Whales roamed the oceans freely for centuries and scientists are interested in where they eat, live and reproduce to better understand these giants. Humpback whales can be identified by their flukes. Humpbacks have a unique fluke pattern that is consistent across their lifetime; it is as distinctive as a face.

Credit Tom Fernald/North Atlantic Humpback Whale Catalog

Credit Tom Fernald/North Atlantic Humpback Whale Catalog

Callosity patterns, the roughened skin patches on their heads, are used to identify North Atlantic right whales.

The ability to identify these leviathans makes it possible to track where they go. It was recently discovered that one North Atlantic right whale, approximately the size of a school bus, was sighted in Madagascar and again in Brazil. It had traveled a quarter way around the world, a whopping 6,500 miles.

Figuring out how far whales could travel was possible once whales could be individually identified. Without a way to identify a whale, researchers could not be sure that the whale in one location was the same in a different location. Whale identity made it possible to determine the population of different species.

Metrics allow scientists to understand whale behavior, where they feed, breed, and migrate especially now that the Gulf of Maine, where whales feed, is one of the fastest warming places on earth. Scientists have declared an “unusual mortality event” due to the high number of humpback deaths during the first eight months of this year. Researchers estimate the whale population from whale sightings at the same locations every year. They use 36,000 photos of 9,500 animals to track the whales. Digital algorithms help, but in the end matching still requires a person looking through thousands of images one-by-one.

A lot of emphasis is put on metrics to assess whether things are going in the desired direction. Using the right metrics, ones that provide the information to give an accurate picture is even more important. There may be times when the metric doesn’t exist. In those cases, a combination of metrics can be used as a proxy. That can get complicated especially if the measurement becomes part of a bigger equation.

Creating a new metric that is specific and accurately measures a function may serve your purpose better in the long term. Just as important is collecting reliable data for the measurements.

Have you looked at your metrics to see if they are capturing the data you want or are you using a series of proxies? If you haven’t looked recently, it may be time to check for changes in your data sources because you might not be measuring what you thought.