While our extensive field testing gave us very good ideas about the merits and problems with each individual shirt, we also found it necessary to devise more controlled tests for some metrics, where shirts could easily be compared side-by-side. The details of how we tested for each individual metric are described below.
Testing for comfort was largely done in the field while out running. We made special note of what kind of stitching was used in all the seams, what sort of fabric weave was used, and how the fit affected our comfort levels, and then paid special attention to these factors while running. We also had multiple people try on each shirt, one after the other, to help verify our opinions about comfort.
This is a difficult metric to accurately assess for. Our assessments began with field testing on runs in warm temperatures, paying attention and keeping notes about our experiences. We paid special attention to the effects of shirts with mesh panels versus those with more solid materials, and the locations of those panels. Based on the accumulation of experiences and notes, we came up with scores for breathability for each product.
This metric, which constitutes half of the story of how well a shirt aids in evaporative cooling, was much easier to test. We simply dunked every shirt in water until it was dripping wet, then hung them all up to dry next to each other. While we opened windows and turned on the ceiling fan to allow some airflow, there was no sunlight or direct wind on the shirts. We then checked the shirts every 30 minutes until they were dry, arranging them and grading for drying speed.
When assessing for versatility, we tried to determine how effective the shirt would be for the maximum amount of other uses. Much of this testing took place by using the shirts for other activities such as hiking, climbing, and skiing. We wore every shirt on cold days under warmth layers to see how well they fit as a layer. We then looked at perceived durability, since we didn't have time to test all these shirts for the long haul, as well as fit, and how they affected the versatility of the shirt.
To test features, we first identified what features a shirt had, and then set about testing them. While it was not possible to test whether a shirt's UPF rating was working or not (we never got sunburned, FYI), it was possible to test features like the media loop on the New Balance Ice 2.0. We also tested the odor controlling properties of some shirts by comparing them after they had sat in the laundry basket for up to a week.