"Strength training" is a complicated idea, especially for endurance athletes. While there's a specific clinical definition of "strength" - it's the maximum (singular effort) amount of force that you can produce through muscular contraction, this is rarely what most people - and especially most endurance athletes - actually mean when they talk about getting "stronger." Ultimately, strength training from a practical standpoint has come to mean, essentially, "any non-sport-specific movement that's hard and relatively short." But the further adrift we get from explicit and clear meaning, the harder it is to have meaningful conversations. I won't wade into the especially murky waters of "functional training" and "functional movement" and their purported promises around "resilience" and a great deal else, though I think it is a topic worth exploring at some point.

Running exists in essentially the opposite space. All running is basically productive as training for running. If someone says, "I ran 30 minutes" or "I ran 5 miles," I think we likely have a fairly similar idea of what that entailed. Cycling is mostly productive, though the massive draft benefit can render long rides in a group much less productive than people realize. "I rode 10 miles" has a much wider range of possible meanings than our theoretical runs. Swimming is likely the closest analog to strength training. Especially when people do "drills," which - especially for adult-onset swimmers - are often of dubious benefit. Someone who says, "I swam for 30 minites" could, in my experience, have done almost anything ranging from something I think most triathletes would agree is "swim training" to something that more closely resembles water aerobics. This isn't meant as a criticism in any way, more just to try to frame the difficulty in having a practical conversation about "strength training." Strength training is a lot like drills for swimming. It can mean almost anything depending on to whom you are speaking.

Even when done effectively, strength training is almost always what I'll call a "secondary" training stimulus. It doesn't directly make you faster. Ideally, it supports your primary training - your swimming, biking, and running - which is what actually makes you faster at swimming, biking, and running. There are some notable exceptions. Explosive lower leg exercises - heavy calf raises and plyometrics - improve lower leg stiffness, which appears to directly correlate to improved running economy.

Though as is typical, there's regularly competing research that shows no benefit to strength training over endurance training alone. As is often the case with small sample sizes, teasing apart signal and noise and drawing larger, more general conclusions is always a challenge. During my professional career, I had an on-again-off-again relationship with strength training, but it was mostly off. If I had more time - or, more specifically since I have essentially unlimited time, energy - to dedicate to training, it was best spent swimming, biking, or running. I got more into strength training since retiring, though it started in a fairly unique way.

Upon retiring from professional triathlon to work at Zwift, I remained - at least in my own mind - an athlete first and foremost. I'd been focused on sport above all else for virtually all of my adult life, and while I had no qualms about giving up racing, it was much harder than I realized to give up what I felt was essentially my identity. I did some gravel racing, but it was too much like Ironman to really captivate me. I was competent and had some good successes - KOM jersey at the 2018 Belgian Waffle Ride and 6th overall at the 2018 edition of the then-Dirty Kanza (now Unbound Gravel), but I did not love it. I discovered track cycling that fall - the Carson velodrome being only about 10 minutes from Zwift's offices in Long Beach - and fell in love. Pursuiting brought me back to my roots as a rower with the just long enough to be extremely unpleasant and just short enough to be really fast. In another life, track cycling could certainly have been my sport. While it wasn't totally insane, it was quite obviously (in hindsight anyway) a ludicrous goal when I decided I'd try to make the Olympic team in the pursuit. This seemed more plausible than it likely ever was because the US men's pursuit team is not particularly good (within the context of the very pointy end of elite track racing).

They failed to even qualify for the Tokyo games, which unfortunately meant that the best American pursuit rider of the past decade (and possibly the best American pursuit rider ever with a World Championship, World Record, and first ever sub-4 4Km to his credit) Ashton Lambie never got to race in the Olympics. Putting all of that historical context aside for a moment, the biggest obstacle for me as a pursuiter - especially a team pursuiter (now the only timed endurance race in the Games) - was that I am not a very good starter. This has always been the case. I wasn't a great starter as a rower. My poor starting ability as a swimmer was really the cause of probably 80% or more of my relative weakness as a triathlete. I just don't have explosive power. This does come with some significant benefits though, as that same physiology largely kept me from being able to blow myself up during long races. But having a peak power of - reliably - high-900w and - on great days - 1000w simply was not sufficient, especially when starting 110"+ fixed gears from a gate as is required for pursuiting.

Thankfully, the recipe for getting faster at starts is relatively simple. It's about strength. Clinical strength. And it's quite clearly established that heavy lifting provides that. But I had two problems. The first was the heavy squats and deadlifts are very challenging to do on your own, especially safely. These are exercises that require a spotter. The second was time. Nevermind the weird delusions involved in thinking that my commute prevented me from being able to get to a gym where I could lift effectively was somehow not going to be a problem when it came traveling to track races in foreign countries. I figured I'd cross that bridge when and if I came to it. In the short term, I needed a way to lift heavy and lift safely.

My long time trainer and physical therapist recommended that I look at flywheel training. There seems to be some debate about the origin of flywheel devices; the story I heard - which is a common one - is that it was pioneered originally by NASA (links to a presentation on a current device) as a way for astronauts to preserve muscle tone in the absence of gravity. But it may be quite older. The research of Per Tesch is a reasonable starting point, and whether that's the actual beginning or not is probably not hugely important. Fundamentally, the science is what it is - moment of inertia is preserved even when weight is not, flywheel or "yo-yo" training devices are not new, but they are newly popular. Driven by - as much of the innovation in sports science seems to be these days - companies in Scandinavia, flywheel training works by providing high eccentric loading. Eccentric loading has long been known to offer significant benefits - Eccentric Muscle Contractions: Risks and Benefits and especially The effects of eccentric versus concentric resistance training on muscle strength and mass in healthy adults: a systematic review with meta-analysis - that possibly even exceed the benefits of classic, concentric load strength training. For a quick primer, eccentric load is basically the "negative" load that's applied when slowing against gravity. When squatting, the concentric load is when you stand up, and the eccentric load is when you lower yourself down into the squat position. The classic example most runners and triathletes know is running downhill and - especially post-race - the agonizing experience of walking down stairs.

Flywheel training works quite simply. You apply torque to a flywheel and accelerate it, giving it rotational inertia. That applies some load, but it's not particularly stimulating, especially as compared with a very heavy concentric contraction. But, upon reaching the end of your range of motion, the momentum then attempts to pull you back down. Especially for a squat, where the flywheel and gravity are working in concert, this force is significant. Combined with the relatively higher load experienced during eccentric contractions, you are able to effectively replicate the stimulus of heavy, classic concentric loading with much less weight. Eccentric loading may even be more beneficial than concentric loading, but most studies are at least convinced that it's equally good.

The particular device he recommended was made by a Swedish company Exxentric. Their kBox device is used by a number of professional/elite track cyclists. I was drawn in particular to the kBox Lite because it is - as its name implies - light and quite portable. Unlike a traditional squat rack, I could bring this with me to work. I also liked that it had a digital measuring device that hooked up to a phone app, allowing you to actually track progress. After using the kBox for a while, I found that my peak power on the kBox matched up almost exactly to my peak power during track starts. I don't use the KMeter as much anymore - I don't need specific numbers to tell me if I've gotten a good workout in - and definitely consider at an optional - but worthwhile - add-on. But I did value the quantitative feedback it provided. But really, it was the portability that sold me. I would take the escalator up to the Zwift office with my backpack holding my laptop, a change of clothes, and my squatting harness and deadlift attachment bar on my back, my flywheel bag with two large and one medium flywheel - more than enough load for even the hardest workouts - over my shoulder, my kBox slung over my other shoulder, and my track bike rolling beside me.



In addition to its portability, the primary reason my trainer recommended the kBox was that flywheel training is relatively quite safe and also does not require a spotter. While you can injure yourself doing just about anything, the specific way in which flywheel trainers work mean that they can only impart as much load to you as you can impart to them. A flywheel will only pull you down as hard as you can pull up on it. If you do a workout like 5x max effort squats, you'll inevitably see a drop off between reps. When doing classic squats or leg press, you might replicate this with a spotter by having them pull plates off between reps, allowing you to fully exhaust your muscles, but this requires a skilled spotter. With a flywheel, this is just what happens. It's virtually impossible to get yourself into a situation where you do one rep too many and fail partway through and need to figure out a safe way to jettison the bar. The specific stimulus of eccentric loading is very similar to plyometrics, but again it's safer than jumping up onto (and down from) a box, though the newer foam boxes for plyos do help a bit here. But plyos require a lot of coordination to do well, and flywheel squats just aren't that demanding.

You can achieve a lot of the same type of stimulus safely with dedicated machines - leg press, etc. But unlike a dedicated machine, a flywheel trainer has nearly the universality of a bar and freeweights. I mostly did regular parallel squats - the king of exercises - but split squats, straight leg deadlifts, high-pulls, and almost anything else imaginable is very doable using the same simple device. While it's best for lower body exercises, you can do a fair number of upper body exercises as well, especially if you use the included floor anchor so that the device stays put without requiring body weight. My total cost for the kBox Lite with kMeter including most of the accessories I thought I'd want - belt for squatting (closer to leg press), harness for squatting (closer to traditional squat), bar (for bent over rows and straight leg deadlifts), single handles for most upper body work, and two large and one medium flywheel - ran me $3155 in 2018, and I expect it will last me a lifetime.

Since I shifted my overall focus in training to be entirely pursuit-centric - I rode the track 1-2x a week, I stopped running, put on about 10lbs, and semi-regularly supplemented with more traditional strength training. I'm reticent to try to tie any specific gains to the kBox directly, especially given what I've said about strength training being largely secondary. But in terms of pure performance, I saw my peak power on the road get up over 1200w and my track starts were regularly in the 1000-1100w range, and I could on good days top 1100w. I've since returned to running and dropped those 10lbs, but I've managed to hang onto most of the explosive gains, largely I think because I never stopped working on it nor did I stop strength training. Running makes me happy, and I love being a recreational multisport athlete. The forced lockdown of 2020 and 2021 helped me finally retire in my own mind, and so I don't have the same obsession with seeing myself as a competitive athlete, but the kBox remains a regular part of my routine, mostly because it makes me feel good and is so approachable. It offers the sort of workout options that I think would have kept me more consistent in regards to strength training when I was racing, especially because it's so convenient. It the rare home gym device that actually lives up to its promise.

While the kBox is adequate for upper body exercises, it's not ideal. In particular, the fact that the force - barring some very creative mounting (which people do in fact do; you can mount it upside down on a ceiling and do eccentric pullups…) - always comes from below you does limit what you can do. I have particularly problematic shoulders - shoulder pain, especially after my accident - was a constant battle. And I think that more upper body strength really might have helped my starts in open water. The Exxentric kPulley fills this gap. There's a fair bit of overlap between what you can do on each device - as with the KBox, with some creative mounting you can do almost anything, including squats, but certainly each has its own set of optimal uses. The kPulley Go - which Exxentric provided for this article - is particularly portable, so much so that I'd say it's viable to bring to the pool to do dryland before or after swimming. Its adjustable height - especially when used with the rail system - really makes it ideal for upper body work, especially overhead exercises. Any of the classic stretch cord exercises can be done on a kPulley but with a much more flexible load profile. As I work to try to keep up with my oldest son, who's gotten into competitive rock climbing, the kPulley gives me a lot of great options for focusing on the health of my admittedly sub-optimal shoulders.

You can put together a very complete kPulley setup for less than $2000. Along with a kBox, the total cost for a home gym that takes up less space than a treadmill, is as portable as a bike trainer, and truly offers access to literally every exercise imaginable in a safe and approachable way is about $5000 total. The system is virtually maintenance free - flywheels are super simple and if, somehow, you do wear out a cartridge bearing, it's about as simple as replacing any cartridge bearing on your bike.

While I think the convenience is perhaps the biggest selling point - training for triathlon takes a lot of time; adding in gym time to that is just a big ask, I think it's really the approachability that flywheel training offers that I love. Telling someone to add in strength training has always felt like telling someone to do drills in swimming. At best, it might do no harm. At worst, it makes them slower or injures them. But based on my own experience, flywheel training is both safe and effective.



I'll cover triathlete-specific workouts in future articles, but fundamentally, I split my workouts into two basic groups - 30sec efforts and 5-rep efforts. 3-5 reps seems to be the sweet spot in terms of maximizing benefit. The unique falloff that you can get with the flywheel makes it easier to do 5-rep-max sets than it is on a traditional machine or with freeweights. The 30sec reps is something that's a bit more unique to flywheel. Unlike with freeweights where momentum both is a bit dangerous and - typically - makes things easier, momentum on the flywheel is what makes it hard. 30sec reps are some of the most uniquely awful things that I think I've done to myself. I consider these to be the more challenging of the two - and perhaps the one to be more cautious about, since I do worry more about form breakdown; 30sec is a long time at high effort - and the one I do less often, but I think it does provide both a unique challenge and a unique stimulus. But the 3-5 rep maximal effort is simple and great for the core exercises like squats and other big muscle group exercises in 3-5 sets - removing flywheels as needed; lower moment of inertia isn't necessarily "easier" in the same way that lighter weight is easier. For smaller muscle groups I typically will do higher reps - 8-12 - with a bit more focus on technique. There are coaches who specialize in strength training programs for flywheel devices, though I think for most triathletes it's probably not something you need. Squats, calf raises, straight leg deadlifts, and hip flexion are the main exercises I'd suggest based on evidence and then for swimming I think it really depends on where you're at and, in particular, if you do or do not have specific weaknesses in your shoulders. Exxentric's own YouTube channel is also an invaluable resource that I refer to regularly. In particular, the leg-specific workout and total body workout are great examples that take you through very complete 20-30min workouts.

As with nutrition, bad advice on this topic is worse than no advice at all. Having discovered flywheel training and the Exxentric devices in my post-racing career, I do feel like this is something that I would have and could have used effectively. Swimming, biking, and running is the most effective training that a triathlete can do to get faster. But I believe effective - and low-risk - strength training can be a wonderful complement, and the flywheel devices from Exxentric provide that in a compact and comprehensive package.

[The author purchased his kBox Lite and accessories through normal retail channels. Exxentric provided the kPulley Go system and some accessories in support of this article. Slotwitch may receive a commission from Exxentric on some sales.]