Strava's Fitness metric is deeply flawed

Honestly, no one should be measuring their fitness based off of what Strava says, nor would I recommend using Garmin's VO2 max. Strava is just a social media for runners and a subpar running log, it's fun to use but shouldn't be much more than that.

It massively over-values long runs vs intervals and races.

Run a PR over 10k? That will be 2 points less fit tomorrow.

Run 20x400 at 5k pace? Looks like you're getting kinda soft over there.

Run 20 miles with a heart rate of 130bpm? Woah! You're kinda a stud.

In terms of trends it's not entirely without its uses, but it's pretty dumb

AndySmith wrote:

It massively over-values long runs vs intervals and races.

Run a PR over 10k? That will be 2 points less fit tomorrow.

Run 20x400 at 5k pace? Looks like you're getting kinda soft over there.

Run 20 miles with a heart rate of 130bpm? Woah! You're kinda a stud.

In terms of trends it's not entirely without its uses, but it's pretty dumb

Exactly this for both Garmin and Strava...

Also, I am currently 46 fit on Strava. 46 what?

Raphael Warnock Knock wrote:

I notice that as my fitness gets better, the fitness gauge goes DOWN on strava.

Seems, as you improve and are able to run with a lower heart rate, strava takes down your fitness.

Garmin, using their estimated VO2 Max does the opposite.

So after completing a recent PR 1/2 marathon, my estimated VO2 Max went up with Garmin, yet Strava took down my fitness level.

I think they should tweak their algorithm, so it does a check of comparable runs, and if you run similar times with lower avg HR, your fitness should go UP not down.

You are on to something.

Keep in mind that all these gadgets like strava, fitbit..... are just to entertain people and make them feel good.

All people want to see is how many calories they "burned" so they can eat more junk food.

In the last year, I increased my mileage by 50%, did harder workouts, decreased my marathon time by 34 minutes and my 5K by 90 seconds and strava says my 'fitness' went down from 103 to 85. Lol.

So, yeah. Strava's fitness gauge seems way off, for me at least.

My garmin also sends data to Runalyze and their fitness gauge seems a lot closer to what I'm felling and how I'm performing, at any point in time. It's worth checking out if you're a running-data geek like me.

GettingFasterDude wrote:

In the last year, I increased my mileage by 50%, did harder workouts, decreased my marathon time by 34 minutes and my 5K by 90 seconds and strava says my 'fitness' went down from 103 to 85. Lol.

So, yeah. Strava's fitness gauge seems way off, for me at least.

My garmin also sends data to Runalyze and their fitness gauge seems a lot closer to what I'm felling and how I'm performing, at any point in time. It's worth checking out if you're a running-data geek like me.

Thanks I'll look into that.

@Strava - fix this!

The Strava fitness score literally has an inverse correlation to my actual fitness. The only thing it accurately identifies is when I miss a bit block of training.

3 months after starting to run in about 1:33 HM shape = 112

Immediately after running. 15:45 5k 1.25 years later = 39

GettingFasterDude wrote:

In the last year, I increased my mileage by 50%, did harder workouts, decreased my marathon time by 34 minutes and my 5K by 90 seconds and strava says my 'fitness' went down from 103 to 85. Lol.

So, yeah. Strava's fitness gauge seems way off, for me at least.

My garmin also sends data to Runalyze and their fitness gauge seems a lot closer to what I'm felling and how I'm performing, at any point in time. It's worth checking out if you're a running-data geek like me.

I signed up for Runalyze my monotony score is 2.37. Should I worry, says anything over 2 is critical. I guess I need to vary my training

I like generally like strava and try to support the company but it’s things like this that make me think they are doomed to fail.

It’s an awful algorithm and while I would never get anything out of it anyways, it should at least make some sense. My “fitness peak” workout 1.5 years ago is a LR 30bpm higher and 30s/mi slower than one I did a week ago (which dropped my fitness score). I’m now at my lowest strava fitness score ever while being the most fit I’ve ever been. In the end, who cares but strava should be embarrassed by this joke of a metric.

Raphael Warnock Knock wrote:

I notice that as my fitness gets better, the fitness gauge goes DOWN on strava.

Seems, as you improve and are able to run with a lower heart rate, strava takes down your fitness.

Garmin, using their estimated VO2 Max does the opposite.

So after completing a recent PR 1/2 marathon, my estimated VO2 Max went up with Garmin, yet Strava took down my fitness level.

I think they should tweak their algorithm, so it does a check of comparable runs, and if you run similar times with lower avg HR, your fitness should go UP not down.

Based on the last 21 days, you have to run a longer amount of time than average with the same average heart rate, or run the same length of time than average with a higher average heart rate, or run a longer amount of time with a higher than average hear rate in order to gain fitness.

If you run 10 miles every day and can now do it with a lower HR and 15 minutes faster than you could 2 months ago, you are not doing anything to improve beyond your current shape. You may be faster but you are not gaining anything because you are not working as hard as you used to.

Agree 100%. This fitness thing is bad. But their segment matching algorithm is laughable. I've sent a bunch of support requests, but doesn't seem like they really care. Whoever said that they build these "features" as entertainment is exactly right.

Walmslord wrote:

I like generally like strava and try to support the company but it’s things like this that make me think they are doomed to fail.

Raphael Warnock Knock wrote:

GettingFasterDude wrote:

In the last year, I increased my mileage by 50%, did harder workouts, decreased my marathon time by 34 minutes and my 5K by 90 seconds and strava says my 'fitness' went down from 103 to 85. Lol.

So, yeah. Strava's fitness gauge seems way off, for me at least.

My garmin also sends data to Runalyze and their fitness gauge seems a lot closer to what I'm felling and how I'm performing, at any point in time. It's worth checking out if you're a running-data geek like me.

I signed up for Runalyze my monotony score is 2.37. Should I worry, says anything over 2 is critical. I guess I need to vary my training

I doubt you should worry. But it could mean varying things (distance and/or intensity) a bit might help (maybe).

Interestingly, when I did my marathon training this winter (Pfitzinger 55-70 mpw, runs from 5-22 miles) runalyze liked my lack of monotony and it was always green. Now that I've shifted gears and I'm doing 5K training (Pfitzinger's 60 mpw 5K, most runs 8-12 miles) it's constantly red. I'm still varying things quite a bit, with two hard workouts per week and slow days slow. Plus, the overall mileage is about the same, in the 60's mpw. It think it likes it when you vary the length of your runs a lot, and that this particular algorithm is geared more towards marathoning, than shorter distances. I haven't found the monotony thing as helpful as the overall fitness, fatigue and marathon shape readouts.

When in doubt, go by how your body feels.

Strava uses the training impact factor for each workout to calculate a fatigue factor. The impact factor is basically a weighted sum of the heart strokes. I do not know exactly but I guess it uses a factor of the max HR to estimate the impact at any moment during the workout and sums this. So on a long run you can gather a lot of heart strokes compared to a hard interval workout where the average heart rate is not verymuch higher than the longer easy runs, but the workouts might be shorter.

The more training impact the higher fatigue factor. Fitness is a factor that moves up in case the fatigue is higher than the current fitness and down in case the fatigue is lower than the current fitness. That means that in the start when the fitness was low and the fatigue estimate was high above it, the fitness moved fast up. Then when the fitness came up to the levels your fatigue factor on average was, it flattened out. Strava fitness does not show your fitness in the sense of your race ability, they are just numbers based on HR.

VO2max estimates of Garmin uses a simple relationship between pace and heart rate vs your max HR. So if the HR is getting lower on the same pace or the pace increases on the same HR the VO2max estimate will increase.

If you train by HR it is natural to increase the pace without thinking about it to keep the HR in the zone. If you train by pace, the pace needs to be adjusted along the improvement of the race pace or else your easy runs will get easier and easier and Strava is right that your training impact factor goes down on those workouts and then they count less and makes the fatigue lower (which is right) and when the fatigue is lower than the fitness level, the fitness level will decrease.

I would check the paces of all your easy running to possibly adjust the easy pace to reflect your racing paces (like 5k/10k/HM paces)

Strava fitness algorithm is no magic or complex thing, it is using quite straight forward mathematics based on your HR through the workouts. This is well tested so I don't see anything needing to be fixed. If fixing, see to that your max HR is input correctly and that you use a HR monitor that is accurate. Or else all will be crap

Raphael Warnock Knock wrote:

ojdbdhyd wrote:

Sorry, in the last 21 days, I ran a 9 minute PR in the 1/2 marathon, and strava thinks my fitness is worse.

Based on the last 21 days, you have to run a longer amount of time than average with the same average heart rate, or run the same length of time than average with a higher average heart rate, or run a longer amount of time with a higher than average hear rate in order to gain fitness.

If you run 10 miles every day and can now do it with a lower HR and 15 minutes faster than you could 2 months ago, you are not doing anything to improve beyond your current shape. You may be faster but you are not gaining anything because you are not working as hard as you used to.

Were you more fit or were you more rested? Sounds to me like you set a pr because you decreased your training load.

Garmin told me my activity level is dropping and my fitness is decreasing. Oddly I am training more now than I have in probably 10 years! I quit looking at all the metrics cause they dont make sense.

Couldn't care less about Strava estimations. Runalyze is spot on on predictors for me.

silviuburceadev wrote:

Couldn't care less about Strava estimations. Runalyze is spot on on predictors for me.

Funny you mention it, but Runalyze also uses training impact (TRIMP) based on HR and they use the same line of thinking around fatigue and fitness. Maybe presented differently and with slightly different formulas, but basically the same idea.

It seems to be primarily based on increasing intensity. So if you have plateaued or are in a recovery phase it will go down. It's basically a measure of "improvement" rather than an absolute fitness score. So if you aren't improving (based on the limited information they have) by increasing mileage, pace, or both, then it won't go up.

If I am building my base or in a marathon training phase, it tends to rise really quickly. When I'm training for a 5K or something where you don't need to run as many miles or ramp up intensity as much, it stays flat or goes down.