Training Load Explained: ATL, CTL, TSB and What They Actually Mean for Coaching

Your athlete just uploaded a ride file. Inside it: 3,600 seconds of data, recorded at one data point per second across a dozen channels — power, heart rate, cadence, speed, GPS, elevation, temperature. That's roughly 43,000 individual data points from a single one-hour session.

Now multiply that by 30 athletes, each logging six to nine sessions a week across swimming, cycling, and running.

This is the data reality of modern endurance coaching. And the three acronyms that are supposed to make sense of it all — ATL, CTL, and TSB — are the ones coaches either swear by or quietly admit they don't fully trust.

Let's fix that.

The Performance Management Chart: where ATL, CTL, and TSB live

Before we unpack each metric, it helps to understand the system they belong to. The Performance Management Chart (PMC) is a model originally developed by Dr. Andrew Coggan for cycling power data. It's since been adopted across endurance sports and is built into platforms like TrainingPeaks, Today's Plan, and most modern coaching software.

The PMC models three things: how fit an athlete is, how fatigued they are, and the balance between the two. That's it. Three lines on a chart that tell a story about readiness.

The inputs vary by sport — Training Stress Score (TSS) for cycling, rTSS or pace-based equivalents for running, swim-specific stress scores for the pool — but the underlying logic is the same.

CTL: Chronic Training Load (your athlete's fitness)

What it measures: The rolling average of training stress over the past ~42 days.

What it tells you: How much training load your athlete has been absorbing consistently. A higher CTL means a higher sustained training volume. It's the closest thing the PMC gives you to a "fitness" number.

How to think about it as a coach: CTL is the slow-moving line. It doesn't jump overnight. When you see CTL climbing steadily, your athlete is building. When it plateaus, they're maintaining. When it drops, they're either tapering, recovering, or — and this is where your coaching judgment matters — losing consistency.

What it doesn't tell you: Whether the athlete is actually getting faster, or just accumulating stress. A high CTL from lots of low-quality junk miles looks the same as a high CTL from structured, purposeful training. CTL measures load, not quality. That distinction matters.

Coaching scenario: You notice an athlete's CTL has been climbing steadily at about 3-4 points per week for the past month. They're 8 weeks out from an A-race. Good — they're building at a sustainable rate. But another athlete's CTL has jumped 15 points in a single week because they did a training camp. That's not fitness — that's a fatigue spike disguised as a fitness gain. CTL will eventually reflect the camp, but in the short term, you need to look at ATL.

ATL: Acute Training Load (your athlete's fatigue)

What it measures: The rolling average of training stress over the past ~7 days.

What it tells you: How hard your athlete has been training in the very recent past. ATL moves fast — it responds to this week's sessions, not last month's.

How to think about it as a coach: ATL is your early warning system. When ATL spikes sharply above CTL, your athlete is accumulating fatigue faster than their body has adapted to handle. That's fine for a planned overreach week. It's a problem if it's happening because the athlete decided to add extra sessions without telling you.

What it doesn't tell you: Whether the fatigue is productive. A structured overreach block and a chaotic "I felt good so I went hard four days in a row" week can produce the same ATL spike. Context matters — and context is something no chart provides on its own.

Coaching scenario: Your triathlete's ATL has been sitting at 75 for weeks (their normal training rhythm), and this week it's suddenly at 110. You check the sessions: they raced a local sprint triathlon on Saturday and then did a long ride on Sunday because they "felt fine." That's the kind of fatigue spike that shows up as a flu or a niggling injury ten days later if you don't intervene. ATL told you it happened. Your job is to decide what to do about it.

TSB: Training Stress Balance (the readiness indicator)

What it measures: CTL minus ATL. Literally: fitness minus fatigue.

What it tells you: Whether your athlete is fresh (positive TSB), absorbing load (negative TSB), or on the edge (deeply negative TSB). It's the metric coaches use most often when making race-week decisions.

How to think about it as a coach: TSB is the metric that answers "Is this athlete ready to perform?" A positive TSB in race week — typically somewhere between +10 and +25, depending on the athlete — suggests they've tapered well and are fresh enough to race. A deeply negative TSB (say, -30 or worse) tells you they're buried in fatigue and need recovery.

What it doesn't tell you: Everything. TSB is a blunt instrument. It doesn't account for sleep, nutrition, life stress, illness, or the athlete's subjective feeling. An athlete can have a perfect TSB on paper and feel dreadful because they've been working 60-hour weeks and sleeping five hours a night.

Coaching scenario: Your swimmer's A-meet is in 10 days. Their TSB is currently at -15. Based on their taper response in past seasons, you know they need about 12 days to reach a TSB of +15. The math works — but only if they actually rest. You write the taper, flag them for screening, and check in mid-week. TSB gave you the starting point. You provided the judgment.

The metrics nobody talks about: what the PMC misses

The PMC is powerful for what it does — modelling the fitness-fatigue relationship over time. But here's what it won't tell you, and what experienced coaches account for instinctively:

Session quality. An athlete who nails their intervals at prescribed power and heart rate is getting a different training stimulus than one who hit the same TSS by going too hard on an easy day. Same load, very different outcomes.

Training distribution. Two athletes with identical CTL can have completely different training polarisation. One might be doing 80/20 easy-to-hard. The other might be living in the grey zone. The PMC doesn't distinguish between them.

Multi-sport interactions. For triathletes, the PMC typically tracks each discipline separately. But fatigue doesn't respect discipline boundaries. A massive bike week affects running quality. A heavy swim block affects everything. Cross-discipline fatigue accumulation is one of the hardest things to model — and one of the most important things for triathlon coaches to track.

The human stuff. Sleep, stress, travel, illness, motivation, confidence. These don't have a data channel on your athlete's Garmin, but they influence performance and injury risk as much as any training metric.

This is where coaching judgment fills the gap that no metric can close.

How to use these metrics across a squad

Understanding ATL, CTL, and TSB for one athlete is manageable. Understanding them across 30 athletes is a different problem entirely.

This is the data challenge that most coaching tools weren't designed to solve. TrainingPeaks gives you a PMC per athlete, but no squad-level view. You're opening individual athlete pages one at a time, checking their charts, mentally flagging who needs attention, and trying to hold 30 different fitness-fatigue stories in your head simultaneously.

The coaches who manage this well have systems. Some use spreadsheets with colour-coded cells. Some check PMC charts in a specific order each Monday morning, starting with athletes who raced or had heavy weeks. Some keep a running notes doc with flags.

But even the best systems are manual, time-consuming, and dependent on the coach remembering to check. The athlete whose TSB quietly drops to -35 without anyone noticing is the one who ends up injured.

The question isn't whether these metrics are useful — they are. The question is whether your workflow lets you actually see them across your whole squad, consistently, before someone needs your attention.

Making training load practical

If you're looking to sharpen how you use ATL, CTL, and TSB in your coaching, here's what the best coaches tend to do:

Trust relative trends, not absolute comparisons. Higher-end devices — which most triathletes tend to use — are fairly consistent with each other. An ATL of 80 on a Fenix and a Forerunner will be in the same ballpark, maybe +/- 5–10 points. But across a mixed squad with different hardware, those small gaps add up. Broad-stroke thresholds and trends within an individual athlete are reliable. Comparing exact numbers between athletes on different setups is where you need to be careful.

Watch the rate of change, not the number. A CTL that climbs 5 points per week is sustainable. A CTL that jumps 15 points in one week is a red flag regardless of where it ends up. The direction and speed of change are more reliable coaching signals than any absolute value.

Use TSB for taper planning, not daily decisions. TSB is most useful on the macro scale — guiding race-week readiness. Thresholds like +10 to +25 are guidelines, not gospel. An athlete who races well at +8 and another who needs +20 are both normal.

Combine with subjective data. A simple 1-10 readiness score each morning, compared against PMC trends, gives you a richer picture than either source alone.

Screen before you analyse. You don't need to deep-dive every athlete's PMC every week. You need to know which athletes to deep-dive — the ones whose numbers have shifted, who missed sessions, or whose trends are heading the wrong direction.

That screening step — figuring out who needs attention across your whole squad — is the most time-consuming part of working with training load data. And it's the step that determines whether the metrics actually improve your coaching, or just sit in a chart nobody has time to read.

Simma reads your athletes' sessions automatically and tracks training load across your whole squad — so you can see who needs attention before you open a single file. Join the early access waitlist.