ACWR for Hybrid Athletes: What the Acute:Chronic Workload Ratio Actually Tells You
If you have read anything about training load in the last decade you have run into the acute:chronic workload ratio, usually as a chart with a comfortable green band and an alarming red zone, and a strong implication that staying in the green keeps you in one piece.
The idea underneath it is genuinely useful. The green band is oversold. Here is what the number is, what it can honestly tell you, and the criticisms worth knowing before you let it drive decisions.
The arithmetic
ACWR = this week's training load ÷ your average weekly load over the previous four weeks
Acute is roughly the last seven days. Chronic is your rolling baseline, usually 28 days. If you have been averaging 400 units a week and this week you did 600, your ratio is 1.5.
There are two common ways to compute it. A plain rolling average treats every day in the window equally. An exponentially weighted moving average gives recent days more influence, which responds faster when your training changes and is generally considered the better of the two. The difference matters less than people arguing about it online would suggest.
What it is actually measuring
This is the part that gets skipped, and it is the whole point: ACWR does not measure how much you train. It measures how unusual this week is compared to what you have been doing.
Two athletes can both sit at exactly 1.0 with nothing in common. Someone training three hours a week and someone training fifteen both score 1.0 if this week looks like their last month. It is a change detector, not a volume measure, and it is blind to whether the baseline it is comparing against was sensible in the first place.
The numbers you will see quoted
Most charts mark 0.8 to 1.3 as a sweet spot and treat anything above about 1.5 as elevated risk. Those figures trace back to Tim Gabbett's work and the studies around it, largely in team sports such as cricket, rugby league and Australian football.
Two things about that lineage matter. They are associations observed across groups of athletes, not laws of physiology. And they come from sports whose training looks nothing like a hybrid athlete's week.
So: a research-based ramp-rate signal, not a prediction about you, and nothing here is medical advice.
What it does not tell you
It hides how much you are actually doing
A ratio of 1.4 on top of a large, well-established base is a different proposition from 1.4 on top of three weeks of barely training. The number is identical. Your situation is not.
The maths itself has been challenged
This is the part most articles leave out. The acute week sits inside the chronic window, so the numerator is part of its own denominator. Lolli and colleagues argued this mathematical coupling can manufacture correlations that look meaningful but partly reflect the construction of the ratio rather than anything about the athlete.
There is also plain regression to the mean. An unusually big week tends to be followed by a quieter one regardless of what any model says, which can make a threshold look predictive when it is describing arithmetic.
Attempts to reproduce the specific sweet-spot bands in other populations have been inconsistent. The general principle that sharp ramps are worth noticing has held up better than any particular cutoff.
It does not know what kind of load
A 1.4 driven by one extra long easy run is not the same stress as a 1.4 driven by adding a second heavy lifting day, even when the totals match. The ratio flattens everything into one number, which is what makes it convenient and what makes it crude.
It is only as good as the load number underneath it
For hybrid athletes this is the one that actually bites. If your load figure comes from your watch, it is built from heart rate, which means your lifting is either missing or represented by whatever your watch guessed from a wrist sensor during a set of squats.
Your ACWR is then a running ACWR wearing a hat. It will sit calmly at 1.0 through a week where you added two heavy sessions, because as far as the model is concerned those hours barely happened. The ratio is not wrong; it is faithfully describing a picture that is missing half the week.
So is it worth tracking?
Yes, if you demote it from verdict to prompt.
The useful question it answers is: am I asking my body to absorb a week substantially bigger than anything I have done recently? That is a real question, it has a real answer, and almost nobody can answer it from memory — which is exactly why a number helps.
What it cannot do is tell you that you are fine because you are at 1.1, or that you are in danger because you are at 1.6. Read a sharp ramp as a reason to pay attention to how you feel, look at what caused it, and decide deliberately rather than by accident.
Working it out yourself
You do not need an app for this. One row per session in a spreadsheet: date, activity, duration, and a load figure. If you need a load figure that covers both lifting and cardio, the session-RPE method in the previous post is the simplest honest option.
Then two columns: this week's total, and the average of the previous four weekly totals. Divide one by the other. You need about a month of data before the chronic figure means anything, which is the main reason people give up before it becomes useful. There is a step-by-step version if you want to calculate your ACWR by hand, lifting included.
Disclosure
I built LOAD partly because I got tired of maintaining that spreadsheet. It pulls cardio from Strava or Polar and lifting from Hevy or manual entry, fuses them into one number, and tracks the ratio over the combined figure rather than the cardio half. It is free for individual athletes, and it backfills 12 weeks from Strava when you connect so the chronic baseline is real from the first day.
It also shows the ramp rate as a signal to look at, not a green light or a red one. Given everything above, that is the only honest way to present it.
The methodology and citations are public at loadtraining.app/faq.