There is a Tuesday, somewhere around your first year of training, when the weight you added last week does not move. Not slower. Not uglier. It does not move. You rack it, you sit down, and you decide the program is broken.
The program is not broken. Linear progression is. The two are easy to confuse, because for the first eight to eighteen months they are the same thing. That is the whole logic of a good novice program: you add weight every session, the bar obeys, and you mistake the obedience for a law of nature. It was never a law. It was a runway, and you have reached the end of it.
The reflex at this point is to find a new program. The better move is to change one rule. Stop adding weight every session. Start adding reps.
That rule has a name. Double progression. You progress two variables instead of one, but never both at the same time. You hold the weight and chase reps. When the reps arrive, you add weight and the reps reset. Two ladders, climbed one rung at a time, alternating.
What it looks like on the page
Pick a weight and a rep range. Say you put squats at 3x6-8. That notation means three sets, and your job is somewhere between six and eight reps per set.
Week one you get 8, 7, 6. You do not touch the weight. Week two you get 8, 8, 7. Still the same weight. Week three you get 8, 8, 8. Now every set has hit the top of the range, so the weight goes up, and next week you are back near the bottom: 7, 6, 6 at the heavier load. Then you climb again.
Linear progression moves the load and holds the reps. Double progression holds the load and moves the reps, then moves the load once. That single inversion is the whole idea. Everything else is detail.
The detail that matters most is the one nobody writes down.
The range is a decision, not a default
Most guides hand you 3x8-12 and call it done. That number is not a law either. It is a fatigue decision disguised as a default, and the disguise costs people progress.
A six-to-eight range keeps you near the strength end. The loads stay heavy enough that the movement still trains the movement, not just the muscle. A ten-to-fifteen range buys you more reps before you have to add weight, which is useful on a movement where adding weight is awkward or where your joints have opinions. Wider ranges hide more fatigue and forgive more bad days. Narrower ranges keep you honest and keep the bar heavy.
I program tighter ranges than most coaches. On a main lift for a client past their first year I rarely go above eight, and often sit at 3x4-6. On accessory work where the point is tissue and not load, I will open it up to twelve or fifteen and stop caring about the exact number. The range is the dial. You set it by asking what the exercise is for, not by copying what the last article told you.
So before you load the bar, answer one question. Is this lift here to build strength, or to build the muscle that feeds a different lift? The answer picks the range. The range picks how often you add weight. None of that should be automatic.
The failure mode lives in the reps you add
Here is where double progression quietly breaks, and where almost no one looks.
The promise of the method is that reps are free progress. They are not. A rep is only progress if it is the same quality as the rep before it. The moment you start grinding, the bar slows, the brace softens, the bottom position gets sloppy, and you are no longer adding reps to the same exercise. You are adding a worse exercise on top of a good one and calling the total an improvement.
I tell clients to count reps the way a referee counts them. If the rep would not pass on a platform, it does not pass in your log. A squat that cuts depth by two inches under fatigue is not the eighth rep of your set. It is the first rep of a different, easier movement. Write down seven.
This is the difference between double progression done honestly and double progression done as bookkeeping. The bookkeeping version chases the number on the page. Set after set creeps up by a rep, the log looks like progress, and the lifter cannot understand why the eventual weight jump feels like a wall. It feels like a wall because the reps that earned it were counterfeit. The load went up against reps that were never really there.
Bar speed is the tell. On your working sets, the early reps should move at a pace you would call brisk. As you approach the top of the range, the last rep is allowed to be a fight, but it should still look like the same lift. The day the last rep moves at half the speed of the first and changes shape on the way up, you have found the real top of your range for that session. Stop there. Log what was clean. The honest seven beats the fictional eight every time, because next month the seven is still true and the eight was a story you told the bar.
When to add the weight, and how much
Add weight when every set hits the top of the range with reps you would show a referee. Not when most sets do. Not when the first set does and the rest limp behind it. Every set, clean.
How much you add depends on the lift, not on a fixed plate. On a deadlift or squat, the smallest meaningful jump is usually five kilos and sometimes ten. On a press, where the runway is shorter and the leverages are crueler, two and a half kilos is often the right move, and microplates that let you add a single kilo are not vanity. They are the difference between staying on the bottom rung for two weeks instead of two months. The smaller the jump, the longer double progression keeps working before you need anything more complicated.
And it does eventually need something more complicated. Double progression is the bridge, not the destination. It carries most lifters comfortably through the intermediate years, and for a while it carries them well past where linear progression quit. But the same Tuesday is coming again, further out. When holding the weight and chasing reps stops yielding new reps for weeks at a stretch, you have reached the end of this runway too, and the next conversation is about waving your loads across a week, or training to a target RPE instead of a target rep, or building blocks that accumulate and then intensify. Those are later. They are not today.
Today the fix is smaller than the panic that announced it. The bar stopped moving every session because it was always going to. You do not need a new program. You need to add reps before you add weight, keep the reps honest enough to pass on a platform, and pick the range on purpose instead of by habit. Do that, and the runway in front of you is years long.