Hi @cryptopodium, 

Thanks for the post. Without looking at the specific activity that is causing this, it's difficult to know what is going on. Several different factors can contribute to changes in your Fitness and Freshness score. Here are the main ones:

1. The intensity of your activities has changed.
2. The frequency of your activities has changed.
3. Your heart rate zones were recently updated. (Updating your heart rate zones will only affect future activities)
4. If your activities are using power, your FTP may have recently been updated. You can update your FTP here: https://www.strava.com/settings/performance

If you still have questions or need any clarification, please submit a support ticket and include links to any activities where you started seeing this change. This will help our team to further investigate.  

It’s nothing to do with the fitness and freshness scores, I’m just talking about Strava’s estimate of my FTP (visible on the power curve). 

How can Strava estimate my FTP at 274w when my maximum 20 minute power is 264w? It’s been reasonably accurate in the past (at around 255w).

Hey @cryptopodium, 
Sorry about that.  For your FTP, we recommend entering your FTP for best results.  Please take a look at these help pages for additional info: 
Strava Training Glossary for Cycling
Power Analysis Features

You can enter your FTP on the website by navigating to the upper right and clicking on Settings and then My Performance. 

 I hope this helps!

Yeah, I could set it manually to what I think it is. But Strava is trying to estimate it - what algorithm is it using to do that, and why i sit getting it so wrong?

In the past this estimate has been fairly accurate, now it isn't. What's different?

Yes your 20 min max power is much lower but how hard have you pushed? Have you done FTP tests of at least 20 min?

I fairly regularly do max 20 - 23 minute efforts (I use Strava climbs for this, specifically the Epic KOM and Innsbruck). These are usually all-out. The point us that Strava has absolutely no basis from my data for overestimating my FTP by 15 or 20 watts... I have never sustianed those figures for an hour, neither have I ever sustained 105% of those figures for 20 minutes. It's obviously using some weird algorithm to do this, but we're not allowed to know what it is (the answers from Strava here have been utterly stupid, bordering on obfuscatory).

Right now the estimate seems to be about right, but a few months ago it was vastly overestimating. If I was going to guess what was happening based on the figures, I might assume that there was a simple bug whereby the algorithm was calculating FTP as 105% of a max 20 minute effort rather than 95% (i.e. using the inverse of the appropriate calculation).