even feel looser and more confident than you did at eight o'clock, which is its own warning sign.
At a single home table there's nothing to check that feeling against. You go down in 3NT on board 22 and it looks like a bad hand. Maybe it was. Or maybe four other tables made it because they counted the diamonds and you didn't. Without the other results you'll never know which story is true, and the default story (bad luck) is always the comfortable one.
This is where comparing results earns its keep. When your scores sit next to everyone else who played the same deals, a pattern can show up that no amount of honest self-reflection would reveal: solid through board 12, then a slow, steady slide. It's surprisingly common for a player convinced they held rotten cards all evening to find their first half was well above average and their second half handed it all back. Tools like Bridge@Home bring that kind of board-by-board comparison to a home game, and the late-session slump tends to be one of the first things it exposes.
Habits that keep the late boards sharp
None of this takes heroic willpower. It takes routines that keep running when you're tired, which is sort of the whole point of a routine.
Make the trick-one pause non-negotiable. Count your winners or losers before you call a card from dummy, every hand, including the easy ones. Easy hands are where tired players bleed. The case for planning before trick one holds all night, but it doubles in weight after board 15.
Count one suit, always. On nights when you can't track the whole hand, count trumps. Just trumps. It's the count that saves the most contracts, and it's cheap enough to survive a long evening. In notrump, pick declarer's long suit and count that instead.
Leave the last board at the last table. Replaying a disaster in your head while the next auction is happening is the fastest way to create a second disaster. Scribble a note on the scorecard if you want to talk about it later. Then let it go. Recovering after a bad board is a skill of its own, and most of it is simply refusing to carry the previous hand forward.
Choose simpler lines as the evening goes on. A plan you can execute at 70% sharpness beats a clever one you'll botch at trick nine. That isn't cowardice. It's knowing your own condition and playing to it.
Get up. Between rounds, stand, walk to the kitchen, drink a glass of water. Two minutes away from the table resets more than you'd think. And yes, the second glass of wine is a bidding decision too.
Maybe the session is just too long
Sometimes the fix isn't better habits. It's fewer boards.
Club sessions run 24 to 27 boards because that's what the movement needs, and serious players build stamina for it over years. A home game has no such excuse. If your group reliably falls apart after board 16, play 16 and finish with everyone still thinking clearly. Nobody earns anything for finishing more hands badly, and a shorter night that ends on a sharp, close result is a better night than a long one that dissolves into guesswork and yawning.
Pay attention to when the chatter changes, too. When the post-mortems get shorter and somebody starts checking the time, the table has already told you where the session should have ended.
The takeaway
Focus at the bridge table isn't a talent you either have or lack. It drains at a fairly predictable rate, and it drains quietly, one uncounted trump at a time. Build two or three habits that work on autopilot, cut the session to a length your group can genuinely handle, and then look at the evidence. If your late boards look worse than your early ones against players who held the same cards, you haven't found bad luck. You've found the easiest improvement available to you.