A lab threshold number is a hypothesis, not a plan

Testing produces a precise figure from an artificial protocol, and athletes then train to that figure for months without checking it against reality.

The report comes back with a number. A pace, a heart rate, a wattage, presented to two decimal places and accompanied by zones. It looks like measurement. Much of it is inference.

Threshold testing works by pushing an athlete through a series of increasing stages and sampling blood at each one to find where the readings start climbing steeply. The problem is that the answer depends on choices made in the protocol. Longer stages give a different result from shorter ones. The starting intensity matters. What the athlete ate that morning matters, because being low on stored carbohydrate suppresses the readings and flatters the result. Testing on a treadmill and racing on a road are different tasks for the same legs.

None of that makes the test useless. It makes it a snapshot taken under specific conditions, and the useful information is usually in how the snapshot changes across a season rather than in the absolute figure.

What happens instead is that the number becomes doctrine. The athlete trains at exactly that pace for eight weeks. The pace does not update, because retesting is expensive and inconvenient. Fitness improves, or worsens, and the prescription stops matching the athlete some time in week three without anybody noticing. Then the target is missed in a session, and rather than concluding that the number is stale, everyone concludes that the athlete is tired.

The field alternative is cruder and I trust it more. Regular time trials over a fixed course, or a session repeated identically every few weeks, measured the same way each time. It captures everything the lab cannot: heat, terrain, sleep, whether the athlete can actually hold this on a road with wind in his face. The number that comes out has no decimal places and no authority, which is exactly why it stays honest.

The best use of a lab is diagnostic. Something is wrong and you want to know where. It is a poor instrument for daily prescription, and it is sold as one because a printout with a graph on it justifies a fee in a way that a coach's stopwatch does not.

Test if you can afford to. Then treat the result as a question rather than an instruction.