Rethinking Series: Acute:Chronic Workload Ratio vs. Suddenness — Worth Chasing?

---

What is ACWR?

---

Definition & Calculation

If you have some experience with load management, you have definitely heard of the acute:chronic workload ratio (ACWR). The foundation of ACWR is the hypothetical relationship between training loads, fitness, injuries, and performance (see Figure 1).

Figure 1. Hypothetical relationship between training loads, fitness, injuries, and performance (adapted from Gabbett¹)

Based on that, the main concept behind ACWR is to avoid "spike load" and thereby minimize the risk of negative adaptation or injury.

Well, "spike" is relative; we have to compare both chronic and acute loads, so isolating either doesn't give us the whole picture.

Basically, when the load input is too high (overuse) or too low (underprepared), it can negatively impact performance.

This makes sense, but the question turns out to be "Is there an optimal load for the relationship between chronic and acute load?"

A common rule of thumb has been to keep weekly load increases under 10–15%¹.

However, before we go deeper about whether this optimal load exists, let's go back to the OG paper on ACWR by Hulin et al.². Originally, the calculation of ACWR is

"The average of the past 7 days (one week) / The average of the past 28 days (4 weeks)"

Acute load (7 days) is meant to represent "fatigue," and chronic load (28 days) "fitness."

Just a quick reminder that there isn't strong evidence that using a "7/28" window is proper or ideal, nor is it based on human physiology or the adaptation process.

This ACWR calculation is a "Rolling Average (RA)".

Both windows roll forward one day at a time.

However, the biggest problem with RA is that it weights every day in the window equally, so a session from four weeks ago counts as much as yesterday's.

In 2017, Williams et al. proposed exponentially weighted moving averages (EWMA) to calculate acute and chronic loads, which give more weight to recent loads³.

Here is the calculation of EWMA:

“EWMA_today = Load_today × λₐ + (1 − λₐ) × EWMA_yesterday”

λₐ = 2 / (N + 1)

N is the chosen time decay constant, typically 7 and 28 days for acute (‘fatigue’) and chronic (‘fitness’) loads.

Previous research showed that EWMA_ACWR was more sensitive than RA_ACWR⁴ (see Figure 2).

Figure 2. EWMA_ACWR vs RA_ACWR in a simulated dataset (concept from Murray et al.⁴)

Criteria

Now we know EWMA might be a better calculation than RA since it is more sensitive and fits better in real-world situations. Next, we need to understand what a good or bad number is as a reference for ACWR.

The first study investigated the relationship between ACWR and injury risk in elite cricket fast bowlers². It found that an ACWR above 1.5 was associated with a higher injury risk.

This brings us to another OG paper, Gabbett's review¹, which proposed the criteria that are still widely used in the field today (see Figure 3)!

• Below 0.8 = Under-training zone

• Between 0.8 and 1.3 = Optimal zone

• Between 1.3 and 1.5 = Warning zone

• Above 1.5 = Danger zone

Figure 3. Criteria of ACWR (based on Gabbett¹)

One caveat is that the author notes: "It is possible that different sports will have different training load–injury relationships; until more data is available, applying these recommendations to individual sport athletes should be performed with caution."¹

Even so, these zones are now applied to almost every sport, even though they were proposed in 2016 from data in only a few sports.

We will discuss in the next section whether there is or has been a study investigating ACWR in other sports or across different variables.

Issues of current ACWR

---

Recent studies have discussed the issues with ACWR.

Issue 1: Inconsistent Study Outcomes

Pooled injury rates varied widely across sports; hockey showed an ES of 0.00 because its only study recorded zero injuries (see Figure 4).

Figure 4. Different sports or different variables for ACWR (Qin et al.⁵)

Issue 2: Unsubstantiated "Sweet Spot"

While the 0.8–1.3 interval is often considered optimal, its pooled estimate (ES = 0.56) has a very wide 95% CI (0.14–0.94), which indicates substantial statistical uncertainty, suggesting that current ACWR criteria alone are not a perfect fit for clinical decision-making or injury risk assessment⁵ (see Figure 5).

Figure 5. Pooled effect size for the 0.8–1.3 "sweet spot" with its 95% CI (data from Qin et al.⁵)

Issue 3: Lack of Causal Foundation (Association ≠ Causation)

The "Ice Cream and Shark Attacks" Analogy: Altering ACWR to prevent injuries without establishing causality is like banning ice cream sales to prevent shark attacks⁶ (see Figure 6).

Figure 6. A shared cause can make ACWR and injury move together (based on Impellizzeri et al.⁶)

Issue 4: Conceptual & Biological Flaws

Arbitrary Time Windows: The standard 1-week (acute) and 4-week (chronic) windows were selected for scheduling convenience (e.g., 7-day weekly microcycles) rather than reflecting actual biological adaptation rates or tissue-specific recovery times^7,8 (see Figure 7).

Figure 7. ACWR on the same simulated loads with three different window lengths

Issue 5: Mathematical & Statistical Pitfalls

1. Failure to Properly Normalize: The ratio fails to consistently normalize acute workload by chronic workload, instead introducing non-physiological noise and statistical artifacts^7,8. For example, an athlete going from 100 AU on day 1 to 200 AU on day 2 shows a ratio of 2.0 (a 100% jump), while another going from 1,000 AU to 1,100 AU shows only 1.1 (a 10% increase), even though both added exactly 100 AU. Because a ratio only captures relative change, the same absolute increase can look alarming or harmless depending on the starting point.

2. The Recovery Week Paradox: Following a planned recovery/tapering week, chronic workload drops, causing the subsequent ACWR to spike artificially; reducing workload further to avoid this spike leads to unintended undertraining (see Figure 8).

Figure 8. How a two-week taper inflates the ACWR of the following normal weeks

Is there a better way? "Suddenness"

---

Possibly, but what follows is a hypothesis, not an established method.

Big shout-out to Adam Virgile for sharing, discussion, and inspiration.

Given the five issues above, it's clear that there are a couple of conditions that have to be fulfilled for a modified version:

• Avoid ratios

• Drop the fixed 0.8–1.3 threshold

• Match time windows to the schedule

  
  

However, it can't be denied that a "spike" load will still increase the risk of injury.

In this case, instead of comparing chronic and acute loads, compare against each athlete's own normal variation to determine whether the change is a spike.

Lastly, maintain the EWMA's characteristics, as it is more sensitive and better suited to real-world situations.

To simplify these concepts, I name this metric "Suddenness"; it's basically a "z-score EWMA with the current and yesterday's change, with a different time window based on the schedule". Mathematically, because Δ EWMA = λ × (today's load − yesterday's EWMA), Suddenness measures how far today's load departs from the recent 7-day average, scaled by the athlete's own variability over the past 28 days. It is a standardized difference, not a ratio.

For example:

Below is a simulated 70-day player load dataset. The table shows days 25–45; the interactive dashboard below compares three scenarios on the same data. Because the z-score requires 28 days of ΔEWMA history, the first value appears on day 30.

• Load = Playerload

• EWMA 7D = Past 7-day window (match the game schedule)

• Δ EWMA = current day EWMA - yesterday EWMA

• Z-score = 28-day window of Δ EWMA (without the current value included)

Day	Load	EWMA 7D	Δ EWMA	Z-score / Suddenness
25	520	406.85	37.72
26	650	467.64	60.79
27	480	470.73	3.09
28	750	540.55	69.82
29	430	512.91	-27.64
30	560	524.68	11.77	0.14
31	480	513.51	-11.17	-0.24
32	0	385.13	-128.38	-2.22
33	510	416.35	31.22	0.47
34	600	462.26	45.91	0.76
35	450	459.20	-3.07	-0.07
36	550	481.90	22.70	0.35
37	370	453.92	-27.97	-0.51
38	0	340.44	-113.48	-1.92
39	1700	680.33	339.89	5.47
40	580	655.25	-25.08	-0.42
41	440	601.44	-53.81	-0.72
42	490	573.58	-27.86	-0.38
43	0	430.18	-143.39	-1.70
44	630	480.14	49.95	0.54
45	590	507.60	27.47	0.29

Table 1. Example data set, days 25–45 (the dashboard below compares three scenarios)

Table 2. Interactive dashboard

Application

---

Since multiple factors can cause injury, load management should cover multiple aspects as well.

Besides "spike load," the other key question is magnitude: how heavy is the load?

• "Magnitude" refers to "how heavy is the load for the player?"

• "Suddenness" refers to "how sudden is the load for the player?"

Combining the two gives us a better idea of where the stress comes from and how (See Figure 9).

And the next questions are "does this load make sense? And why? What action should we take?" "We have to know where we are before knowing where we are going!"

All these questions can help you take the "context" into consideration at that moment to interpret the data.

The green zone doesn't mean it's automatically better, and the red zone doesn't mean it's bad either.

Again, we need context to interpret the data.

Figure 9. Load Magnitude & Load Suddenness

Summary

Before we wrap this up, we have to emphasize again that injuries are multifactorial. ACWR was designed to flag load spikes, but it has often been interpreted as a tool that reduces injury rates. The fact is that it isn't "one size fits all"; it shouldn't be, and never will be when it comes to mitigating injury. That's why we have to develop a better, more reliable, and adaptable approach in this chaotic, multi-factor environment to make decisions and take action. Otherwise, we won't know why or what happens if we chase a metric that can't represent what we believe.

Suddenness is a theoretical calculation that hasn't yet been tested in research, which I will dive into in a series once I start my doctoral degree in a Sport Science program at the University of Miami. One known limitation: rest days (load = 0) create large negative changes every week, which inflate the standard deviation and can make true spikes harder to flag in schedules with many off days.

If you found this article useful, share it and tell me what you think in the comments.

Follow me on IG and LinkedIn so you don't miss the next part of the series.

Until next time!

Kenny Ho, MS, CSCS, CPSS

Sport Performance Coach

Founder, FILTER Human Performance

References

1. Gabbett, T. J. (2016). The training—injury—prevention paradox: should athletes train smarter and harder? British Journal of Sports Medicine, 50(5), 273-280.

2. Hulin, B. T., Gabbett, T. J., Blanch, P., Chapman, P., Bailey, D., & Orchard, J. W. (2014). Spikes in acute workload are associated with increased injury risk in elite cricket fast bowlers. British Journal of Sports Medicine, 48(8), 708-712. doi:10.1136/bjsports-2013-092524

3. Williams, S., West, S., Cross, M. J., & Stokes, K. A. (2017). Better way to determine the acute:chronic workload ratio? British Journal of Sports Medicine, 51(3), 209-210.

4. Murray, N. B., Gabbett, T. J., Townshend, A. D., & Blanch, P. (2017). Calculating acute:chronic workload ratios using exponentially weighted moving averages provides a more sensitive indicator of injury likelihood than rolling averages. British Journal of Sports Medicine, 51(9), 749-754.

5. Qin, W., Li, R., & Chen, L. (2025). Acute to chronic workload ratio (ACWR) for predicting sports injury risk: a systematic review and meta-analysis. BMC Sports Science, Medicine and Rehabilitation, 17(1), 285. doi:10.1186/s13102-025-01332-x

6. Impellizzeri, F. M., Tenan, M. S., Kempton, T., Novak, A., & Coutts, A. J. (2020). Acute:chronic workload ratio: Conceptual issues and fundamental pitfalls. International Journal of Sports Physiology and Performance, 15(6), 907-913.

7. Lolli, L., Batterham, A. M., Hawkins, R., Kelly, D. M., Strudwick, A. J., Thorpe, R. T., . . . Atkinson, G. (2019). The acute-to-chronic workload ratio: an inaccurate scaling index for an unnecessary normalisation process? British Journal of Sports Medicine, 53(24), 1510-1512.

8. Wang, C., Vargas, J. T., Stokes, T., Steele, R., & Shrier, I. (2020). Analyzing activity and injury: Lessons learned from the acute:chronic workload ratio. Sports Medicine, 50(7), 1243-1254. doi:10.1007/s40279-020-01280-1