In the past four and a half years, I've gone through a lot of running watches: Garmin 220, 235, vivoactive 3; Suunto Spartan Trainer, and now a Coros Apex. The Suunto has some definite strong points and some definite weak points (especially the software) so I figured it was time to try something new. I'd had my eye on the Coros Pace for a while. When I went to check it out I saw that the newer flashier Apex was available for the exact same price, so I jumped on it. There still aren't many reviews out there, even fewer with the final hardware and software as regular people will see them, so here's mine.

First, as an every-day watch, it's very nice. The Spartan had the best battery life and worst screen I've seen so far. The Apex beats it handily in both categories. The screen is quite readable both indoors and out, with an automatic backlight that actually works. On battery life, it has gone only from 87% to 74% in the past day which includes a four-mile run. That's only 13% per day; even the Spartan would probably lose twice that much. Of course, those numbers could be inaccurate, but if they hold up it's impressive. Also, this is the 42mm model. It's 46mm sibling has even better battery life, at the cost of being a bit chunkier - less so than any of the others except the vivoactive, and way less than the aptly named Spartan.

But every-day wear isn't why I bought this. It's probably not why you're reading this. How does it do as a running aid? Let's break it down.

• GPS Acquisition
  This is one of the things I love about the Spartan. Immediate lock-on, every time. Such a simple thing, but it was a problem with all of the Garmins. The vivoactive in particular often took forever to acquire, and sometimes failed entirely. The Apex didn't acquire as quickly as the Spartan, but it wasn't problematically slow either. So far, so good.
• GPS Accuracy
  Another strong point for the Spartan, and frankly the main reason (along with acquisition) why I put up with its other shortcomings. The Garmins were all mostly reliable, but sometimes showed me 300 feet into a lake or 100 feet underground. Unfortunately, on its first run the Apex was more in the latter category. I traversed two of my MapMyRun courses today, and only got credit for one because the GPS track showed me totally off course for the other.
  If this were a mature product (e.g. like the Spartan was even when I bought it) I'd totally rip into Coros for this. But I've seen these kinds of "teething problems" with every new watch, and the Apex is very new. It's probably even a new chipset. Still, this has to get better. I see there's another firmware update already since my run; maybe that will help. I'll also experiment with using GPS only instead of GPS+GLONASS, since there are reports that the combination is actually worse than GPS alone on many chipsets. The hardware also supports BDS, but the firmware doesn't yet, so that's another possible area for future experimentation.
• Pace Measurement
  If GPS accuracy is off, pace will be off too, and indeed that seems to be the case. There's also something else going on. The pace as shown by the watch itself was all over the map from seven to eleven minutes per mile. I can only hit seven on a good day, and I don't hit eleven even on a bad one, so something's definitely wrong there. The numbers shown in the app or in MapMyRun were much more reasonable, though. So was the overall average in all cases. It clocked me at 8:37/mile overall, and that's pretty much what today's run felt like.
  There are many possible explanations for this disparity. One is that the watch just didn't have enough data and/or enough CPU horsepower to do any better, but more powerful systems can do a better job of processing the raw results. Another is that the Spartan has been misleading me, either because "FusedSpeed" doesn't work as well as advertised or because of other smoothing/averaging that I could tell it was doing. It's possible that what the Apex is showing me is more instaneous numbers, which actually should vary more than I'm used to.
• Other Measurements
  As far as I can tell, elevation readings were pretty accurate. The Apex does have its own barometer, and compass too, so that's kind of expected. I didn't get a good look at heart rate, since I wear the watch on the outside of my sleeve when it's freezing out. Readings at rest have seemed reasonable. I might be able to get some data on this next week in Seattle.
• Cold Resistance
  Since I run all through the winter, this is important to me. Again the Suunto is king. The Garmins all struggled a bit, and the vivoactive was a complete failure in that environment. Today wasn't cold enough for a real test, but the Apex did OK. I'm sure I'll have better info later in the season. My other concern is that the Apple Watch-like dial/button might not be ideal for gloved hands. It didn't seem particularly troublesome today, but if I have to take off my gloves when it's 10F or less to deal with it I won't be happy. I'd honestly prefer old-fashioned side buttons.
• App Integration
  Definitely a weakness at this point. There is basically no app integration right now, though Strava at least is promised. However, you can mail a TCX/GPX/FIT file to yourself from the app. Then it's the old download, unzip, upload tango. Not ideal. It did work, though. With the GPX file I can even see metrics and get credit for courses and challenges (except for the GPS-accuracy issue mentioned above). That doesn't seem to work with the FIT file. This is another area that will have to improve, but I can put it up for a while.

So there you have it. Overall, I'd say the watch hasn't made it out of probation yet. If the GPS had performed better, especially if the app integration had also been there, I might have declared victory already. That didn't happen. On the other hand, it's still too early to declare defeat either. The clock's ticking.