If you have spent any time looking at training apps, you have probably seen the acronyms ATL, CTL, and TSB drift past in a chart you did not entirely trust. They are not jargon for the sake of jargon. Each one is a real, simple piece of math, and together they are the closest thing the sport has to a dashboard for whether you are getting fitter, more tired, or more ready.
Here is the plain version.
One stress number per run
You cannot track load without first quantifying how much load each run delivered. The standard tool for that in running is TRIMP, short for Training Impulse, a metric Eric Banister proposed in the 1970s. The simplest version multiplies the duration of a run by a heart-rate-derived intensity weight. A long, slow run and a short, hard one can end up with very similar TRIMP scores even though they feel nothing alike, which is the point. TRIMP scores them on a common axis: total physiological stress.
Smart Runner computes a TRIMP score for every workout the app sees, whether it came from the Watch, from Apple Health, or from a manual entry.
ATL - Acute Training Load
ATL is the exponentially weighted moving average of your daily TRIMP scores over roughly the last 7 days. Recent runs count more, older runs count less. In practice it answers a simple question: how tired should I be right now?
When ATL is climbing, you are stacking stress faster than your body is clearing it. When ATL drops, you are recovering. A hard week is visible as a sharp ATL spike in the chart, the kind of spike that should be followed by a cutback week to let the number come down.
CTL - Chronic Training Load
CTL is the same calculation over a longer window: an exponentially weighted moving average over the last 42 days. It moves slowly. One huge week barely nudges it. Three months of consistent training builds it brick by brick.
CTL is the closest thing we have to a number for fitness. A higher CTL means you have built more aerobic infrastructure: more capillaries, more mitochondria, more economy, more time on feet absorbed without breaking. Your CTL on race day is a good predictor of how durable you will be in the back half of a marathon.
CTL has one important honest caveat. It rewards volume of stress. It does not directly measure whether that stress was the right type. A runner who only ever runs easy can build a high CTL and still lack the threshold and VO2max work to race fast. CTL is necessary, not sufficient. The intensity distribution still matters, which is why Smart Runner also tracks your easy-to-hard ratio separately.
TSB - Training Stress Balance
TSB is just the difference: TSB = CTL - ATL. The point is to ask whether your fitness number outpaces your fatigue number, and by how much.
Think of it as the form dial:
- TSB very negative (worse than -20): You are deep in a hard block. Lots of fatigue piled on top of slowly rising fitness. Expect feeling heavy and slow.
- TSB mildly negative (-5 to -15): The productive zone in a build phase. You are absorbing real work and adapting.
- TSB around 0: Steady. Fitness and fatigue balanced. Good for tune-up races and quality workouts.
- TSB +5 to +15: The classic peak window for a goal race. Recent training has tapered enough that fatigue has cleared, but not so long that fitness has decayed.
- TSB above +20: Too rested. You have lost some adaptive sharpness. Usually shows up after extended time off or an over-aggressive taper.
Why the rolling shape matters
The shape of your TSB over the cycle is more informative than the value on any one day. Smart Runner draws the curve so you can see the rhythm: hard week pushing TSB down, cutback week letting it come up, the slow climb out of negatives in the last two weeks before a race. A good taper looks like a smooth pull from -10 to +10 over about 14 days.
A taper that ends near 0 is usually too short. A taper that ends above +15 is usually too long, or the training before it was too light.
What it does not capture
Training load is a strong signal, not the whole story. ATL/CTL/TSB will not see:
- Life stress. Sleep, work pressure, a sick child, dehydration. Internal load goes up; the model does not know.
- The right kind of stress. 60 minutes of zone-2 easy and 60 minutes of red-line tempo can produce similar TRIMP scores, but they drive different adaptations.
- Race-specific readiness. A marathon-specific block of long-pace work is different from a 5K block at the same TRIMP. Race-day readiness depends on what you trained, not just how hard.
This is why the model is a coach's tool, not a coach. It tells you the shape of the build. It does not tell you whether you trained the right zones in the right order, which is where the methodology decisions in Pfitzinger, Daniels, and Canova come in.
How Smart Runner uses it
The Insights tab in Smart Runner exposes the same three numbers, drawn over the last several weeks. The plan engine uses them at every weekly recalculation:
- If your ATL has spiked beyond a safe range relative to CTL, the next week dials down volume.
- If your CTL has stalled flat for too long, the plan introduces a progression week.
- In the last 14-21 days before a goal race, the plan moves TSB toward the +5 to +15 sweet spot.
Plus, on the Stats screen, the Acute:Chronic Workload Ratio (ACWR) gives a more conservative early-warning signal for sudden overload. The combination, TRIMP into ATL/CTL into TSB, with ACWR as the safety check, is the load layer that sits behind every adaptive run on the calendar.
The short version
Get a stress score per run. Average the last week. Average the last six weeks. Take the difference. Watch the difference move. That is training load.
The math is simple. The trick is using it to coach yourself, not just track yourself.