HISTORY QUALITY
How missing workouts affect your training-load history
A model only sees the activities available to it. Before explaining a changed training trend, check whether the history changed.
Separate no run from no record
An empty date can mean you did not run. It can also mean the outing has not imported, was recorded under another sport or was removed while resolving a duplicate. Those situations have different practical meanings even if a chart currently shows the same empty space.
Start with the source activity list and compare dates, distance and start times. Do not create a replacement before checking whether the original is merely delayed. A duplicate introduced during troubleshooting can create the opposite error later.
A missing activity changes two summaries
These load values are illustrative inputs, not values inferred from the distances shown.
| Activity | Distance | Load units |
|---|---|---|
| Tuesday | 8 km | 60 |
| Thursday | 10 km | 70 |
| Sunday | 12 km | 80 |
| Complete week | 30 km | 210 |
| Without Sunday | 18 km | 130 |
The incomplete log contains 12 fewer kilometers and 80 fewer load units. A later restored activity changes the recorded history, even though the runner did not perform a new workout at the moment of restoration. A recalculated trend can therefore move retrospectively.
A present activity can still lack a load input
A distance total may include an outing for which a heart-rate-based load calculation lacks usable inputs. Inspect the activity rather than assuming every imported run contributes to every trend. A manually entered duration can preserve the log while leaving detailed sensor questions unanswered.
Runome's load calculations use their required activity and profile inputs. Values imported under a similarly named provider metric need not follow Runome's current method. Keep source load labels separate until the formulas and settings have been checked.
Repair the history before explaining the curve
Once the record is complete, compare equivalent date windows and note any recalculation. Keep a short audit note: which activity was restored, which date it belongs to and which summaries it affects. Preserve its original source if further troubleshooting is needed.
Follow the import data guide when the workout exists but a metric does not. Read date-window differences before assuming a weekly discrepancy is missing data. A lower model value by itself is neither an injury prediction nor a diagnosis of reduced fitness.