Fitness trackers are all the rage these days. It seems like everyone has one, but there are so many to choose from. Watches, bracelets, head gear . . . you name it. Author John Bradley [1] from Outside Magazine [2] shares with us his results of what the best one is.

After two months testing 16 different fitness trackers, I'm sitting on the sofa watching Game of Thrones, jiggling my wrist to push the step count on my Garmin Forerunner smartwatch to the 20,000 mark.

I am a shallow, petty man.

A local 10K in the morning, some yard work in the afternoon, and the constant motion that comes with being the father of an infant brought me to 19,841. Even though I've already determined step counts to be a pointless metric for athletes.

I'm too obsessive-compulsive to just leave it at that. Yet, walking around the block to reach a specific count of an imprecise measure of an activity I don't even care about seems less honest than just lying to the Garmin. Better to deceive the watch than myself. So I sit there waving my arm back and forth while Jon Snow strikes an uneasy alliance with the Wildling leader.

It's no secret that fitness trackers are a mess of a success story. Manufacturers shipped 9.7 million of them in 2013, a number expected to hit 135 million by 2018. The proliferation of wearables was supposed to change our lives [3]. We'd run faster and sleep better, get injured less and lose weight. The problem, of course, is that change is hard. While wearables have undoubtedly helped spur millions of people to be more active, the effect for many can be temporary. Indeed, a report last year from consulting firm Endeavour Partners found that more than half of the people who buy fitness trackers eventually stop using them [4]. A third do so within six months.

"What's the common experience for individuals? They get a fitness tracker, and it sparks them to start walking," says John Bartholomew, a professor of health education at the University of Texas at Austin, who specializes in exercise psychology. "So they walk three miles in the morning, and that gets them 4,000 steps. Over the course of a day, maybe they get another 3,000 and do an extra walk to get to 10,000. After a couple of weeks it's, 'I do my walk in the morning, and then I go about my day and I hit my goal.' The novelty of the information is removed. The step count is no longer useful. And that's why people set these devices down."

But that's a sweeping generalization, data averaged out across the masses. Move a couple of standard deviations away from the center of the bell curve—out toward the motivated fitness junkies who wake up at 5 a.m. for pre-office workouts—and things look different. I started using a bike computer to track speed and mileage 25 years ago and have been getting real-time performance data like power and heart rate for more than a decade. As both an athlete and a tech journalist, I wanted to know how wearables were evolving for more specialized users: those of us accustomed to, say, comparing steady-state power output on Strava. The masses can have—and abandon—their Jawbone Ups. What's out there for us?

I spent months working my way through fitness trackers built into watches, bracelets, belt clips, apparel, and jewelry. (See here [5].) I didn't start a new training routine or pick up any new sports; I didn't change my bedtime or set weight goals. I incorporated these gadgets into my life, not vice versa, and then watched what happened. Eventually, I settled on a few that made me not just a fitter athlete but also a more effective worker and a better husband and father.

And I have the data to back that up.

The classic feedback loop for behavior change is: act, measure, learn, modify. If you've ever housebroken a dog you've done this, and you know how important it is to provide feedback immediately following an action if change is going to happen. You also know that there's no point in continuing to offer it once Biscuit has learned to do her business outside.

For activity trackers to spur change, they need to be part of an effective feedback loop, and they need to continue offering novel information. Data telling you what you performed—a workout or a certain number of steps—is not novel. Data telling you how you performed—speed, heart rate, power—is. A lot of wearables, including most step counters and sleep trackers, fulfill only the measure step. As a study in the Journal of the American Medical Association noted earlier this year [6], "The gap between recording information and changing behavior is substantial … and while these devices are increasing in popularity, little evidence suggests that they are bridging that gap."

Critics of fitness trackers—and there are many [7]—fault them for not operating more like medical devices. They want their wrist straps to give them their resting heart rate and also tell them that their LDL cholesterol is too high and to prescribe statins and a meal plan. In March, Wired ran a story titled "Fitness Trackers Won't Really Help Until They Tell Us What to Do [8]." It argued that fitness trackers mainly measure problems without offering solutions.

Here's the thing: in a lot of cases, the problem isn't that fitness trackers don't offer advice; it's that they're actually prohibited from doing so.

The FDA defines a medical device as a product that is "intended for use in the diagnosis of disease or other conditions or in the cure, mitigation, treatment, or prevention of disease or is intended to affect the structure or function of the body." These devices are subject to onerous regulations and approval processes. Given our understanding about the roles that exercise and healthy lifestyles play in disease prevention, apps and wearables designed to measure and promote exercise butt right up against those regulations.

Earlier this year, the FDA issued new guidelines [9] meant to help clarify things. While apps and products that help people record workouts or make decisions about health and wellness could meet the definition of a medical device, the FDA says it will "exercise enforcement discretion," meaning they're mostly cool—for now. But the agency says it will look much more closely at products that use "built-in features such as light, vibrations, camera, or other similar sources to perform medical device functions."

In other words, as soon as a company starts pulling data from a sophisticated piece of hardware that reads physiological markers like blood pressure and oxygen levels, regulators might get involved. That means engineers are doing far less with wearables than they could be.

"We know there's a lot more potential," says Alex Frommeyer, founder of Beam Brush [10], an electric toothbrush that pairs with an app to measure and promote dental hygiene (and is regulated as a medical device). "If the regulatory environment was completely different in how you could look at the prescriptive and diagnostic side of the world, we could have considered investing more deeply in those technologies."

For now, the available fitness trackers can measure and nudge, and that's about it. But if you're already motivated, that's plenty, especially if the data you're recording remains novel and engaging.

"From an athlete's perspective, you are somewhat variable in terms of both your workouts and how you respond to them," Bartholomew says. "Your heart rate's not always going to be the same, for example. Tracking for athletes continues to give interesting information in a way that tracking for the general population does not."

Moments after I set out for a seven-mile evening run, a woman's voice, calm and Siri-like, interrupts the Japandroids song blasting in my Jaybird headphones: "You are averaging 156 steps per minute. You should be averaging at least 162." Over the next hour, she'll continue to check in with cadence numbers as well as mile splits and reminders to keep my back straight and to strike the ground with good form. "Try shortening your stride," she suggests.

The voice is from my Moov Run and Walk app [11]. The company, which launched last year on Kickstarter, also has free apps for cycling, swimming, and the gym. The 1.5-inch puck strapped to my right ankle, which pairs with all the Moov apps, collects data on cadence, range of motion, stride length, and impact. Once I end my workout, I can view all this in the app's graphical report, with a table of my splits and with charts showing metrics for performance and form down to degrees in my range of motion and the G-forces of my foot strikes. The Garmin Forerunner 920 XT [12] watch on my wrist does the same stuff, but the Moov's in-ear reminders have me hooked. The 920 is there for post-workout analysis and trends—something Garmin presents better than any other company in this space.

In the swimming pool, the two devices switch roles. The 920, with its huge display and broad capabilities around timed intervals and customized workouts, provides in-the-moment feedback. The Moov sits quietly on my opposite wrist until I get back within range of the Swim app running on my iPhone. It grabs the data from the puck and gives me an in-depth look at my workout, not just total distance and times for every lap, but also time and distance per stroke, fastest splits, and longest uninterrupted swim.

Crucially, Moov also explains what the numbers mean and what I should be shooting for. My 2.26 seconds per stroke is well off the 1.5 to 1.7 seconds that the app says strong distance swimmers average. So I start focusing on my arm turnover. Within three weeks, I'm down to 2.12 seconds and I'm getting a bit more distance with each stroke. I am objectively a better swimmer.

These devices are measuring the same things myriad other trackers do—motion, distance, speed. At this point in our technological history, the trick isn't gathering metrics; it's presenting them. Garmin and Moov package tens of thousands of data points in a way that's both understandable and addictive. Proof? Moov says greater than 75 percent of those who bought the device are still using it more than a year in.

"Intrinsic motivation is based on enjoyment," Bartholomew says. "Learning about yourself and testing yourself and demonstrating competence are inherently enjoyable. So from an intrinsic-motivation perspective, you need to keep learning. As long as you're getting useful information, you'll continue to track."

And some companies are finally delivering on that promise.

Moov and Garmin are the pinnacle of this first wave of fitness trackers. The next generation is something else entirely. The tech startup Athos [13] has come out with compression shirts and shorts embedded with sensors that provide feedback that even the best personal trainers can't match.

The garments send the data to a smartphone app that displays it as a real-time heat map on an anatomical diagram. Set your phone in front of the squat rack and you can see how well you're using the target muscles, if you're engaging them with the right intensity, and where you have imbalances.

At Unbreakable Performance [14] in Los Angeles—a gym with numerous pro-athlete clients, including about 30 NFL players—trainers use Athos to gauge users' initial form and guide their workouts. "I get to see which muscles are firing and in what order," says trainer Gabe Rangel. "If I see that the left hamstring is firing and the right isn't, I know something's wrong. If we're doing sprints and I don't see glutes firing, I know they're not getting full extension."

As I push my way through a heavy set of squats, the eight sensors on the shorts immediately pick up on the weakness in my left hip that a physical therapist diagnosed a couple of years ago. The heat map shows left and right firing at different times and intensities. I press a button on the screen after my set and the app displays muscle-effort data for my glutes, hamstrings, and inner and outer quads. All the numbers are lower for my left side, save for the inner quad, which was doing almost 50 percent more work than my right inner quad, presumably to compensate for the imbalance.

For my next set, I reduce the weight and watch, in real time, as my focus on form slightly evens things out: another smart feedback loop leading to real change; no need for a visit to my PT.

The Athos app can also play back entire workouts, which is helpful for things like running and cycling.

Still, as much as I appreciate what Athos offers, I use it only once a week or so, to check in on my form. Even the 920 and Moov come out only when it's time to train. After exhaustive testing, the only tracker I wear daily has nothing to do with athletic performance at all. At least on the surface. It's called Spire, and it was based on research at Stanford University. It clips inside the waistband of my pants and monitors the rise and fall of my abdomen to record breathing patterns.

Spire uses this data to gauge whether I'm stressed, focused, or relaxed—states that it displays on a smartphone app. I can set goals—three hours of focus per day, two hours relaxed—and the app shows progress on an elegant clover-shaped graph. Customized alarms notify me when I've been tense for five minutes or after a calm streak of at least three minutes has ended.

Additionally, once every day or two, I plug in my headphones and choose a Calm, Focus, Energize, or Meditate session. The app then walks me through a few minutes of breathing exercises geared toward putting myself in those states. As I follow the commands for inhales and exhales, I can watch the oscillating line on my screen—my breathing pattern—become deeper and more rhythmic.

"This is one of the few actionable things you can do," says Spire [15] cofounder and CEO Jonathan Palley. "You can't really consciously change heart rate. What you can do, though, is take a few relaxing breaths. A long exhale isn't something you do when you're in danger. It triggers your mind to think, OK, if I have time to slowly breathe out, then nothing is going to attack me right now."

Being relaxed has a tangible effect on your performance. Athletes with high stress levels are more likely to get sick or injured. In 2012, Bartholomew and a co-researcher at the Yale Stress Center found that stress hinders recovery [16] after a hard workout. They recommend that "athletes and others undergoing bouts of strenuous exercise should undertake strategies to obviate the negative effects of chronic mental stress."

Spire can tell you you're stressed before you're even aware of it yourself. "Eighty-four percent of the time," Palley says, "when Spire surfaces that someone is becoming stressed, the user's breathing pattern will change within 90 seconds." Another feedback loop in action.

Contrary to criticism that mobile technology takes users out of the moment, Spire has taught me to be more connected—to listen to my breathing for a sense of how I'm reacting to things and to use that knowledge to be more present in what I do at work, at home, and while training. My wife noticed enough of a change that she's getting one for herself. I'm certainly no Zen monk, but I'm better than I was.

I started with 16 fitness trackers and worked my way down to four. That means I abandoned 75 percent of the ones I tested, all within just a few weeks. That's an awful return. But the four I didn't discard have become part of my regular routine—one that is now healthier and more productive. So is this product category failing or succeeding?

Both. Understanding why makes all the difference.

Three key mistakes have hurt people's relationships with fitness trackers. The first is a tendency to lump them all together. An entry-level pedometer is not the same as an Apple Watch with optical heart-rate detection, and that Apple Watch is not the same as a Garmin Forerunner 920XT that can record swim workouts, analyze running form, and estimate V02 max. Just because a basic step tracker doesn't change things doesn't mean the entire category is without merit.

"The idea that a single wearable is for everyone is wrong," Palley says. "These are specific tools designed to help you do something better that's both very personal and very diverse."

Second, from the consumer perspective, the focus on hardware is misguided. People buy a tracker, then build their goals around what it can measure. The trick is to start with what you want to accomplish, determine which data best reflect it, then find the hardware and apps that deliver that data. Goals, data, technology, in that order.

The third mistake is the most important: a failure on the part of both buyers and manufacturers to understand feedback loops and the psychology of behavior change. The data must be novel and relevant, but also presented in a way that encourages and measures transformation.

"The big challenge with apps or fitness-tracker technology is that, in general terms, they're asking a huge percentage of their customers to do a new behavior," says Beam founder Frommeyer. "So the premise is, you have to wear something new, every day—a physical change to your appearance. And after you make that change, you have to adjust your behavior."

A few months ago, the Apple Watch entered this space [17] with huge expectations. It's a watch—and an elegant one at that—so for most people it doesn't require a change in appearance. But the only attractions for me lie in the productivity features—things like texting, screening calls, and checking my schedule from my wrist. Those ultimately weren't enough to keep me coming back. (And I'm an Apple guy: iPhone, iPad, two MacBook Pros, and two Apple TVs.) In terms of fitness tracking, the Apple Watch is all steps and daily goals. Even the standalone exercise app measures workouts only in terms of time, distance, heart rate, and calories. It's a diary, not a coach.

That was the takeaway for most of the wearables I tested. Of the trackers I started with, only the Garmin, Moov, Spire, and Athos shorts have hooked me with their data while asking nothing new of me. I simply go about my workouts or my daily routines as I normally would, except with continuous opportunities to adjust and improve. It's not about hitting huge 24-hour goals built around arbitrary metrics. It's about thinking, from moment to moment, how I can train and live a bit better.

That gap the JAMA researchers found between recording information and changing behavior hasn't closed, but it's starting to narrow.

Check out more great stories from Outside Magazine:

• The "Trail Is Better Than Pavement" Lie [18]
• The Next Big Sports Food Is . . . Algae? [19]
• Your Sorry Attempts at Pull-Ups Are About to Be a Thing of the Past [20]