COMPARISON METHODS
Choose steady segments for a running-efficiency comparison
A whole-run average blends everything in the outing. Select a segment that matches the question, and keep the selection rule consistent.
Select by structure, not by the result
Decide which portion represents the effort you want to study before looking for its best efficiency value. You might choose a continuous section of a repeated flat loop, using the same start and finish landmarks on each visit. Document the rule and any exclusions.
A segment chosen because it has the lowest heart rate will tend to answer a different question from a segment chosen because it matches the workout structure. The same is true if you move the endpoint until an unfavorable late section disappears.
Separate the parts of a mixed outing
Consider this illustrative 50-minute record. The labels describe the example's structure; they do not prescribe a workout.
| Time | Recorded context | Comparison use |
|---|---|---|
| 0–10 min | Settling into the run | Keep separate from the chosen steady portion |
| 10–25 min | Continuous loop section | Possible first comparison segment |
| 25–27 min | Stopped at a crossing | Do not hide the pause handling |
| 27–42 min | Continuous loop section | Possible second segment if route and pace match |
| 42–50 min | Deliberately faster finish | Analyze as a different effort |
The two 15-minute portions are candidates, not automatically equivalent. Check terrain, direction and speed before using them together. Equal duration does not guarantee equal distance or effort.
Align the streams
Speed and heart rate need to describe the same time window. If heart rate is absent for part of one portion, shorten both signals to a defensible shared window or leave the comparison unresolved. Do not use speed from the complete activity with heart rate from just one favorable segment.
Specify whether you divide segment averages or average sample-level ratios. Both are defined calculations, but they can yield different values. For a repeatable personal comparison, use the same method and preserve the raw inputs.
Report the selection with the result
When reviewing the available charts in Runome, note the start and end of the portions you are comparing. Describe the excluded stops and any route mismatch. A result such as “two continuous 15-minute sections, same loop direction” tells a future reader much more than an unexplained efficiency percentage.
Drift versus decoupling shows why the selected inputs matter. The warm-up example demonstrates how changing the analysis window changes a number even when the underlying recording is untouched.