Acute:Chronic Workload Ratio: A Practical Guide for Coaches

What ACWR is trying to measure

The acute:chronic workload ratio compares a short-term training load window (typically the past week) against a longer rolling average (typically four weeks), on the theory that a sharp spike relative to an athlete's recent baseline — not high load in isolation — is what drives injury risk. The concept was popularized by research proposing a "sweet spot" ratio (commonly cited as roughly 0.8–1.3) associated with lower injury rates in the underlying data.

Why coaches adopted it quickly

ACWR offered something training-load monitoring hadn't really had before: a single number, calculated from data many teams already collected (GPS distance, session-RPE, or similar), with a specific target range attached to it. That combination — simple, actionable, and framed as injury-preventive — made it an easy sell across professional and amateur sport alike.

Where the evidence gets more complicated

More recent methodological critiques have raised substantive concerns with how ACWR is typically calculated and interpreted:

• The same underlying training data can produce meaningfully different ACWR values depending on the calculation method used (rolling average vs. exponentially weighted moving average), which undermines comparing a single "sweet spot" range across studies that used different methods.
• Some of the statistical relationships reported in early ACWR studies have been argued to be partly a mathematical artifact of how acute and chronic load are calculated from overlapping data, rather than purely a physiological signal.
• Individual variation in injury tolerance is large enough that a population-level "sweet spot" may not transfer cleanly to guidance for a specific athlete.

None of this means workload monitoring is useless — it means the specific 0.8–1.3 target range popularized in early research should be treated as a rough heuristic, not a precise threshold.

How to actually use it

A more defensible approach treats ACWR as one input into a broader monitoring picture, rather than a standalone alarm system: