TSS & the Garmin Recovery feature

I’ve been using a garmin 520 for nav/data recently. The 520 has a ‘recovery advisor’ which flashes a recovery time (in hours) AFTER a ride and flashes a recovery condition at the BEGINNING of a ride. Something pithy like ‘Good’ or ‘Fair’. I’ve always dismissed it as hokum & have never talked to anybody who takes it seriously.

TSS, on the other hand, is fairly broadly accepted. But I’ve always said TSS OVERestimates long low intensity efforts and UNDERestimates short intense efforts. So it’s OK as a relative gauge if you typically train within a fairly narrow band of intensities but not so good if you normally do a bunch of Z2 Mile Piles and then try to switch over to a similar TSS load of Threshold work. Or even sweetspot work. You’ll crush yourself.

Is Garmin 520’s Recovery Advisor a better measure of training stress than TSS? Here is why I ask…and also a sideways humble brag…

A couple weeks ago I did a gravel race. Here is time spent in each zone:

Recovery: 19 min
Aerobic: 50 min
Tempo: 1 hour 10 min
Threshold: 2 hours
Super Threshold: 45 min
Aerobic Capacity: 57 min
Anaerobic Capacity: 20 min

This weekend I did a time trial. Here is time spent in each zone:

Recovery: 1 hour 8 min
Aerobic: 4 hours 40 min
Tempo: 3 hours 52 min
Threshold: 1 hour 57 min
Super Threshold: 6 min
Aerobic Capacity: 1 min
Anaerobic Capacity: 0 min

TSS for the latter ride was 553. TSS for the former 729. That’s about what my expectation would have been: the second ride had more training stress. Therefore, I would say it required more recovery.

Ok. So Garmin 520 fitness advisor said 42 hours recovery for the first ride…which I would normally ignore but happened to see as I was using the Garmin to navigate back to my car. After the second, 729 TSS ride I was trying to take a snapshot of the Garmin right after the race & it flashed 18 hours recovery time. That’s pretty much diametrically opposed to what TSS analysis would lead me to believe.

However the passage of time has proven Garmin Fitness Advisor to be exactly correct. Maybe not down to the hour but certainly with respect to relative recovery time post event. After the gravel race I don’t know that I felt fully recovered even after a week of taking it easy. After the time trial I feel ready to go right now.

Is Garmin Fitness Advisor under rated? Should we be paying more attention to it and less attention to TSS?

Firstbeat Recovery Advisor is used under license:

If you poke around the Firstbeat website there might be a white paper on the science.


Now I’m curious. I know FirstBeat utilizes HRV if your device can, I wonder if their “Recovery Advisor” uses HRV in it’s calculations if available? I like my Whoop but I’d rather ditch it and use a real fitness tracker if possible.

Something is off with your FTP settings in Garmin because those TSS numbers do not align at all to the efforts you are describing.

@RONDAL, well maybe that’s part of my problem, then!

The TSS numbers are taken from another platform (not TR or Garmin)…one that I’m sure we’re all familiar with. Can you give me some more guidance? What is it that catches your eye as being wrong?

Maybe it’s something I can fix…or maybe I just shouldn’t use that other site to provide TSS data.

I believe it does. My understanding is that HRV is obtained from the chest strap paired to the Garmin.

I found this white paper:


Go into trainingpeaks and go check your Zone Settings for FTP and ensure the FTP that TP has for you is the same for TR and Garmin.
If it’s different, as it appears, it will cause issues with the calcs that are occurring. You’ll need to delete the files and then re-upload to TP as they dont seem to be able to back-calculate if changes occurred.

Alternatively just check the TSS #'s generated by TR for the rides assuming that TR’s FTP is accurate.

All recovery data requires weeks to months of input to arrive at a reasonable prediction. But also realize that it’s not just tss that identifies recovery, your atl and ctl need to be factored in. Theoretically you could do an identical tss ride the same way on different weeks and the second ride could require decreased recovery time due to ctl increasing. If you haven’t been recording a lot of data with garmin related to tss and recovery your output is limited.

Just some interesting follow up observations re: Garmin 520 Recovery Advisor. Recall from the original post that a 100 mile gravel race with some pretty intense pacing scored 553 TSS but Recovery Advisor said 42 hours of recovery…meanwhile a 210 mile ride that went above threshold for 7 minutes total scored TSS of 729 but Recovery Advisor said 18 hours of recovery.

So I think the primary difference in recovery after those two rides was not due to the TSS of the rides themselves but rather due to the accumulated training stress in the days leading up to the rides. A few days previous to the 100 mile gravel race I had done a TrainerRoad workout Seiler -1 (he of polarized training fame) that was pretty intense. Recovery Advisor was keeping a rolling total, I think. Prior to the 210 mile race I was tapered fully because I took that race seriously.

Last week most of my workouts were recovery. Recovery Advisor reported recovery times of 8 to 10 hours. Saturday was a big ride:

52 minutes threshold
27 minutes super threshold
40 minutes aerobic capacity
15 minutes anaerobic capacity

And I set some new threshold records.

Even though I was in a very recovered state Recovery Advisor told me I should take 72 hours to recover. So even though it was a much shorter ride in terms of time and distance (60 miles vs 100 & 200) it was a much bigger recovery time. Ok…next day I did a 90 minute TR recovery ride & after that ride concluded Recovery Advisor told me I needed 47 hours of recovery. That was basically the 72 hours Recovery Advisor told me I needed after the intense ride less the time between when the intense ride ended and the recovery ride happened.

Apparently Garmin 520 Recovery Advisor has some smarts. Whether or not 72 hours is the right recovery time after that ride, who knows. It’s potentially a nice feature in what is kind of a down scale bike computer GPS product, though.

I would have to think your FTP was not set properly. You have recorded 6.5 hours or race time. The recorded TSS of 553 indicates you did the equivalent of 5.5 out of the 6.5 hour race at your FTP. That doesn’t seem possible.

Well, for sure I like to tell myself my FTP is set way too low. The problem is when I try to bang it out for an hour it usually turns out to be spot on.


Anyhow, don’t make me the defender of some other folk’s software. It’s the tool that I have. Which is maybe another good reason to lean on Recovery Advisor but I would say the jury is definitely still out on Recovery Advisor. It’s interesting but not convincing for me.

I don’t have my 510 anymore but I was having a problem with my TSS on Garmin connect. The settings defaulted to 200 FTP and would not let me adjust it.

One of the things about the Firstbeat Metrics is that many of them are based on Excess Post-Exercise Oxygen Consumption (EPOC). This is a method that tends to under-emphasize time spent purely in the aerobic zone, and give greater weight to time spent in the sub-threshold and above zones (at least compared with TSS). Unfortunately Garmin doesn’t give any visibility to the raw EPOC numbers, but I did get a good understanding of how they worked with an older Suunto watch.

This may well account for the the different RT metrics from your ride.

VO2Max is also a factor in the EPOC and Recovery Time - a high VO2Max (as estimated by Garmin) will result in a lower EPOC and Recovery Time estimate for the same workout., so if your VO2Max changed between those rides, this may also explain things.

1 Like

After screwing around with Garmin Recovery Advisor (GRA) for quite some time I feel like I have some data & experience to really draw some conclusions about it.

First of all, most know but for those that don’t, GRA is just FirstBeat HRV technology licensed for use in Garmin recreational GPS devices. So if that colors your opinion of the feature…there you go. Or, if you just want to google FirstBeat and read up on them you can.

After fiddling with GRA for quite a while my conclusion is it’s pretty good. For sure it’s better than floating rudderless through a sea of overtraining. I tried a few of the TR canned training plans but let GRA ‘schedule’ my workouts…the first couple of weeks end up getting compressed and then workouts start to get spaced out more than the workout plan originally had them.

But GRA wants to keep me at a TSS that (IMO) doesn’t produce optimal results. On the other hand, it’s a training load I could probably handle week after week, year in year out. If I really want to put an edge on my fitness, though, GRA probably won’t get me there.

Additionally, GRA will prevent overreach/supercompensation that we all try to achieve before a big event. GRA just keeps you on a level path of moderate training stress. Not sure how this would play out if you continued to do it for a long time…let’s say 24 months or something. Maybe the TSS trend would reach an inflection point and start to trend up again? I’m not willing to committ that much time to it…

Anyhow, here’s a quick chart showing GRA tss/fitness/fatigue/whatever trends vs non-GRA. This is for a rider that is working hard to claw their way up to mediocrity. Also shown in this chart: what happens to your fitness when you crash & need surgery.

1 Like

GRA on Garmin watch? Thanks for sharing.

No. Just Edge devices.

1 Like

I think it’s there (to different extents) on Forerunner devices?

Yes GRA is also on some Forerunner devices. My 520 is getting old, would be nice to pick up a 530/830 and give GRA a chance when base starts in a month.

Do it! 530 a noticeable upgrade: brighter, clearer, larger screen.

1 Like

It appears to be there on Fenix 5 plus also. And there’s a widget to see what GRA thinks is current recovery time beyond just having it flash up at the end of the ride. Physio TrueUp keeps things in sync between edge and fenix then.