This discussion has been locked.

You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Garmin killed the runner in Forerunner

I've upgraded from an FR630 to the 645M about 4 weeks ago and I'm super disappointed. In my eyes some cool new features have been added, but the main reason for the existence of this watch - to measure activities and help runners in training - is almost totally useless.

As I said in another thread, a measuring device has to be (A) consistent (B) accurate. I will not talk about accuracy, there's another thread for that. The 630 fulfilled (A), yet the 645 doesn't (all comparisons are made in runs on the exact same paths and deal with 'lap pace' - since current pace is something that has never been trustworthy):

My 630 would lap (1km) every time at the same place, or 2-4m before or after when I ran my usual easy runs. The 645 is nowhere near that.
On an out and back run, my 630 would lap on the return almost exactly where it did on the out run.
On my 630 my reported lap pace would be more or less in line with what I perceive as my pace after about 70-130m inside the lap. The 645 shows me ludicrous paces starting a new lap, and settles to something plausible after 500m at the earliest, and sometimes not at all. Not at all means that the lap pace is reported as something too high or too low, and then lap pace for the beginning of the next lap is even more ludicrous (for example lap 1 starts at 5:20, and slowly slowly goes down towards 4:43 at the end of the lap, even though I'm going at around 4:25. The next lap, lap pace starts at 3:10, and slowly slowly rises)
On my 630, taking a turn under some tree cover did throw it a bit off (for example after 600m in a lap pace would be 3:41 and after the turn would quickly jump to 3:44), the 645 just can't handle these turns, the pace can jump to 3:56 and doesn't recover by the end of the lap, though I know I'm running a constanct 3:40-45.
The 630 can't handle U-turns, especially under some tree cover, the 630 did so very well. Taking a turn after 500m in a lap, lap pace can immediately jump from 4:33 to 4:59, and not recover by the end of the lap.

Example: Thursday I ran a 6km tempo at about 3:45.

The first 4km had some turns in them, their reported paces were 3:52-3:54. The pace was "ruined" when taking the turns. Also, I've noticed how sometimes the lap pace could jump up/down considerably even when the lap is almost over (meaning the lap pace should be a more solid average and hard to change). The last 2km were on a stretch, and those were 3:46 and 3:42 - in line with expectations.
On my cool down I reached my run end point, but wanted another 2.4km to make it 15. I ran 1.2km north, turned around, and on the return the watch measured 1.08. Come on!
At home I used a mapping service, the total run measured 15.4km on a map, where the watch reported 15km even.

I can't trust this watch. It's GPS/algorithm is far inferior to those of the 630. It is inconsistent with it's measurements, and therefore I see no reason to even discuss accuracy.

I wanted to upload an image to show an example trace of this run that could partially explain why the measurement is so off, but the forum won't accept my files. I'll try again later.

Top Replies

kalamarios over 6 years ago +1

Kudos for the best title I've seen in this forum (and so true BTW).

It sounds like an extreme statement but I honestly think that Garmin doesn't care about runners that much anymore.

All Replies

0 scotthunter2 over 6 years ago

A Garmin representative contacted me, and said that while they have been receiving complaints about accuracy and lap pace - they have not seen them to the extent that my reports show, and therefore they will replace my unit. I will report once the new unit arrived, hoping to stand corrected and that a non-defective 645M is not inferior to the 630 in these basic functionalities.

I don't think that a new device will fix the lap pace issues unfortunately. I replaced my 645 with a new one and I still don't believe what the lap pace is telling me sometimes. It makes it really hard to pace yourself in a race. I did a 10km race last night and although my average pace was around 6:45 min/mile, at some points the watch was telling me that I was doing 7:30 - 8 mins per mile on my lap average, even though I knew by feel that I was keeping a steady pace as it was a relatively flat course. I can understand variability in live pace, but lap pace should not be so erratic. Admittedly the course was though a forest, but I had GPS + Galileo enabled.

I've only kept it as I think it's more comfortable for wearing 24/7 than other watches in Garmin's current line up and doesn't scream sports watch, even though sometimes I feel it's just a very expensive stopwatch in as far as helping you pace a race.

0 PatrickNSF over 6 years ago

I replaced my 645 with a new one and I still don't believe what the lap pace is telling me sometimes.

Having used a 645, 935 and Fenix, the only way I’ve been able to obtain a reliable lap pace is to use a foot pod, in my case a Stryd. Even with a Stryd, I generally don’t rely on whatever the watch tells until after the first 1/4 mile of each mile.

0 scotthunter2 over 6 years ago

Having used a 645, 935 and Fenix, the only way I’ve been able to obtain a reliable lap pace is to use a foot pod, in my case a Stryd. Even with a Stryd, I generally don’t rely on whatever the watch tells until after the first 1/4 mile of each mile.

On the 630 & 935 I previously owned, lap pace with GPS was usable. On the 645, I tend to ignore it. I could use my footpod for pace only, but mine is calibrated for my 5km pace so I don’t know how reliable it would be for long runs and recovery runs.

0 Former Member over 6 years ago

I could use my footpod for pace only, but mine is calibrated for my 5km pace so I don’t know how reliable it would be for long runs and recovery runs.

I don't use a footpod. Never have, probably never will. So, this may be a dumb question. What does calibrated for your 5km pace mean? Is a footpod calibrated by stride/cadence and only reliable if you maintain the same stride/cadence every time afterwards? It was my assumption that once calibrated, the point of the footpod was to let you know if your calibrated pace was faster or slower than normal. If that's the case, I'm confused how running longer or doing recovery runs would be affected. Is my assumption on how a footpod works incorrect?

0 talsela over 6 years ago

I don't think that a new device will fix the lap pace issues unfortunately.

I'm also afraid of that but really hoping to somehow be wrong and find that the replacement 645 operates as or better than my late 630. If this is not the case, how do you explain that it's more or less you, I, and maybe a couple of other people that have noticed the, putting it mildly, not-so-subtle issues with lap pace? Is it that nobody looks at the watch during a run or pays much attention to lap pace?

Having used a 645, 935 and Fenix, the only way I’ve been able to obtain a reliable lap pace is to use a foot pod, in my case a Stryd. Even with a Stryd, I generally don’t rely on whatever the watch tells until after the first 1/4 mile of each mile.

I find that very sad. With my 630 I felt like lap pace was pretty accurate after about 150-200m. I thought with Stryd it should be even better. With the 645M (which should be replaced this week in order to check if it's defective) I can't trust lap pace even after 950m. In fact, sometimes (run) average pace changes in a way that indicates it's not accurate even 4-5km in a run.

0 scotthunter2 over 6 years ago

God help me next time I do a half marathon and going for sub-90. I need to be bang on 6:52 min per mile. I don't know how people can use this watch for serious racing or when they are going for PB's and need to pace their splits perfectly to achieve a goal time.

0 tmk2 over 6 years ago

Is a footpod calibrated by stride/cadence and only reliable if you maintain the same stride/cadence every time afterwards?

I think the general idea is that the simple garmin footpod (stryd is another beast altogether) or any of its clones, becomes a little less accurate if you deviate significantly from the calibration pace. I use a foodpod regularly and I noticed that it maintains satisfactory accuracy within the paces I usually run (3:30 min/km - 6:00 min/km). "Satisfactory" being of course relative.

0 WillNorthYork over 6 years ago

TMK17 A footpod does more than just measure cadence -- it's not the case that it simply assumes a fixed stride length and uses the measured cadence to determine your speed -- which is what I used to think and which seems to a common assumption. Apparently the Garmin footpods use a 3-axis accelerometer whereas Stryd has a "9-axis sensor". The idea is they're both measuring movement in 3d space tho, so it's not like the footpod will always think you're running the same pace whenever you run the same cadence as last time.

http://fellrnr.com/wiki/Footpod
"The first generation Footpods use a 3-axis accelerometer, where second generation Footpods typically use 9-axis sensors which combine accelerometers, gyroscopes, and magnetometers for far greater accuracy. This page focuses on the first generation Footpods, though Stryd is far superior."

But like tmk2 said, Garmin footpods only seem to be good within a narrow pace.

0 Former Member over 6 years ago

Heres my thoughts

The lap pace isn't great, I don't use it that much but at the start of the lap its all over the place, its usually showing much slower than I think its going. I've noticed when doing some 20 secs strides the pace on the watch at the end of the strides is significantly slower than the pace that ends up on Garmin connect for the short section. Its as though there is a big lag.

I also find that it seems to struggle with a GPS signal when running under tree cover but it seems to correct itself. Quite often ill be running a steady pace and then the average pace for the whole run will increase by 2-3 secs a mile (even after running 5-6 mile) then ill look again 1 min later and its back to where it was.

Funnily enough I've done 5 races with it and the distance as been almost spot on.

These are the distances the watch measured

marathon 26.2
10mile-9.99
10k-6.23
10k-6.21
5k-3.12

So overall I think it gets the distance correct over the whole run but seems to have problems with the pace at the start of each lap. I should add all my races were road races and id suspect the accuracy wouldn't be as good on a trail race with tree covering.

0 robm over 6 years ago

Although I am one of the few with a "working FR645M", I also can confirm that the 'current pace' readout varies too much to be useful. I've only really noticed while watching it during speed walking, but it can vary up to a minute as I traverse 1 km. The pace for each km seems correct at the end, but part way through it may report up to a minute slower although it does not seem like I have let off any effort. I have been attributing this to varying GPS signal over my walk; I'm about 6 degrees south of the equator and have not tried with GLONASS or Galileo.