How to use the COT Shape tab and controls

Use this tab when you want to review or refine the cost-of-transport curve in calibration.

This is not the first calibration step.

Most users should come here only after:

  • finishing Ingest
  • checking Speed Shape
  • checking ΔHR Shape

What this tab is for

The COT Shape tab answers this question:

Does the modeled COT curve stay aligned with the FIT-derived proxy across the grade range in the dataset?

Read it this way:

  • compare the modeled curve with the FIT-derived proxy
  • check the COT Shape Score
  • change controls only when the chart and score both support the same diagnosis

If you cannot name the mismatch you are correcting, do not change the controls.

Step 1: Open the COT Shape tab

From the Calibration Workspace, click COT Shape.

The tab has two main parts:

  • the chart and advisor on the left
  • the controls on the right

Step 2: Read the chart before touching anything

Start with the chart, not the controls.

Check these first:

  • does the modeled curve follow the proxy across most bins
  • is the mismatch mainly uphill, downhill, or broad across the curve
  • is the mismatch repeated across bins or confined to noisy edges

If the chart already tracks the proxy closely enough, do not keep editing.

Step 3: Check the advisor

Use the advisor as a confirmation layer, not as the first thing you optimize.

Pay attention to:

  • score
  • label
  • coverage
  • mismatch notes

Use it like this:

  • if coverage is weak, avoid aggressive tuning
  • if the mismatch is concentrated in one region, make only one targeted change
  • if the chart is acceptable and the score is not perfect, do not keep editing for score alone

The advisor is most useful when it confirms a mismatch you can already see in the chart.

Step 4: Make small, targeted adjustments only after the earlier tabs support the current baseline

Before changing anything:

  • make sure the earlier tabs are not the real problem
  • make sure you can identify which side of the curve is off

Good reasons to adjust:

  • the uphill side is consistently high or low against the proxy
  • the downhill side is consistently high or low against the proxy

Bad reasons to adjust:

  • one noisy bin is odd
  • the score is not perfect
  • you want to try the controls without a specific target

Step 5: Recheck the chart after every small change

Do not make several edits before re-reading the chart.

Use this rhythm:

  1. identify the mismatch
  2. make one small change
  3. recheck the chart
  4. recheck the score and notes
  5. stop when the specific mismatch is no longer the main issue

If one change improves one region but worsens the rest of the curve, back it out and reassess.

Step 6: Reset when you lose the thread

Use Reset on one control, or Reset All, when you lose track of which edit caused the current result.

Reset when:

  • you changed several controls without confirming each step
  • the latest edit fixed one region and damaged another
  • the dataset is too weak to support more tuning

Each control change should have one clear reason.

Recommended order

Use this sequence:

  1. Read the chart.
  2. Read the score and mismatch notes.
  3. Confirm the mismatch is broad enough to matter.
  4. Make one small targeted change.
  5. Recheck the chart and score.
  6. Stop once the specific mismatch is no longer the main issue.

When not to use this tab much

Do not spend much time here if:

  • FIT coverage is weak
  • the advisor says the dataset is insufficient
  • Speed Shape or ΔHR Shape still shows the main unresolved issue
  • you are still trying to establish the first usable route preview

In those cases, improve the FIT folder or return to the earlier tabs first.

What to do next

After reviewing COT Shape, continue with:

Still need help? Send a Little Note to Backpacking Light Send a Little Note to Backpacking Light