You tested FTP in January. Twenty minute test. Good pacing. Brutal final five minutes. The number came back 268 watts.
Now it is April and you are still training from 268.
Everything looks precise. Sweet spot intervals at 240. Threshold at 265. VO2 work above 300. TrainingPeaks gives clean numbers. Your power meter reads to the watt.
But the anchor is stale.
That is how power meters lie. Not because the device is wrong. Because the reference point is old.
FTP Is Not Fixed
Functional threshold power is not your identity. It is a moving estimate.
It changes with training, fatigue, heat, sleep, illness, weight loss, travel and life stress. It can improve across a block. It can fall when fatigue builds. It can look stable while your ability to repeat efforts changes underneath.
If you train for twelve weeks from one test, the zones become fiction.
Sometimes that fiction is too easy. You improve, but the zones do not move. Threshold work becomes tempo. VO2 work becomes hard threshold. You think you are training intensity, but you are not reaching the target.
Sometimes the fiction is too hard. Fatigue builds and your actual threshold drops. Now every threshold session is above threshold. You are not being tough. You are accumulating debt.
Why Cyclists Avoid Retesting
Cyclists avoid FTP tests because they hurt and because the result threatens the ego.
If the number goes up, great. If it goes down, suddenly every recent session gets questioned.
But avoiding the test does not protect your fitness. It just protects the story.
The better approach is to stop treating FTP as a verdict. It is calibration. Nothing more.
A bathroom scale tells you the number today. FTP testing tells you where to set the training zones today.
Field Tests Have Problems Too
The classic twenty minute test is useful, but it is not perfect.
Some athletes are excellent at suffering for twenty minutes and overestimate threshold. Others are poor testers and underperform. Pacing errors matter. Heat matters. Motivation matters. The 95 percent calculation is a shortcut, not a law.
Ramp tests have their own problems. They often flatter anaerobic athletes and punish steady diesel riders.
That does not mean testing is useless. It means the number needs context.
How Often To Recalibrate
For most serious amateur cyclists, every six to eight weeks is reasonable during a training block.
You do not always need a formal test. A hard race, a well executed threshold workout, a long climb or a maximal 40 minute effort can all inform the number.
Watch the sessions. If threshold intervals feel too easy for two weeks, your FTP may be low. If workouts that should be controlled keep turning into survival, your FTP may be high or your fatigue is too deep.
The power meter gives the data. Your training history gives the meaning.
The Hybrid Athlete Problem
Hybrid athletes have it harder because fatigue comes from more places.
A cyclist only tracking bike load misses the cost of squats, sleds, running and conditioning. You can show up to a bike session with fresh looking CTL and dead legs from yesterday’s gym work.
The power meter will not know why you cannot hold the numbers.
It just reports the failure.
The P247 View
Power is one of the best metrics in endurance sport. But precise data built on a stale assumption is still bad guidance.
The question is not “what is my FTP?”
The better question is “what number should guide the next block, given my current fitness, fatigue and goal?”
That answer changes.
The athlete who recalibrates keeps training honest. The athlete who clings to an old number trains inside a story.
X Thread
1/ Power meters feel objective. But if your FTP is stale, your zones are fiction.
2/ January FTP does not automatically belong in April. Fitness changes. Fatigue changes. Heat, sleep and stress change the number too.
3/ If FTP is set too low, threshold work becomes tempo. If set too high, every session becomes recovery debt.
4/ Retest or recalibrate every six to eight weeks during a block.
5/ Power is useful because it is honest. Do not corrupt it with an old anchor.