Don't always hold out for the perfect five-card pure sequence. Sometimes melding the short three-card run right now is stronger. The hooks it creates for your future singletons are worth more than the extra length would be.
A sub-optimal meld is one you lay down before it reaches its best possible form. You have 4♠-5♠-6♠-7♠ in hand. The temptation is to hold and hope for the 3♠ or 8♠ to make it a five-card run. The sub-optimal move is to meld the three-card version now (say 4♠-5♠-6♠) and keep the 7♠ in hand as a reserve.
On paper you lose a point of "meld quality." In practice you gain something bigger. A short meld on the table works as a discard hook for the rest of the round.
Every card in your hand that matches either end of that run can now be dumped safely. That's the tactic.
In Rummy the last few turns of a round are dangerous. Any card you can't shed becomes deadwood. Any card you shed into the discard pile might feed an opponent.
A three-card table meld gives you two hooks. Both ends can accept extensions. If you draw a 3♠ later, it goes on the low end. If you draw a 7♠ later, it goes on the high end. You just burned a card for zero risk.
A hand-held four-card run gives you zero hooks. The whole group is stuck in your hand. If you can't finish it in time, all four cards count against you.
Hooks compound. Two three-card melds on the table means four possible discard destinations. That's four cards per round you can shed safely instead of throwing into the pile.
Turn eight. Your hand has 4♠, 5♠, 6♠, 7♠, and seven other cards including some singletons like K♣, 3♣, 8♠.
Option A. Hold everything. Wait for the 3♠ or 8♠ to extend. Deadwood is 22 (4+5+6+7) sitting in hand.
Option B. Meld 4♠-5♠-6♠. Deadwood drops to 7 (just the 7♠). And now the 7♠ is a guaranteed free burn on your next turn (place it on the run). You also just gave yourself a hook for a 3♠ if one arrives.
Option B wins by almost any measure. The only reason to hold is if you're confident the 3♠ and 8♠ are both live and you have three more turns to close.
Three situations always favour the short meld.
Deadwood is high. If your hand is over 25 deadwood, melding anything is a win. The three-card run drops you into the safe zone in one move.
The stock pile is thin. If under 10 cards remain in the stock, waiting for the extension is expensive. Meld now, burn the extra cards over the last few turns.
You have singletons that match the run's ends. If you're holding a 3♠ and an 8♠ alongside the 4♠-5♠-6♠, melding the run first turns both singletons into free extensions. That's two burns for one meld action.
Some situations favour holding for the longer meld.
You're chasing a pure sequence for a first-meld requirement. Some Rummy variants require a specific score or meld shape for your first declare. If a three-card run doesn't qualify but a four-card one does, you have to wait.
Your table position is safe. If you're already melded and your remaining deadwood is low, an extra turn of waiting costs almost nothing. Might as well try for the longer version.
The extension card is highly likely to arrive. If the 8♠ has not been discarded and no opponent has picked from the pile in three turns, the 8♠ is probably in the stock. Hold one more turn.
If you are new, the rule is simple. Meld as soon as it's legal. Never hold a valid three-card group hoping it becomes a four-card group. That's an intermediate move at best, and it's easy to get wrong.
Beginners lose more points from unmelded groups sitting in hand than they gain from extra meld length. Trust the short meld. Move on.
Intermediates make the trade-off explicit. Before deciding to hold or meld, count the singletons in your hand that could extend the run.
Two or more matching singletons: meld the short run. Every singleton becomes a hook.
Zero matching singletons: holding is fine if the extension card is likely.
One matching singleton: usually meld, unless the extension card is nearly certain.
Experts add one more angle. A short meld on the table looks weaker than it is. Opponents see a three-card run and think "that player hasn't closed much yet." They stay loose.
Meanwhile you have hooks that let you shed cards for free. Every turn you extend one, opponents realise too late that you were closer to declaring than they thought.
This is a slow trap. It works best against opponents who read the table for threats. Against loose opponents the effect is smaller because they weren't reading closely anyway.
Waiting "one more turn" for the perfect meld. The perfect meld often never arrives. Meanwhile deadwood piles up. Cut losses. Meld the short version.
Forgetting to extend after melding. The point of the sub-optimal meld is the extensions. If you don't use them, you didn't gain anything. Watch for extension opportunities every turn.
Melding a short run when the long version is one turn away. If you've seen enough discards to know the extension card is live and coming, patience pays. Don't meld reflexively.
Ignoring set opportunities. Sometimes the "sub-optimal" choice is between a three-card run and a three-card set built from the same cards. Compare hooks. Sets have fewer natural hooks than runs.
Play the next five rounds and track every time you meld a three-card group. Note how many times you extended it during the same round.
Most rounds you'll extend at least once. Some rounds you'll extend twice. The hooks pay off almost every time.
If you find you never extend, you're not using the tactic right. The short meld is only valuable if you actively look for extensions on every following turn.
Short melds are not lazy. They are tactical. A three-card run on the table gives you hooks, drops your deadwood, and signals false weakness. Compared to a four-card run held in hand, the trade is almost always worth it.
Stop waiting for the perfect meld. Lay down what you have. Let the extensions come to you.
On Play Rummy, the sub-optimal meld is one of the moves that separates strong regular players from casuals. Build the habit.