How to Use the Acute:Chronic Workload Ratio (ACWR) Safely

When Tim Gabbett introduced the acute:chronic workload ratio framework in 2016, a subsequent analysis of 1,334 professional cricket fast bowlers demonstrated that athletes with an ACWR above 1.5 were 2.1 times more likely to sustain a non-contact injury in the following week compared to athletes with an ACWR in the 0.8–1.3 range. That finding drove rapid adoption across elite sport. The problem: subsequent re-analyses and methodological critiques (Impellizzeri et al., 2020) showed the original mathematical formulation contained a significant statistical artifact, and the celebrated 0.8–1.3 'sweet spot' was partly an artefact of regression to the mean. This has not invalidated the ACWR framework, but it has substantially changed how it should be interpreted and used.

This guide explains how ACWR is correctly calculated, what load metrics work best, what the current evidence actually supports, and how to integrate ACWR monitoring into practical training decisions without over-relying on a single number.

What Is the ACWR and What Does It Actually Measure?

The acute:chronic workload ratio compares how much training an athlete has done recently (acute load, typically the past 7 days) to their established long-term training capacity (chronic load, typically a rolling 28-day average). The ratio is calculated as:

ACWR = Acute (7-day) Load ÷ Chronic (28-day) Average Weekly Load

Conceptually, the chronic load represents the athlete's current fitness — the amount of work their body is adapted to absorb. The acute load represents the stress being placed on that body right now. When acute load significantly exceeds chronic load (ACWR well above 1.0), the athlete is being exposed to more stress than their current fitness level is adapted to absorb. When acute load is substantially below chronic load (ACWR well below 1.0), the athlete is in a detraining phase — which itself carries injury risk when training is rapidly resumed.

What ACWR does NOT measure directly: tissue-level structural stress, psychological stress, sleep quality, nutrition status, or whether the type of training (e.g., switching from aerobic volume to high-intensity intervals) matches the chronic base. ACWR is a coarse aggregate measure. Its value lies in detecting large, sudden changes in training demand — the circumstances most strongly linked to injury in the epidemiological literature.

Calculating ACWR: Rolling Average vs EWMA

The original Gabbett (2016) method used a simple rolling 7-day acute load divided by a rolling 28-day average. This mathematical formulation contains a coupling artifact: the acute load (week 1 of the 28-day window) is mathematically included in the chronic load calculation, which inflates correlation statistics and narrows the apparent injury-risk range. The exponentially weighted moving average (EWMA) approach, proposed by Williams et al. (2017), resolves this by calculating acute and chronic loads independently:

• Acute EWMA load: Applies a decay constant (λ) of 2/(7+1) = 0.25 — more recent days weighted more heavily.
• Chronic EWMA load: Applies a decay constant of 2/(28+1) = 0.065 — slower response to recent changes, better reflecting true physiological adaptation.

EWMA formula: EWMA today = Load today × λ + EWMA yesterday × (1 − λ)

In practical terms: use EWMA when calculating ACWR in a spreadsheet or monitoring software. The difference between rolling-average and EWMA ACWR is small at stable loads but becomes meaningful during rapid load changes — precisely when injury risk identification matters most. Most current athlete monitoring platforms default to EWMA; verify which calculation your software uses.

What Load Metric Should You Use?

The choice of load metric profoundly affects ACWR accuracy. Different load measures capture different aspects of training stress:

For most strength and power athletes, sRPE remains the most practical external-load metric because it captures gym, field, and skill training in a single consistent currency. Velocity-based mechanical load can augment sRPE specifically for resistance training sessions where intensity variation matters most.

The Sweet Spot Controversy: What the Research Actually Shows

The original claim that maintaining ACWR between 0.8 and 1.3 protects against injury (Gabbett, 2016) generated widespread adoption but has since been substantially qualified. Key findings from subsequent research:

• Impellizzeri et al. (2020) showed that the ACWR's predictive validity for injury risk is largely context-dependent and smaller in magnitude than originally reported once the mathematical coupling artifact is removed.
• Cross et al. (2019) meta-analysis of 18 studies found that pooled evidence did not support a specific 'sweet spot' range — the relationship between ACWR and injury risk varies substantially by sport, population, and load metric.
• Individual chronic fitness level moderates risk: an ACWR of 1.4 in an athlete with a high chronic load (well-conditioned) represents less absolute stress than the same ACWR in an athlete with a low chronic load (deconditioned).

Current consensus (Gabbett et al., 2021 update): ACWR is most valuable as a warning flag when values are clearly elevated (above 1.5) or when the direction of change is rapid (a 40%+ acute spike in a single week). It should never be used as the sole arbiter of training load decisions. Athlete context — freshness, readiness markers, tissue tolerance history, subjective wellbeing — must be integrated.

Step-by-Step: Setting Up ACWR Monitoring

1. Choose one primary load metric for each training context (e.g., sRPE for all sessions; optionally add GPS distance for field sessions). Consistency across sessions matters more than using the theoretically optimal metric.
2. Build a minimum 4-week baseline before using ACWR to make training decisions. ACWR values in weeks 1–3 are unreliable because the chronic load has not stabilized. A common error is reacting to elevated ACWR values in the first month of a new monitoring program — these values reflect insufficient chronic load data, not real elevated risk.
3. Log every training session within 30 minutes of completion. sRPE accuracy degrades with delay; athletes consistently underestimate session intensity when logging 12+ hours later.
4. Calculate EWMA acute and chronic loads using the formulas above, or import data into a monitoring platform that does this automatically. Review ACWR trend (direction of change) as much as the ACWR value itself.
5. Flag ACWR values above 1.4 for discussion — not automatic load reduction. Consider athlete readiness markers (sleep quality, CMJ, HRV), session quality from the current week, and proximity to competition before making a load change decision.

Practical Decision Rules for Coaches

The following decision rules should be used as prompts for coach-athlete conversation, not automatic prescriptions:

• ACWR 0.8–1.3, stable trend: Normal training range. No load adjustment needed. Check readiness markers as usual.
• ACWR 1.3–1.5, rising: Caution zone. Check athlete subjective wellbeing (fatigue, soreness, mood). If CMJ is more than 5% below baseline, reduce session volume by 20%. If athlete reports feeling good and CMJ is normal, proceed with planned session but avoid adding new loading stressors.
• ACWR above 1.5: High spike territory. Unless it is a planned pre-competition spike in a conditioned athlete with high chronic base, reduce upcoming week's load by 20–30%. Do not schedule maximal strength testing or high-intensity plyometric sessions until ACWR returns below 1.4.
• ACWR below 0.6 for 3+ consecutive weeks: Detraining risk. Gradual re-escalation required — weekly load increases capped at 10–15%. Jumping from a low chronic base to high acute load (e.g., first week back after a holiday) is one of the highest ACWR spike scenarios in practice.

Integrating Velocity-Based Load with ACWR

Standard ACWR calculations using sRPE or GPS distance do not capture the neuromuscular load of resistance training sessions with sufficient granularity. A heavy squat session at 85% 1RM for 5×3 reps has a very different tissue stress profile than a jump training session at the same perceived effort rating. Velocity-based mechanical load (VBL) can fill this gap:

VBL per set = Mean concentric velocity (m/s) × Load (kg) × Reps

Summed across all sets in a session, VBL provides a work-output-based internal load metric that reflects both load and intent-driven force production. When an athlete is fatigued, their MCV drops at a given load — this automatically reduces VBL even if external load (kg) remains constant. This self-correcting property makes VBL a more sensitive neuromuscular load proxy than lifted tonnage alone.

Practical integration: use VBL as the load metric for resistance training sessions in your ACWR calculation. Use sRPE × duration for on-field or conditioning sessions. Sum both into daily and weekly load totals. This hybrid approach captures the full training demand across modalities more accurately than any single metric.