How does Base grade an endurance session?

By comparing what the session asked for against what your watch actually recorded. Where the watch marked laps, each rep is graded individually against the target band. Where it did not, the session is graded on time spent inside the band, or on its overall average. The basis used is always stated rather than left implicit.

Why is my ride graded on an average when my run was graded per rep?

Because the health store exposes cycling power only as a session-level statistic — there is no per-lap power. A run has a real pace for every lap, so each rep can be graded. A ride falls back to the session average, and Base says so rather than presenting an average as if it were per-rep detail.

When does grading happen?

Automatically, once the session syncs from your watch to your health store and Base next opens the Endurance tab. It only inspects a workout in detail when there is an ungraded target of the same discipline on the same day, so it is not re-reading your whole history on every launch.

What if I did the session but it was not graded?

Either there was no planned target to grade against, or the workout has not reached your health store yet. A session with no target still counts toward your week and ramp rate; it simply has nothing to be compared to. The tab distinguishes these two cases rather than showing both as a gap.

Does the grade change later if my threshold changes?

No. The grade is stored as a snapshot at the time it was computed, including the band it was graded against. A grade that shifted retroactively when you updated your threshold would take the whole training log with it, so what you saw at the time is what stays.

How Base Grades an Endurance Session

When it happens

Automatically. Once a session syncs from your watch into your health store and you next open the Endurance tab, Base looks for an ungraded target of the same discipline on the same day and grades against it.

It only opens a workout for detailed inspection when such a target exists. Reading every workout in the window on every launch would be dozens of round-trips to discover that nothing needed doing.

While it works the tab says "Grading a session that just synced…" rather than showing an empty result.

The three bases

The Endurance tab. A graded session shows its verdict against the band it was prescribed at; an ungraded one still counts toward the week, and the tab distinguishes the two.

Basis	Used when	What it grades
Rep by rep	The watch marked laps	Each rep against the band
Time in band	Continuous session with a pace series	The share of measured time inside the band
Session average	No per-rep data available	The overall figure against the band

Why laps rather than inferred intervals

When the watch marked laps, those laps are the reps. The watch ran the workout and recorded exactly where each effort started and stopped.

Base uses them directly and does not try to re-derive intervals from a pace stream, because that would be guessing at something already recorded — and guessing badly, since the boundaries are precisely what a pace stream makes ambiguous.

Why a ride is different from a run

This is the asymmetry people notice first.

A run has a genuine pace for every lap, so every rep gets a real number. A ride does not, because the health store exposes cycling power only as a session-level statistic. There is no per-lap power to read.

So a ride's reps come back unmeasured and the grade falls back to the session average — and says so. The alternative, spreading a session average across reps and presenting it as per-rep detail, would be fabricating precision that does not exist.

Partial splits are dropped

For time-in-band grading on a continuous session, a final partial mile is excluded. Extrapolating a pace from the last forty seconds produces a number with far more error than the splits around it.

That is why the time-in-band figure is documented as a share of measured time rather than of the whole session.

Reading the result

The session result screen shows the verdict, the time in the band, a rep-by-rep table where one exists, and the threshold it was graded against.

That last card is worth checking when a grade surprises you. A session graded against a threshold set four months ago is being held to a standard from four months ago — in either direction.

Grades are snapshots

A stored grade never changes. It keeps the band it was graded against at the time.

This is deliberate. If grades re-computed against your current threshold, then raising your threshold would retroactively turn a string of good sessions into failures — and the training log would stop being a record of what happened. The result screen reads from the stored snapshot rather than doing a fresh device read for the same reason: a grade that changed when health access lapsed would take the log with it.

Ungraded sessions

Two distinct cases, and the tab distinguishes them:

No target — you trained without planning. It counts toward the week and the ramp rate, with nothing to compare against.
Not yet synced — the workout has not reached the health store. It will grade itself when it does.

Neither is shown as a failure.

Common mistakes

Expecting per-rep power on a ride. The data does not exist at that granularity. The average is the honest answer.

Reading a poor grade without checking the threshold. A stale threshold produces bands that are wrong in a consistent direction.

Planning one session and doing another. The grade compares against what was asked for, so a swap reads as a miss.

Assuming an ungraded session did not count. It counts toward the week and the ramp rate.

Grading yourself on time in band for an interval session with no laps. If the watch did not mark reps, the interval structure is invisible to the grade regardless of how well you executed it. Use the lap button.

Treating a single grade as a verdict on fitness. Weather, terrain and heat all move pace at a fixed effort. The trend across sessions is the signal.

What to do next

Use the lap button on your next interval session. It is the difference between a rep-by-rep grade you can actually learn from and a session average that tells you almost nothing about how the reps went.