reckonerFront-rack carry
Two kettlebells racked on the chest, elbows in, and then you walk. Everything wants to pull you into a backward lean, and refusing that lean for the length of the walk is the exercise. It is the carry that asks the most of the upper back and the least of the grip.
Elbows high, ribs down, stand tall.
- Pattern
- Carry
- Trains
- Abs, upper back, traps
- Kit
- Kettlebell or dumbbell
- Loading
- A weight you choose
- Reckoner prescribes
- 4 holds of 30 to 60 seconds, on the fat loss plan. Reckoner does not prescribe this pattern for muscle or strength.
- Progression step
- 10 seconds on the hold
- Skill
- Technical
- Recovery cost
- Moderate
- Not prescribed if you have declared
- Shoulder, wrist, neck
Muscles worked
How to do it
Clean the bells to the rack
Bring one bell at a time to the front of the shoulder, handle across the palm, wrist straight, elbow tucked in against the ribcage. The bell rests on the forearm rather than being held up by the hand.
Set the ribs down before you move
The load in front will try to tip you backward. Pull the ribs toward the hips, squeeze the glutes, and stand tall with the weight over the middle of your feet rather than behind them.
Walk with short, quiet steps
Small steps, normal breathing, eyes forward. If your steps get long the trunk starts to sway, and a swaying trunk means the position has already gone.
Stop when the elbows drop
The elbows sinking and the chest caving are the first honest signs of failure, and they arrive before your legs do. That is the end of the set, and the time you held it there is the number to log.
What goes wrong
Leaning back to balance the load
It works, which is why it happens. Tipping backward lets the skeleton hold the weight instead of the trunk, and the exercise you came to do stops at the moment you do it.
Letting the wrists bend back
The bell hangs off the hand and the wrist folds. It hurts, and it also means the bell has slipped out of the rack and onto your grip, which is not what this carry is for.
Choosing distance over position
Walking further with the elbows dropping is not a better set. Position first, then time, then load, in that order.
How Reckoner progresses this lift
Six weeks of this lift, from 16 kg, run through the real engine. Nothing below was typed in: the prescription, the decision and the sentence beside it all come from the same code that writes your week.
| Week | Prescribed | What it did |
|---|---|---|
| 1 | 4 x 30 s at 16 kg | Longer hold 30 → 40 sYou carried the full time at RPE 7, so the carry gains ten seconds. Once it reaches 60 seconds the load goes up instead. |
| 2 | 4 x 40 s at 16 kg | Longer hold 40 → 50 sYou carried the full time at RPE 7, so the carry gains ten seconds. Once it reaches 60 seconds the load goes up instead. |
| 3 | 4 x 50 s at 16 kg | Longer hold 50 → 60 sYou carried the full time at RPE 7, so the carry gains ten seconds. Once it reaches 60 seconds the load goes up instead. |
| 4 | 4 x 60 s at 16 kg | Load up 16 kg → 20 kgYou carried the top of the window at RPE 7, so the load goes up 4 kg to 20 kg and the carry drops back to 30 seconds. Given the choice between heavier and longer, a loaded lift gets heavier. |
| 5 | 4 x 30 s at 20 kg | Deload 40% of sets offFive weeks of building, then one week lighter. About 40% of the sets come off. The load holds and your reps stay where you earned them, so you keep the strength and lose the fatigue. |
| 6 | 2 x 30 s at 20 kg | Longer hold 30 → 40 sYou carried the full time at RPE 7, so the carry gains ten seconds. Once it reaches 60 seconds the load goes up instead. The lighter week is over, so this is back to four sets. |
Real engine output, not an illustration. Every row was produced by running the engine itself when this page was generated.
The full rule, and why the reps move before the load does, is in how Reckoner decides to add weight.
Questions people ask
Why two bells instead of one?
Because balanced is the easier thing to load. Carrying a weight on one side pushes low back compression far higher than splitting the same total across both hands, and in one measurement doubling the total across two hands still produced less spine compression than the original load in one1. If you want the one-sided version, that is the suitcase carry, and it is a different exercise on purpose.
Is this a leg exercise or a core exercise?
Neither, quite. It is a position held under load while you walk, and the trunk and the upper back are what fail first. Trained as a category, work of this kind produces small to moderate improvements in strength, sprint speed and change of direction in athletes2.
Why is it measured in seconds?
Because a carry is performed against a clock rather than in reps, so it progresses on the axis it is actually done on. The engine adds seconds to the walk the same way it adds weight to a bar.
Similar movements
Reckoner eases the load first when this lift stalls, and offers one of these only if it stalls again after that. A lift already as light as your kit goes has no load left to ease, so the offer comes straight away. Each of them has a page.
Sources
- McGill, Marshall and Andersen, Ergonomics, 2013. Low back loads while walking and carrying: comparing the load carried in one hand or in both hands six participants walking with buckets, where thirty kilograms carried in one hand pushed low back compression above 2800 newtons while the same total split between both hands brought it to 1570, a reduction of 44 per cent, and where doubling the total by carrying thirty kilograms in each hand still produced less spine compression than thirty in one
- Saeterbakken and colleagues, Sports Medicine, 2022. The effects of trunk muscle training on physical fitness and sport-specific performance in young and adult athletes: a systematic review and meta-analysis thirty-one controlled trials and 693 athletes aged eleven to thirty-seven, with trunk muscle training producing significant small to large improvements in maximal strength, local muscular endurance, lower limb power, linear sprint speed and change of direction, and a moderate effect on sport-specific performance