Team Management · Season Leaderboard
Speed Index
Speed Index is the leaderboard metric used in Logbook Companion team analytics. It blends raw speed and power-to-weight efficiency into a single score so coaches can scan a group quickly, then drill into split, efficiency, and workload detail before making lineup decisions.
Why This Metric Exists
Rowing evaluation always carries a tension between two useful truths:
- Lower split reflects who is moving the machine or boat fastest.
- Higher watts per pound reflects who is producing power efficiently relative to body weight.
In plain coaching terms, efficiency matters because it can explain why one athlete may be more boat-relevant than their absolute power alone would suggest. But watts per pound is already derived from watts, and watts is derived from split. Speed Index is built around that idea: once both values are standardized into the same z-score language, they can be blended evenly in a way that is explicit rather than accidental.
That is why a blended metric makes sense at all. Coaches already make this judgment informally. Speed Index simply makes the tradeoff visible in one place so the first pass through a leaderboard is more honest than using split alone, without pretending one number should replace actual coaching judgment.
We are also intentionally aware that this does give raw power extra voice. Split already reflects output, and the relative-power side introduces power again through a bodyweight lens. That overlap is not a bug in the explanation. It is a conscious product choice because we want actual speed to stay primary while still giving explicit credit to athletes who produce more watts for their size.
Why Not Use Raw Split Alone?
A pure split ranking is simple, but it tends to over-reward athletes whose absolute horsepower is carrying the result while under-rewarding athletes who are producing unusually strong output for their size. In rowing that matters because body mass is not free. The boat has to carry it.
Speed Index keeps that tradeoff visible. It does not say efficiency is more important than speed. It says both signals are worth tracking, and that the product deliberately allows the power side to matter more than a pure efficiency-only adjustment would, because relative power is part of the coaching question rather than a side note.
Why The Z-Score Step Exists
Split and watts per pound do not live on the same scale. One is a time measure. The other is an efficiency measure. If you try to combine the raw numbers directly, whichever column happens to have the wider spread on that workout can dominate the result for mathematical reasons rather than coaching reasons.
The z-score step fixes that by translating both signals into the same relative language: how far above or below the group average an athlete sat on that workout. Once both inputs are expressed as comparable standings inside the same cohort, blending them becomes defensible.
How It Is Calculated
Each scored workout is turned into a per-workout Speed Index before it is averaged into the leaderboard view. The workflow is:
Step 1 - Standardize speed
The system computes a speed z-score from each athlete's average split for that workout.
speedZ = -(split - meanSplit) / stdSplit
Lower split is better, so the value is negated to make faster athletes score higher.
Step 2 - Standardize efficiency
The system computes an efficiency z-score from watts per pound on current leaderboard surfaces.
efficiencyZ = (wplb - meanWplb) / stdWplb
Higher watts per pound is better, so no sign flip is needed.
Some coaches prefer to think in W/kg instead. That is valid. Other product surfaces may show both W/kg and W/lb; this page focuses on the current leaderboard implementation, which uses W/lb.
Step 3 - Apply the equal blend
The two standardized values are combined with equal weighting. That keeps the formula simple, but it does not mean the output is neutral with respect to power. Because split already reflects output, the relative-power side intentionally gives the final score another way to reward raw power through a bodyweight-normalized lens:
rawScore = (speedZ * 0.50) + (efficiencyZ * 0.50)
This is the current product model. Think of it as a standardized average across two related signals where the overlap is intentional: speed leads, and relative power still gets a real vote. Coaches cannot change the weighting in settings.
Step 4 - Normalize to a 0-100 score
The raw blended values are min-max scaled so each workout yields an easy-to-read score.
speedIndex = ((rawScore - minRaw) / (maxRaw - minRaw)) * 100
The best scored athlete in that workout lands near 100, the lowest near 0, and everyone else falls in between.
How The Leaderboard Uses It
The leaderboard does not use a hidden rolling workout count anymore. It averages per-workout Speed Index values across the assignments that match the currently selected page filters.
| Control | What it changes | Notes |
|---|---|---|
| Last week | Only workouts from the last 7 days | Useful for short-form form checks |
| Last 4 weeks | Only workouts from the last 28 days | Default team analytics view |
| Season | Workouts from August 1 to today | Matches the academic-season model |
| All time | All scored workouts in scope | Best for broad historical review |
Team, squad, tier, and Tests Only filters all rerank the visible group. That means Speed Index is always relative to the athletes and workouts currently on screen.
How To Read The Score
| Range | Interpretation | Coaching read |
|---|---|---|
| 80-100 | Top end of the visible group | Usually strong speed with enough efficiency to hold position |
| 50-80 | Above group average | Check split plus relative-power columns (W/lb here, and W/kg where available) to see which trait is driving the score |
| 20-50 | Below group average | Often points to development opportunity in speed, efficiency, or both |
| 0-20 | Bottom end of the visible group | Interpret relative to squad level and training age, not as an absolute label |
Relationship To Other Columns
| Column | Meaning | Best use |
|---|---|---|
| Speed Index | Standardized blend of speed and relative power | Quick first-pass ranking |
| Latest Split | Most recent scored split | Check near-term trend |
| Avg W/lb | Mean watts per pound across visible work | Read efficiency directly; if you prefer metric framing, treat this as the same relative-power idea that some result views also expose as W/kg |
Key Limitations
- It is relative, not universal. A Speed Index of 80 in one filtered group is not automatically comparable to an 80 in a different group.
- Weight is required for the efficiency side. Missing weight data means the athlete cannot receive a Speed Index for that workout.
- Small groups exaggerate movement. With only a few athletes, z-score based normalization can swing harder than it will in a full team cohort.
- The metric is a guide, not a final selection rule.Coaches should still read splits, lineup context, technical quality, and boat fit alongside the score.