The 28-Day Load Baseline

A training week only starts to make sense when you compare it with what your body has recently been prepared for. A 28-day baseline gives the current week a practical reference point for duration, distance, elevation, and active days, without turning one noisy week into a conclusion.

A week in isolation tells you almost nothing

You ran 62 kilometers this week. Is that a lot?

The honest answer is: compared to what? For a runner building toward a first marathon, it might be a reckless jump. For someone deep in a training block, it could be a recovery week. The number itself carries no meaning until it is placed next to something. The most useful comparison is not a generic training plan, a friend's Strava feed, or a chart from a book. It is your own recent past.

Absolute numbers and generic plans share the same blind spot: they ignore what your body has actually been exposed to lately. Tissue, tendons, and aerobic systems adapt to the load they have recently received, not to the load a plan prescribes or another runner handles. The same week can be maintenance for one runner and overreach for another. It can also be maintenance for you in March and overreach for you in November.

Most runners already sense this. “That felt like a big week” is a real signal, but it is a noisy one. Memory compresses recent training, over-weights the last hard session, and quietly forgets the easy days. What is missing is not intuition. It is a stable reference point that intuition can be checked against.

That reference point is the baseline: a summary of what you have actually done over the last four weeks.

Why 28 days

The window has to balance two competing demands. It must be long enough that one unusual event does not dominate it, and short enough that it still describes the runner you are right now.

Twenty-eight days sits comfortably between those poles. A sick day, a race, or one unusually big weekend gets absorbed into the average rather than defining it. At the same time, the window does not reach so far back that it describes last season's fitness. What you did five months ago helped build the foundation, but what you did in the last month is closer to what your body is currently calibrated to.

The number also fits the way most runners train. Twenty-eight days is four weekly cycles. If your long run usually lands on the weekend, intervals sit midweek, and rest falls near the same day each week, four full weeks capture that rhythm evenly.

There is a research backstory too. Workload studies often compare the most recent week with a longer preparation window, commonly a rolling average of roughly 3–6 weeks.¹ The 28-day window sits inside that family of chronic-load comparisons.

This does not make 28 days a biological law. Acute:chronic workload ratio research has attracted real methodological criticism, and the exact window length is a convention rather than a universal truth.² The useful way to hold the 28-day baseline is as a well-motivated heuristic: simple, stable, personal, and close enough to recent training to be useful.

Rolling, not calendar

A subtle but important distinction: the baseline should be rolling, not “last month.”

Calendar months create arbitrary cliffs. A huge training week ending on January 30 shapes your January summary and then disappears from a February summary as if it never happened. Your body does not reset its adaptation state at midnight on the first of the month. Your reference point should not reset that way either.

A rolling window updates continuously. Each day, the baseline covers the 28 days before the current comparison window. The question stays consistent: how does the current week relate to the recent training that came before it?

Nothing falls off a cliff. Older training fades out of the window one day at a time, closer to the way its training relevance fades.

The mechanics are simple:

The simplicity is a feature. There is no complicated model to tune, just your own recent history kept current.

Four dimensions, not one number

The temptation is to collapse training into a single figure, usually distance, and baseline that. But a week can look normal in kilometers and still be unusual in ways a single metric hides. The baseline is more useful when it tracks a small set of complementary dimensions.

Duration is the most honest general proxy for load. Time on feet is relatively stable across terrain, pace, and conditions: ninety minutes is ninety minutes whether it was spent on a fast road loop or a slow technical trail. If you baseline only one thing, baseline this.

Distance stays in the set because it is familiar, motivating, and how many runners naturally think about volume. But distance is misleading alone. Ten trail kilometers and ten road kilometers can be different physiological events, which is exactly why distance needs the other dimensions around it.

Elevation gain is the quiet troublemaker. It often turns a normal-looking week into a heavy one without showing up in the distance column. Swap a flat long run for a hilly one and the kilometers barely move while the muscular and tendon load can change substantially. A week at 100% of your distance baseline but 170% of your elevation baseline is not a normal week.

Active days captures frequency, which changes the training stimulus independently of volume. Forty kilometers spread over six days is different from the same forty compressed into three. A sudden shift from five running days to three, at constant volume, means longer individual sessions and longer gaps. A shift the other way means less recovery between stimuli. Both are changes worth seeing.

The point of tracking all four is not dashboard maximalism. It is that the changes that matter often hide behind a stable headline number.

Reading the comparison

The output of a baseline comparison is intentionally plain: a ratio or delta per dimension.

This week might be at 115% of your recent norm in duration, 96% in distance, 160% in elevation, and five active days against a baseline of four. None of those numbers is a verdict. Together, they tell you where to look.

Rough interpretive bands help if they are held loosely. A week within roughly ±10–20% of baseline is usually consistent with your recent pattern. Well above that range is a spike: not forbidden, but worth noticing. Well below is either recovery, disruption, taper, or the beginning of reduced training load. Context tells the difference.

The crucial reframe is simple: elevated is not bad. Progression requires exceeding the baseline. If you never run above 100% of your recent norm, you are probably not building much. The baseline's job is not to keep you at 100%. It is to make sure that when you exceed it, you do so knowingly rather than accidentally.

There is a real difference between “I am pushing volume up 15% this block, on purpose” and “apparently I have been stacking 130% weeks without noticing.”

One week matters less than the drift across several weeks. A single 140% week followed by consolidation can be a training stimulus. Four consecutive weeks creeping from 110% to 135% is a trajectory. Trajectories, not isolated data points, are where both breakthroughs and problems tend to live.

What the baseline cannot tell you

A simple tool earns trust by being clear about its limits.

The baseline measures external load. It knows how long, how far, how high, and how often. It does not know intensity. A week of easy volume and a week of race-pace intervals can produce similar duration, distance, elevation, and active-day numbers while creating very different physiological costs.

It also does not know sleep, life stress, injury history, illness, travel, heat, or how you actually feel. Pairing the numbers with subjective check-ins is not decorative. It is the other half of the picture.

There is also a cold-start problem. The first four weeks of data describe a window that does not fully exist yet, so early comparisons are unreliable. The same applies after a long break. Returning from injury or an off-season, your baseline is legitimately low, and every comeback week can look like a mathematical spike. That is unavoidable. The baseline becomes meaningful again once it has a real month behind it.

The baseline is a lens, not a verdict. A 150% week during a training camp, or a deliberately loaded week before a taper, can be intentional and appropriate. The baseline flags the deviation; it does not judge it. Your plan, goals, and body response supply the interpretation.

Finally, guard against baseline worship. A metric that rewards staying at 100% forever rewards never improving. The failure mode is not only ignoring the baseline. It is optimizing for it.

Where this appears in your weekly report

In your weekly Run Briefing email, the 28-day baseline appears when the current week is compared with your recent training instead of a generic target.

Use it as the second read after the current-week snapshot. First notice what changed this week. Then compare that change with your recent baseline across load, active days, intensity context, and confidence.

The report does not turn the baseline into a decision for you. It uses the baseline to separate a normal week inside your pattern from a week that stands out enough to deserve closer reading.

From data point to context

The shift the 28-day baseline offers is small but useful. The question changes from “did I run enough this week?” to “does this week fit the pattern my body has recently been prepared for?”

The first question has no stable answer. The second has a concrete, personal, always-current one.

Getting started requires very little. Most training platforms already expose duration, distance, elevation, and active days. A spreadsheet with a rolling sum does the job well enough. Give it one full month of honest data, and from then on every week you run arrives with its own context attached.

Read the baseline as a reference frame, not a command. The current week tells you what happened; the 28-day baseline helps you decide how much weight to give it.

References