Same tonnage, 25 minutes longer: the fatigue signal your log misses.
Two sessions with identical tonnage can be very different stimuli. Density, load moved per minute, tends to slip before RPE climbs or top sets fall. A tonnage-only log can't see it.
Week nine looked exactly like week one on paper
The squat day in week nine matched week one set for set. Five sets of five at the same bar weight, the same leg press work after, the same curls to finish. Tonnage matched to the kilo. The session ran 25 minutes longer.
Nobody wrote the extra time down. It hid in the rests that stretched from three minutes to five, the extra warm-up single taken because the bar felt heavy, the long sit on the bench before the last working set. Every logged number said the lifter was holding steady.
Then week ten came, and the top set didn't.
Density drift: the signal that moves first
Call it density drift. When fatigue builds across a block, the first thing to give usually isn't the weight on the bar or the reps you hit. It's how long you need between efforts to keep hitting them.
That's what density measures: load moved per unit of time. Westside Barbell's name for it is VLPM, volume lifted per minute. VLPM (Volume Lifted Per Minute) measures the average weight an athlete lifts per minute during a workout. ([source](https://www.westside-barbell.com/blogs/the-blog/unlocking-potential-with-vlpm?srsltid=AU7gw4UqSC-eVTWHQy4L0SAVGz7EbqCdi9CLkk3yznEveykohTJLGzxp)). Divide total session tonnage by session minutes and you get one number that carries both how much work you did and what it cost you to do it.
And the cost is the part that changes early. A lifter carrying fatigue will still grind out the prescribed sets. They'll just buy each one with more rest.
Why tonnage calls two different sessions equal
Tonnage is the default proxy for work in most logs, and it's a decent one. It answers how much you lifted. It has no opinion about how long the lifting took, so a 55-minute session and an 80-minute session with the same sets score the same.
RPE doesn't rescue it. Rest is how lifters keep RPE honest under fatigue: wait longer and the next set of five feels like an 8 again. So the RPE column stays flat, the tonnage column stays flat, and the top set holds. Three signals, all quiet, all because time absorbed the fatigue first.
By the time a top set drops, the stretched rests have usually been showing up for weeks. They just weren't recorded anywhere a chart could read.
Density as a diagnostic, not a rest prescription
Two different questions get tangled here. One asks how long you should rest. The other asks what your rest lengths are telling you. They need different answers.
On the first question, the evidence doesn't favor rushing. Schoenfeld and colleagues (2016, Journal of Strength and Conditioning Research) found longer interset rests produced better strength and hypertrophy outcomes in trained men than short rests. Schoenfeld et al. 2016 found longer rest periods produce greater muscle strength and hypertrophy gains than shorter rest periods in resistance-trained men. ([source](https://pubmed.ncbi.nlm.nih.gov/26605807/)). Density as a training method exists too. Charles Staley's Escalating Density Training is built around fitting more work into a fixed time window. That's density as a programming lever.
This piece argues the other use. Keep your rests wherever your program puts them. Then watch whether the same session, run the same way, starts taking longer than it used to.
Don't read this as 'rest less'
Shortening rests to push your density number up defeats the point. The signal only works if you rest the way your program says and let the clock record what actually happens. Density is the thermometer here, not the thermostat.
One block, walked through
Take an illustrative lower-body day run the same way for eight weeks: same template, same load progression held flat for simplicity, logged total around 12,000 kg. The lifter rates the top set RPE 8 every week, honestly.
Week one takes 55 minutes. Week five takes 62. Week eight takes 80. Nothing in the tonnage or RPE columns moves. Density falls from about 218 kg per minute to 150, roughly a 31 percent drop, and it falls steadily rather than all at once.
A single slow session means little. Busy gyms happen, and so does waiting ten minutes for the only rack. But a three-week slide on the same template, with every other number flat, is a pattern the lifter can see coming well before the top set gives out.
Week
Tonnage (kg)
Session length
Top set RPE
Density (kg/min)
1
12,000
55 min
8
218
5
12,000
62 min
8
194
8
12,000
80 min
8
150
Illustrative example, not study data. Density = tonnage divided by session minutes. Tonnage and RPE read flat; only density moves.
What changes if density moves first
If density drifts before the other signals, a log without session time is missing the earliest fatigue read it could have. That's a strong claim, and we'll commit to it: tonnage-only logging will always see a stall later than it could have.
The practical change is small. Record when a session starts and ends. Compare like with like, same template against same template, because a heavy squat day and a light pull day were never going to share a density. Then read the trend across weeks rather than any one day.
What to do with a sustained drop is still the lifter's call. It might be a deload. It might be sleep, or a new job, or a program that's simply outgrown its current loads. The chart doesn't decide. It just shows the question weeks earlier.
Fatigue gets paid for in minutes before it gets paid for in kilos.
Where session time lives in the log
Session duration is already data. Any watch-recorded workout has a start and an end, and a log that keeps those timestamps next to every set can put tonnage and time on the same chart without the lifter doing extra arithmetic mid-session.
That's the angle Platepusher takes with history: the sets you've already logged are an experiment you've been running on yourself, and time is one more column in it. Platepusher timestamps every workout with a start time captured the moment a session begins and an end time captured on Finish, with duration always computed on the fly from that difference rather than stored separately; the Apple Watch doesn't keep its own copy of this record, it just mirrors the phone's start time for its live display (its own on-device timer is only used to stream heart-rate to HealthKit). Today the app surfaces duration only as an overall average across recent sessions and shows trend charts for exercise progression, body-part training load, and weekly consistency, it does not yet break duration or session density down per plan template..
Run your next block with the clock recorded. The density line will either stay flat, which is good news, or start sliding while everything else looks fine. Either way you'll know sooner.
Log your next block with session time on the record. Get Platepusher.
Platepusher is a gym log built for lifters with years of training behind them. It keeps every set, every date, and the session around them, and treats that history as the record of an experiment you've already run on yourself. No streaks, no badges, no coach voice telling you what to do. The chart shows what your training has been recording; you decide what it means.
