About - Beyond the Metric

Beyond the Metric is the publication of BTM Analytics, a Canadian company working on measurement, inference, and decision making in human performance.

We can measure a great many things. The harder question is what those measurements actually mean.

Can the result be repeated? Does it represent what we think it represents? Does it hold in another season, another population, or another environment? And is the evidence strong enough to support the decision being made from it?

A number can be measured perfectly well and still be a poor basis for an inference.

That gap is what this publication works on.

The question is not simply whether a number is correct. It is how much confidence that number earns, how far the inference can reasonably travel, and whether that is enough for the decision being made.

We draw on methods from fields that are not always brought together in human performance analysis: information theory, signal processing, physiology, dynamical systems, statistics, and measurement science. Then we test whether those tools reveal structure that conventional approaches miss, and whether that structure survives attempts to break it.

The process is deliberately transparent. The measurement problem is defined. The assumptions are made explicit. Methods are tested for robustness, sensitivity, repeatability, and, where possible, replication before conclusions are drawn.

Not every method survives.

That matters.

The result is a growing body of work with its standards in the open: named data sources, explicit assumptions, stated limitations, reproducible methods, and interactive figures that can be interrogated rather than simply viewed.

Hockey is where much of this work began because its public record is unusually long, detailed, and repeated. That makes it an excellent environment for asking whether a measurement holds up across players, teams, seasons, and contexts.

But the question was never about hockey.

The same problem appears anywhere human performance is measured and turned into a decision, in sport, tactical and occupational settings, and applied physiology.

The number changes.

The underlying questions do not:

What does it actually measure?
Can we reproduce it?
Where does it stop generalizing?
And what decision does the evidence justify?

Standards

The failures get published too

When a finding breaks under robustness, sensitivity, or replication testing, that result matters too. We publish those failures alongside the findings that hold. A method you only ever see succeed is a method you cannot properly judge.

Transparent methodology

Sample sizes, assumptions, statistical methods, uncertainty, and limitations are stated in every analysis. The analytical work comes before the story. Methods are selected because they address the measurement problem, not because they produce the cleaner result. The methods section is not optional.

Replication before certainty

A result observed once is evidence, not a law. Where the data allow it, findings are tested across independent seasons, samples, populations, or contexts. Results that do not replicate are treated accordingly rather than promoted into general conclusions.

Team

Martin Poirier, PhD

Founder and Lead Analyst

PhD in Human Kinetics from the University of Ottawa. CSEP Clinical Exercise Physiologist and CSEP High Performance Specialist since 2011, with more than 40 peer-reviewed publications and research presented at national and international scientific conferences, including the American College of Sports Medicine Annual Meeting, Experimental Biology, and the Canadian Society for Exercise Physiology Annual Scientific Conference.

His work in human performance started with athlete testing and coaching. He trained more than 150 personal training clients and spent seven seasons as lead strength and conditioning coach with the Ottawa Mavericks Volleyball Club, working across athlete development, physical testing, and performance programming.

He then spent eight years in the Human and Environmental Physiology Research Unit at the University of Ottawa under Dr. Glen Kenny, studying how the body responds to heat and the factors that change that response: age, sex, fitness, diabetes, hypertension, dehydration, environmental conditions, and protective clothing.

One result from that work became especially important to how he thinks about measurement: a local measurement of sweat rate does not necessarily represent whole-body heat loss. The measurement itself can be valid while the inference drawn from it is wrong.

That distinction has shaped his work ever since.

Since 2023, his professional research has focused on human performance and national physical employment standards for tactical and enforcement populations, where measurements can influence consequential decisions about readiness, eligibility, and job requirements.

That work is conducted in his professional research role, not through Beyond the Metric. What carries over is the question, not the data: what does the measurement actually tell us, how far can we trust the inference, and is that enough for the decision being made?

ResearchGate LinkedIn

Adam Klockars

Co-Founder and CTO

More than 10 years building enterprise platforms, APIs and developer ecosystems. Senior Engineering Manager at Zoom until early 2026, leading US engineering for the App Marketplace: app onboarding, review and publishing consolidated into one developer flow, a rules and event driven automation framework, partner monetization and billing, and the compliance guardrails around third-party access.

Seven years before that at SurveyMonkey, first running the APIs and integrations group, where he took more than 15 first-party integrations to market including Salesforce, Slack, Tableau and Power BI, and later on data-intensive research products. BASc Computer Engineering, University of Waterloo.