I would spend days looking back playing around. Would also be valuable to all us overthinkers.
DAILY FTP DETECTION
![]()
If my FTP isn’t being updated mid-interval every ride, am I even FTP’ing correctly?
I like to make sure mine is updated for each and every interval.
Im not sure how useful repeated “samplings” would be across a small time horizon except for relatively volatile periods like after illness or long break.
Pair that with %error of power meters and you may not see much difference for a while.
Your not supposed to be the one stirring the pot…that is supposed to be the rest of us.
It would be really interesting if we could go back in time and run FTP detection for any arbitrary historical date
Historic FTP detection would be more interesting. You could check what effect various training blocks had.
If not practical to make it available at every point in time, it would be helpful if it could be run and calculated for every month since a user has been with TR, Could be shown on the TSS graph. Could be called AI FTP Backcast.
That would be awesome and might help with people who want to use it for pacing.
funny how mine goes down for every interval.
Great news, it’s already there
Account > FTP History
Done!
This just gave me a brain-wave. How about a future FTP estimate per pedal stroke? Pedal harder and it starts going up, unless pedaling into excess fatigue at which point it starts going lower and will only start going up again when you pedal appropriately easier, all based on specified future objectives. No workouts or structure, just hop on the bike/trainer and pedal at a power that maximized upward movement of the needle! This would provide the FTP fixated with continuous on-bike gratification.
No its not. What’s there is the “old” AI FTP and its been noted by many users that with the switch in AI FTP models with the latest updates we now have an apples to oranges comparison with no understanding of whether changes in value are due to actual changes or model version changes.
What is being requested is a backcast of the latest model, or a tool to see what the value was on the new model at a particular point in time (but that could be expensive to allow users to run unlimited queries which is why im suggesting backcast the monthly history).
But how does it help you get faster today?
It seems like it only has downsides. People on the thread will complain that the new system lowered there previous FTP even though there current training number is higher.
Nobody is suggesting they remove previous FTPs just show what the values would have been historically under the new model.
How would the history of any metric be valuable to make you faster today? Yet many us enjoy and find value in tracking metrics that are comparable over time as part of our journey in trying to improve in lots of walks of life.
Not suggesting anything be taken away from anybody and no need to use the feature if you personally dont find value it in….
different strokes for different folks i guess. Historical FTP is pointless for me.