Suffer Score, Relative Effort, TSS...
radhartgeorge
Posts: 37
Strava has just released their 'Relative Effort' functionality - it would appear to be a TSS tracker? I think this is a good addition for Strava Premium members but the lack of cross compatibility between the Suffer Score, Relative Effort Score, TSS on various platforms irks me slightly. Has anyone done a comparison of the various training stress metrics and summarised the findings?
0
Comments
-
I did a high intensity 1 hour yesterday and had this new thing pop up on my Strava mobile feed as a "stronger effort than usual" but when I looked online it showed a "massive" relative effort and a big number in red.
The number it showed was identical to my average heart rate - so I assume that my assessment was linked closely on this.0 -
Strava says that the science behind it is based on the work of Dr Marco Altini. He's the guy behind a lot of the Heart Rate Variablity analsysis in HRV4Training
According to their blog entry it seems to be only Heart-Rate based. The blog also says they'll be getting rid of the power-meter based Suffer Score. From their blog:
"With the introduction of Relative Effort, it is also time to bid farewell to Suffer Score."
https://blog.strava.com/relative-effort/
https://medium.com/strava-engineering/q ... a0e3dd6a520 -
Based on one ride so far, it does not look like Relative Effort is anywhere close to hrTSS:
hrTSS: 71
Strava Relative Effort: 36
TSS (power-based): 101
Strava Training Load: 76
Digressing a bit, it's worth noting Strava's Training Load (their equivalent to TSS), is much lower for high intensity rides than the traditional TSS calculation. This is because their Weighted Average Power algorithm does not weight high intensities as much as the Normalized Power algorithm does. For the ride above, my Normalized Power was 280W, while Strava's Weighted Avg Pwr was only 233W. This was a hill interval workout. For steady state rides, the two are very close.0 -
Yes, Suffer Score does not exist any more an Relative Effort is not the same thing so the numbers will be different and will also be different to hrTSS (Suffer Score was just hrTSS, but that obviously differs quite a bit from power based TSS. I found that hrTSS from Strava and power TSS from TrainerRoad was typically 50% different, power TSS was usually around 50% higher).
This week so far I have done 1 ride on Monday and 3 runs (I know, I know - was away for work, no bike...), and I have to say for someone who isn't training "seriously" at the moment the relative effort system seems very intuitive. Being able to look at fatigue from cycling and running together is quite neat, and possibly quite useful for triathlete types.
Just as before though if you are training seriously with power you will probably want to use a proper tool (like training peaks).0 -
Hi Folks,
My quest for answers led me to download and install StravistiX which is a Google Chrome extension and has essentially supercharged my strava experience into something much closer to my (paid for) TrainingPeaks account does. I'd highly recommend it to those of you that like data analysis and cant ustify paying for for a TrainingPeaks account.
Link is here: https://chrome.google.com/webstore/deta ... pckn?hl=en
Unfortunately this also adds another stress measure into the mix but is probably the one I'm going to rely on going forward.
Rad0 -
Strava doesn't know my actual 20min or hour power, so their weighted avg has always been worthless anyway. It's often up to 10% off from my Bolt's calculation with the app. And the Bolt knows the power zones I input to it.
I think I even "broke" the Strava calculation on one ride. The Strava avg power was about 8w higher than the Strava "weighted avg". How do ya figure that?
Personally I am not a big fan of HR. So pretty much solidifies I'll never buy Strava premium if they nix power.
To me HR in cycling training is equivalent to what it would be like to try to dyno tune a race car only using oil temperature or EGT or something. You're measuring the symptoms and not the cause. And a symptom that in some training efforts lags woefully behind in a timescale frame of reference.0 -
burnthesheep wrote:Strava doesn't know my actual 20min or hour power, so their weighted avg has always been worthless anyway. It's often up to 10% off from my Bolt's calculation with the app. And the Bolt knows the power zones I input to it.
I think I even "broke" the Strava calculation on one ride. The Strava avg power was about 8w higher than the Strava "weighted avg". How do ya figure that?
Personally I am not a big fan of HR. So pretty much solidifies I'll never buy Strava premium if they nix power.
To me HR in cycling training is equivalent to what it would be like to try to dyno tune a race car only using oil temperature or EGT or something. You're measuring the symptoms and not the cause. And a symptom that in some training efforts lags woefully behind in a timescale frame of reference.
Couple of points really
- Strava does do power - if you set the fitness and freshness graph to use power only it shows training load and does the conventional calculation. Each ride where power is recorded using a power meter gives a training load figure as it always has.
- Strava does know FTP, you set it in settings, or you certainly do in Premium. Rides are still given a training load figure which is calculated the conventional way (although, Strava used to treat FTP as a "global" value and updating FTP updated all historic rides, which is wrong, but they might have changed that)
Apart from that, HR is absolutely fine for tracking fatigue and load in aggregate, which is the whole point of this relative effort feature as it is designed to include all activities so can be used by e.g., triathletes to track across sports. Not so good for setting training zones for the reasons you've alluded to, plus others. But a fitness and freshness function which works across sports has been on Strava's to do list for some time, remember that not all Strava users are solely cyclists!
If you were serious about training and planning in cycling then you would still go for a proper tool such as TrainingPeaks or Golden Cheetah.0
