Acute:Chronic Workload Ratio (ACWR)

ACWR compares modeled recent load with modeled long-term load. In Trainingload.ai, A:C is same-day Fatigue (ATL) divided by same-day Fitness (CTL).

The ratio can highlight a change in load context, but ACWR is not an injury or readiness prediction. It does not directly measure adaptation, fatigue, recovery, or medical risk.

Example values are for interface preview only.

Formula and data source

ATL and CTL use the same daily-load series:

decay_ATL = exp(-1 / 7)
decay_CTL = exp(-1 / 42)
ATL_today = ATL_yesterday × decay_ATL + daily_load × (1 − decay_ATL)
CTL_today = CTL_yesterday × decay_CTL + daily_load × (1 − decay_CTL)

ACWR_today = ATL_today / CTL_today

If today’s ATL is 60 and CTL is 50, ACWR is 1.20. If CTL is zero, the product returns 0 to avoid division by zero.

Activities, thresholds, and load-source settings can change both numerator and denominator. ACWR is unitless and discards their absolute scale.

Which ACWR is this?

There is no single ACWR algorithm. Trainingload.ai uses a coupled PMC exponential ratio: today’s load enters both ATL and CTL, and both curves use exponential time constants.

It is not a simple “last seven days ÷ 28-day weekly average,” nor an uncoupled variant that removes the recent window from the chronic window. Thresholds from another algorithm, sport, population, or load source cannot be transferred directly.

Reading the ratio

• Above 1: ATL is above CTL.
• Near 1: ATL and CTL are close; this does not mean training is optimal.
• Below 1: ATL is below CTL.

Always inspect numerator and denominator. ATL 12 / CTL 10 and ATL 120 / CTL 100 both equal 1.20, despite very different absolute loads. A new account or very low CTL can also produce a large ratio after one session.

The interface currently shows four bands: <0.8, 0.8–<1.3, 1.3–<1.5, and ≥1.5. These are display rules for reviewing relative load—not proof of “optimal,” “safe,” “dangerous,” “undertrained,” or “overtrained” states.

Practical review

Check that ACWR, ATL, and CTL refer to the same date. Review the absolute values, trace the change to specific activities, and check thresholds and load-source priorities. Then add session quality, sleep, pain, illness, subjective fatigue, and injury history before changing the plan.

References

• Intervals.icu developer explanation of the Fitness and Fatigue formulas
• Gabbett (2016): The training—injury prevention paradox
• Windt and Gabbett (2019): Mathematical coupling in ACWR
• Maupin et al. (2020): Systematic review of ACWR and injury risk
• Impellizzeri et al. (2020): Conceptual issues and fundamental pitfalls
• Wang et al. (2020): Lessons learned from ACWR

Related pages

• Fatigue (ATL)
• Fitness (CTL)
• Form (TSB)
• Training load