MS Band stacked up against Polar Loop (LONG)

[dbregman]

I thought I would add some data about the band for folks looking to make a decision. A bit of background. I have been using some sort of fitness/activity tracker on and off for about 5 years. Here is the list:
BodyBugg
Garmin Forerunner (old version)
Basis Band (Watch)
JawBone UP
Polar Loop (with heart rate monitor)

I won't go into all the devices, but i wanted to compare the MS Band to my latest device, the Polar Loop. Polar is known for fitness and has a long history in the business. My goal with a device is to track activities and workouts. I workout 4-5 times a week a combination of cardio and weights, play tennis 3-4 times a week and climb indoors 2-3 times a week. For me the Polar has been a good way to track the activity, I wear it 24/7 and put on the heart rate monitor when I workout, play tennis or climb.

The issues with the Polar (thus the purchase of the MS Band), are no notifications (I wear a Martian Watch with it for that), and having to put on the heart rate strap for workouts. Other than that I find it to be accurate and it gives a good amount of data via both a phone and web interface.

I am hoping that the MS band will be able to replace the Loop and am wearing both to compare and contrast. So far my results have been mixed. - Equal or better in some areas and not so in others. I will endeavor to update this thread with my findings. I hope you find it useful.

First comparison - 5K run with my 8 year son. NOTE - I am not a big runner and don't think I will use the running feature that often, It just so happens that I was scheduled for a 5k two days after I got the band.

Here is the heart rate from both - meshed as best as I could. The Polar is the red. This is trimmed (see below) I am not sure of the why it has the erratic start.



I would rate this as very close.

Here are the stats:



For the Polar it starts tracking when the Loop picks up the strap - this can be some what annoying since I have to put it on before the activity and it does not always sync right away. I can trim it on the web app to just see the workout, but it will not tell me the calorie count for the trimmed time (just max/min and average HR). Above you can see the Polar shows more calories, but also more time -when the band picked up the strap. Overall I would say it is close since the time before the race was basically at rest. I do like how the MS Band shows the HR recovery below the ending HR as this is a key measurement of fitness (not sure why it didn't show the 2 min recovery unless I stopped the tracking too soon).

Here is an average workout for me:



Here the Loop took a while to get synced with the chest strap and thus the Polar workout is about 5 minutes less. Both show fat and fit calories, although the Polar shows more 'fit' calories then the MS Band does (36% vs 50%). The Polar strap lost connection at the end (the huge drops in HR) when I was doing weighed leg lifts. This often happens with the chest strap when doing ab exercises or push-ups. because of this and the 5 minute discrepancy I would rate them overall pretty close - taking to account that the polar 'lost' the first 5 minutes and the last 10 minutes of the workout.

For those that care - this workout was a mix of weights and cardio (step ups, lunges, jump rope, etc).

Where the comparison starts to get off is tennis.



The MS Band seems not so show the recovery as well as the Polar and has a higher overall HR during the match. I wore the MS Band on my left hand with the face down (I am right handed with a two handed backhand). I am not sure what impact this position has, but I will experiment more with it and post the results.



Due to the overall higher reading of HR during the match the MS Band posted MUCH high calorie readings than the Polar. This is of concern to me. I am going to try playing with out the MS Band in workout mode and see if it picks up the movement as a higher (correct) reading. This would work for me, although I do like being able to see the workout broken out as a separate metric.

Here are my initial findings. I will post an overall end of day metric soon with a rest day without a big workout, as well as how the MS Band preforms with climbing and other tennis readings. Let me know if you find this useful.

 

[realwarder]

I'm curious... are calories burnt only directly related to heart rate?

Some interesting data though... thanks.

 

[B1zzle]

do both the polar and the band have equal information on it's user's height, weight etc?

 

[dbregman]

Yes - they both ask for Sex, age, height and weight. The Polar allows you to estimate your level of fitness and VO2Max with a resting heart rate test.

realwarder - As I understand it - calories burned are a function of how much oxygen the body uses to preform an exercise. Heart rate is the best way to measure that with out a mask that measures oxygen use. That said there are different formula to calculate calories based upon heart rate. Using one on the web I get a different number then both the polar and MS Band based upon the reported average and duration.

 

[Bkr11]

Nice - I was using mine in the Dino Dash as well...though we essentially walked it...I guess 2k is long for Kindergarten ;-)

 

[ytrewq]

My wife has a Polar strap. It is very inconsistent for her -- often, it does not find a heart rate at all (for lengthy periods of time). She's very happy with the Band since we picked it up last Thursday because it consistently finds a heart rate. Perhaps the Polar is more accurate during those windows of time that it finds a heart rate on her, but that's not helpful for the 25-30% of the time that the Polar doesn't register a heart rate at all.

 

[greyskytheory]

this is another comparison I did with my Suunto Ambit 1 with HR on the left and the Microsoft Band on the right. This is the third comparison I've done between the two and the Band is not doing too well for me.

 

[dbregman]

Interesting - what type of activity and how are you wearing you band. What is really interesting is that the Avg HR and the Calories are almost spot on. I guess the question is - what matters more to us the user in the long run. For me I would be happy with spot on Calorie and avg reading and not have the graph represent correctly.

 

[ytrewq]

greyskytheory, I may be confused, but the Ambit reading seems to suggest that, within any five minute window, your heart rate oscillated by more than 20 bpm several times. Your heart rate shouldn't vary that much several times every 5 seconds. Am I misreading? [Whoops --- I did misread. It's minutes, not seconds. In that case, 20bpm oscillations a few times every five minutes is understandable.]

I think a flawed premise in all of these tests is that they all assume that, if there's a difference between the Band and any other device, then the other device must be reading 100% correctly, and any discrepancy is due soley to inaccurate readings by the Band. That is a flawed premise. All of these devices are consumer devices that are designed to get it reasonably close, not hospital-grade devices meant to provide a high degree of precision. The are all off. It would be a statistical miracle for them to give the same results. They should both be in the same general ballpark, but that's about it. Greyskytheory has two devices that give an average heart rate within +/- 1.5% of 145 bpm. That sounds like two devices that are getting pretty similar overall results.

 

[greyskytheory]

greyskytheory, I may be confused, but the Ambit reading seems to suggest that, within any five minute window, your heart rate oscillated by more than 20 bpm several times. Your heart rate shouldn't vary that much several times every 5 seconds. Am I misreading? [Whoops --- I did misread. It's minutes, not seconds. In that case, 20bpm oscillations a few times every five minutes is understandable.]

I think a flawed premise in all of these tests is that they all assume that, if there's a difference between the Band and any other device, then the other device must be reading 100% correctly, and any discrepancy is due soley to inaccurate readings by the Band. That is a flawed premise. All of these devices are consumer devices that are designed to get it reasonably close, not hospital-grade devices meant to provide a high degree of precision. The are all off. It would be a statistical miracle for them to give the same results. They should both be in the same general ballpark, but that's about it. Greyskytheory has two devices that give an average heart rate within +/- 1.5% of 145 bpm. That sounds like two devices that are getting pretty similar overall results.

You're correct that the average hr is close but the max hr is nowhere near each other at all. If doing interval training of some sort and seeing on the Band a max hr of 151 and a max hr on the Suunto of over 180 it's clear to see the Band is not keeping up. This is the theirs workout I've done with the band and the third set of comparisons I've posted and the constant has been that the band doesn't seem to track my hr well. It seems to have problems tracking quick spikes which happens when you do interval training or HIT type of workouts.

 

[dbregman]

More comparison metrics. This was a rest day - although I did play an hour of light tennis - not tracked separately with either band. Here are the screen shots.



You can see that the Polar calculated 8140 Steps with 2583 calories burned and the MS Band calculated 6856 steps and 2153 calories burned. Since my Basil Metabolic Rate (what I burn just being alive) is 1733 I think the band calculated low. I think there might need to be some firmware tweaking done to correct this. If there is anyone from MS reading this forum - I would love to see their take on the data.

 

[valadon]

I wore my fitbit one and my Band all day all yesterday, it was a mostly a rest day with a quick strength session in the afternoon. The workout was tracked with exercise mode on the Band, and was NOT tracked in any special way on the fitbit, was just wearing it as normal.





The fitbit shows that I have burned quite a few more calories, enough for an extra hamburger Which one is correct? I don't know. I hope it would be the Band since it is taking more measurements into account, but it does seem kind of low.

If it helps, I am a 32 year old man, 147 pounds, 5'8"

 

[Raynne413]

The fitbit shows that I have burned quite a few more calories, enough for an extra hamburger Which one is correct? I don't know. I hope it would be the Band since it is taking more measurements into account, but it does seem kind of low.

If it helps, I am a 32 year old man, 147 pounds, 5'8"

That's ok, including one tracked 10 minute workout that supposedly burned 74 calories, and one non-tracked walk for 10 minutes, my band says I've burned 242 calories for the DAY! LOL

 

[valadon]

That's ok, including one tracked 10 minute workout that supposedly burned 74 calories, and one non-tracked walk for 10 minutes, my band says I've burned 242 calories for the DAY! LOL

What does it say on the Band for calories? I am wondering if there is a bug in the phone app that isn't showing your calories. So check the band vs the app, I am wondering the WP is still having daylight time issues. I am on an iPhone and didn't get hit with the Daylight bug so I am wondering if it is a WP issue.

 

[Raynne413]

What does it say on the Band for calories? I am wondering if there is a bug in the phone app that isn't showing your calories. So check the band vs the app, I am wondering the WP is still having daylight time issues. I am on an iPhone and didn't get hit with the Daylight bug so I am wondering if it is a WP issue.

The band says the same as the app. Sadly, though, I'll be returning the band. It is just too big for my wrist and I can't seem to find a positioning that doesn't mess with the circulation in my arm or hand.

 

[TheZuneLune]

I did a spot check with my Polar last night and found it very close. I screwed up doing a real check.

 

[BobLobIaw]

http://tapatalk.imageshack.com/v2/14/11/03/4b1d9d8a69d38139989662f8b4243da0.jpg this is another comparison I did with my Suunto Ambit 1 with HR on the left and the Microsoft Band on the right. This is the third comparison I've done between the two and the Band is not doing too well for me.

This is probably the most disturbing data I've seen so far. If someone is getting his or her HR up to 180+ and the Band only registers a 151 max, then the Band HR function is basically useless for structured interval workouts. Average HR is only so helpful--a reasonably serious athlete needs to know how long he or she is anaerobic or at one's AT as well. It looks like MS may need to tweak the settings; otherwise this device may be of limited usefulness with respect to the HR function.

 

[Joe920]

This is probably the most disturbing data I've seen so far. If someone is getting his or her HR up to 180+ and the Band only registers a 151 max, then the Band HR function is basically useless for structured interval workouts.

Definitely. My main question at this point is whether the raw HR data simply needs this kind of aggressive averaging, or whether there is room to turn it down a bit. I'm hoping the latter.

 

[greyskytheory]

You can see the full comparison at the link below. You'll see that it does pretty well for me while tracking steady state activities but not well during interval training. At one point during my sprints it actually stopped working altogether. I'm hoping my Band is just defective and Microsoft will allow me to exchange it. Hopefully the Band is defective, if not, I hope they can put out some sort of update to increase the accuracy.

 

[greyskytheory]

http://forums.windowscentral.com/microsoft-band/321078-band-vs-ambit-round-two.html