The distinctions in score and stated value proposition need explaining.
70/100 (Strong, with nuance)
68/100 (Moderate-to-Strong)
I would argue the ratio of 70/100 to 68/100 is 1.02 (or 0.97 if you prefer) and this is a distinction without meaning.
domofutu · 59m ago
Great question and fair push-back. Two quick clarifications:
The ETCS isn’t a ratio scale. A 70 vs 68 isn’t “1.02× more true”; it’s a weighted confidence score that blends consistency of RCTs/meta-analyses, effect size, heterogeneity, external validity (who benefits), dosing clarity, and risk trade-offs. Small gaps (especially near band edges) should be read as “roughly comparable confidence,” not a meaningful quantitative jump.
Why 70 (“Strong, with nuance”) for fractures vs 68 (“Moderate-to-Strong”) for respiratory infections? Because the fracture claim is consistently positive when paired with calcium in older/institutionalized adults (clear population + pairing guidance), whereas the respiratory finding shows a modest, baseline-dependent benefit (strongest only when deficient; daily/weekly dosing beats bolus; more heterogeneity and large neutral trials in sufficient populations). Same neighborhood numerically, but the value proposition differs (i.e., Fractures: act when risk is high; pair D3 with calcium; practical, repeatable benefit; Respiratory: correct deficiency first; expect at most a modest protective effect, not population-wide gains).
To make this clearer going forward, I’ll add a short ETCS legend in each post and note that ≤3–5-point differences within a band are “near-ties.” Thanks for the nudge. This is exactly the kind of reader feedback that I'm looking for.
70/100 (Strong, with nuance) 68/100 (Moderate-to-Strong)
I would argue the ratio of 70/100 to 68/100 is 1.02 (or 0.97 if you prefer) and this is a distinction without meaning.
The ETCS isn’t a ratio scale. A 70 vs 68 isn’t “1.02× more true”; it’s a weighted confidence score that blends consistency of RCTs/meta-analyses, effect size, heterogeneity, external validity (who benefits), dosing clarity, and risk trade-offs. Small gaps (especially near band edges) should be read as “roughly comparable confidence,” not a meaningful quantitative jump.
Why 70 (“Strong, with nuance”) for fractures vs 68 (“Moderate-to-Strong”) for respiratory infections? Because the fracture claim is consistently positive when paired with calcium in older/institutionalized adults (clear population + pairing guidance), whereas the respiratory finding shows a modest, baseline-dependent benefit (strongest only when deficient; daily/weekly dosing beats bolus; more heterogeneity and large neutral trials in sufficient populations). Same neighborhood numerically, but the value proposition differs (i.e., Fractures: act when risk is high; pair D3 with calcium; practical, repeatable benefit; Respiratory: correct deficiency first; expect at most a modest protective effect, not population-wide gains).
To make this clearer going forward, I’ll add a short ETCS legend in each post and note that ≤3–5-point differences within a band are “near-ties.” Thanks for the nudge. This is exactly the kind of reader feedback that I'm looking for.