Built a single "Hybrid Score" (harmonic mean of a strength index and a cardio index) to answer whether I'm actually a balanced athlete, or just okay at two things separately
I lift and I run, and I wanted one honest answer to a question two separate apps couldn't give me: am I actually balanced across both, or just decent at each in isolation.
The methodology I landed on: compute a strength index and an engine (cardio) index, each 0 to 100, then combine them with a harmonic mean instead of a simple average. The reason is deliberate: an average lets a strong pillar mask a weak one (a 90/40 split averages to 65, same as a 65/65 split, even though the athlete is nowhere near as balanced). Harmonic mean punishes the gap, so a lopsided profile scores visibly lower than a balanced one at the same total effort. Can't fake the number by being great at only one half.
Inputs right now: logged strength sets (weight, reps, RPE) feeding an estimated strength curve, and Strava-synced runs/rides feeding the cardio side (VO2max-adjacent estimate). No wearable/recovery data yet (HRV, sleep, RHR), that's the obvious next layer, but I wanted the strength x cardio relationship solid first before adding recovery as a third dimension.
I'm running this as a real self-experiment: training for a HYROX in August (goal sub-1:15), logging everything, watching the score move week to week against the actual race outcome as the real-world validation of whether the number means anything.
Curious what this community would track or weight differently. Harmonic mean over average, or would you reach for something else to penalize imbalance? And what's the first non-training-day metric you'd want fused in (sleep, HRV, something else)?