It's well worthwhile to read through the linked article, but I'm going to summarize the main points here.

Most cyclists with a power meter are familiar with the concept of Training Stress Score (TSS). This is an attempt to correlate training load with changes in fitness, and was itself developed based on the earlier heart rate-based TRaining IMPulse model.

Every ride is assigned a TSS, based on its intensity (fraction of FTP) and duration. Quite simply, the longer and harder you ride, the higher the training load (TSS), and the higher the training load the higher your fitness should become. High training loads also lead to high fatigue levels and the potential for overreaching and injury if you increase loads too quickly or beyond levels you can tolerate; the key to effective training is to balance optimal load with recovery.

The problem with the TSS concept is that it tries to track fitness via a single one-size-fits-all number. The way it's calculated means you can get a TSS score of 100 in many different ways. Three examples: riding for 60 minutes at FTP; riding for 60 minutes doing 30 second sprints at 250% FTP every few minutes with easy recovery in between; or riding for 4 hours at 50% FTP.

Anyone who's done these workouts knows that each has very different physiological effects, so clearly they're not interchangeable.

Each of us has a different natural ability and response to training. Some of this is down to genetics, which has a large effect on body type, cardiovascular capacity, muscle typology etc. So it stands to reason that we'll all respond differently to different methods of training.

In the linked article, Couzens discusses two of his athletes, both of whom are fast — winners of major endurance events — but each responds very differently to training. One of them thrives on 25-hour weeks of long slow distance, with low average intensity. He's a 'volume responder': the more he rides, the fitter he gets, almost regardless of how hard his riding is. The other is an 'intensity responder', who does much better on lower volumes at a much higher average intensity: he does a lot of high-intensity interval training and fewer endurance miles.

If we want to optimize our training, it would clearly be beneficial to us to be able to find out if we fall into one of these categories, or if a more balanced approach would suit us better.

The first step in doing this is to move beyond TSS, by separating out the two main terms in the training load equation: volume and intensity.

Couzens shows how to do this using your power and heart rate data. You look at each of your training blocks and get the totals for volume (hours ridden) and intensity (TSS per hour). You also have to calculate your relative VO₂ capacity, which requires an extended maximal effort at the end of every block.

Once you've got this information, the first thing to do is plot VO₂max score versus training load (TSS per day). This gives an idea of how fitness responds to just the raw TSS numbers, with volume and intensity not separated. Mine are shown below.

I took the liberty of adding in r² values for each of these graphs, to judge the strength of the relationships plotted. As you can see, for me there is only a very weak correlation between my training load and fitness; it appears that a higher TSS is probably beneficial, but the effect is rather underwhelming. You can see that there were a couple of blocks in which I had a much higher training load than in the others, but considering the extra effort involved, this didn't make a great deal of difference.

Hopefully we can get more insight by separating out volume and intensity. First, fitness versus training volume:

Ouch! Not only does increasing volume not help, if anything there's an inverse relationship: the more I ride, the worse I get! I'm not taking that too seriously, though, since the r² value is so tiny (you can see the data points are all over the place, and don't cluster tightly around the line of best fit, especially given a big outlier on the left that's skewing things).

What it does mean is that, at least in the range of the data (about 13-18 hours riding per week), extra volume doesn't do much for me: I may as well ride at the bottom end of the range so I can save some energy and recover faster.

At this point all my cycling buddies who've been trying to get me to ride less over the last couple of years are laughing their asses off at me. Still, there is one way that higher volume helps: calorie burning. I'm currently on a weight loss program, and due to the way VO₂max is calculated (the body's oxygen consumption, in millilitres per kilogram body weight per minute), any loss in body weight does actually improve relative fitness. And clearly, the more miles I rack up the easier it is to lose weight.

Furthermore, one important long-term factor in fitness improvement is cycling economy. This increases very slowly over many years, and the biggest factor in its development is total hours ridden. So maybe that's another reason to ride more, but not if it's at the expense of other aspects of my training.

Other than that, there's no reason for me to ride more than (at most) the low end of the volume range I've been riding at. Okay, great. Let's move on and look at fitness versus training intensity.

Now we get to the good stuff: a tight cluster of data points with a good strong correlation (r² = 77%), and, even better, it's pointing in the right direction!

This is a nice steep line, and needless to say it's given me all kinds of ideas about how to experiment with my training over the next few months. Lately I haven't really been enjoying my very long rides, and now I don't have to think twice about reducing their frequency, while I instead shift my focus onto intensifying my training.

The first thing I'd like to see is a wider range of values on the X-axis: I need to do some blocks with an average intensity score in the 60s to see what happens. I don't want to extrapolate too far, but I'm optimistic that if I do this (and get the training/recovery balance right) I'll see some good improvements in the near future. But more data points are certainly required.

What about improvement over time? When I started riding three years ago my fitness level was very low, so there was a lot of scope for improvement, and I certainly saw huge gains over the first twelve months or so. Considering the low starting point, this would probably have happened almost no matter what kind of riding I did. But since I wasn't training with power during that time, I can't include this in my analysis.

But over the last year that I've had my power meter there's been no obvious relationship between my fitness and time, so it seems that I'm at a point where if I want to see significant further improvement, I've got to pay attention to the details of my training. Just guessing what's best won't work anymore: I need to put together a series of specific, focused and consistent blocks with a test at the end of each one, modifying my training based on the results.

Here's a figure from the article, showing how athletes group by type:

The guys who win everything are in the green group, most people are in the blue group, and we all know a few purples!

It appears that I'm in the yellow group (actually, off the top of the scale). Unfortunately, until recently, I spent most of my time training like a volume responder; I nearly killed myself for much of 2019 repeatedly stringing together 20-hour weeks, during which time I seemed to get slower and slower, although this was probably down to exhaustion.

In fact, the likely explanation for the total lack of correlation between volume and fitness for me is that any gains from extra volume were offset by fatigue. If I'd taken time out for recovery when I was in the high volume period last year things might have looked a little better. Indeed, this year I've done just that, recording my best performances across the board in the last few weeks.

This analysis has reinforced what I've increasingly been suspecting: that I need to focus on riding harder, not longer. But intensity isn't a monolithic entity, so I'll need to go into even finer detail. I'm going to be keeping the volume relatively low for a while (averaging around 12 hours per week for the next couple of blocks), while I focus on intensifying my training. But there are several ways I could do this.

If I want to bring my average intensity for a block up to 65 TSS/hour, I could just do a lot of tempo rides (riding at 80% FTP), or I could do a combination of endurance rides and threshold and/or VO₂ sessions, or just do a mixture of fairly short base training rides and very hard anaerobic sessions. And these will all have different physiological effects.

The first option is out, as I want a combination of easy and hard rides, not a lot of moderate ones. So it's mostly a question of getting the right balance between work above and below VO₂ capacity, within the right proportion of hard and easy days. It's time to experiment, and, as ever, I'll be posting updates on how things are going over on my Training Notes page.

If you've got your own power and heart rate data, and you perform regular maximal long efforts of at least 20 minutes, then you've also got the information you need to do this analysis¹. I think it would be well worth your while; maybe it'll lead you to a new way of training that works better for you.

1 You can calculate TSS scores and training hours using the Strava plugin intervals.icu (as I discuss here). Get the numbers and plug them into a spreadsheet.

#cycling #training #volume #intensity

The 4000-m team pursuit cycling world record: theoretical and practical aspects Schumacher and Mueller, 2002

This is an extremely interesting research article, not least because it gives a detailed description of the training practices of world-class athletes. Such information is generally hard to come by, so when I find some I give it my close attention!

The paper actually begins in quite unremarkable fashion, detailing what were at the time the necessary power requirements (an average 480 Watts for the four team members over 4 minutes¹) to secure a world record in the track cycling Team Pursuit event, modeling the workload and physiological demands of the event, and giving an equation to estimate winning times in future competitions based on regression analysis of previous winning times.

However, for our purposes the most interesting material is the included discussion of the team's training program. Given that the event in question takes place over a distance of just 4 kilometres and thus has a high anaerobic component (with a need to maintain 100% VO₂max power – well above lactate threshold – for the full duration of the race), you might imagine that the riders would focus heavily on anaerobic training, such as high-intensity intervals. In fact, there was surprisingly little of this until the final days immediately prior to the event, at the 2000 Olympic Games.

Overall, the training in the months leading up to the event was dominated by high mileage, low intensity rides (3-8 hours/day; 29,000-35,000 km/year), referred to in the article as 'basic training'. This is exactly the same as the old-fashioned base training rides employed by road racers in the off-season, done at 30-50 bpm below lactate threshold heart rate (Z2 in a 5-zone system).

This training led into periodic workload peaks in the form of one or more road stage races, which increased in frequency and difficulty as the Olympics date grew nearer, and culminated in a short period of track training to finish each of 3 macrocycles (see Figure 1; note that the darkest boxes are the track training, and that blank sections involved unstructured light training at home).

The final track training period covered the 8 days immediately prior to the Olympic Games, and the two earlier ones were just 4 days each. That's 16 days of event-specific training in 6 months! This track training was divided into 'evolution' training (6 minute blocks within 5bpm either side of lactate threshold heart rate), and event-intensity 'peak' training (1- or 2-minute blocks at maximal 4-minute power). Both evolution and peak training were performed at around 130 rpm with 20 minute breaks between repeats. As shown in Table 2, a maximum of only 7 evolution and 2 peak blocks were carried out in a given day.

The training distribution outside of the stage races and track training (i.e. 70% of the training days) was 94% below anaerobic threshold, 4% around threshold, and 2% above threshold. These numbers would change with the inclusion of the stage races, although even the tougher races would also consist predominantly of low-intensity riding, punctuated with periods where increased effort was demanded.

The purpose of the track training was to boost anaerobic capacity, as well as allowing focus on discipline-specific technical and motor skills. As we have seen, however, this didn't require a great deal of time to accomplish (about 8% of total training days in the six months leading up to the Olympics).

The remainder of the paper discusses in further detail specifics related to the determination and track-order of team members.

However, for us the lesson is clear (and there's a reason that base training is so named): aerobic adaptations are the major physiological determinants of performance, even in an event consisting of just 4 minutes of maximal effort. Anaerobic capacity must be added on top of this.

Because of the much more rapid recovery following low-intensity riding compared with training at threshold or higher intensities, this type of training can be repeated consecutively for many days, allowing a training volume (and associated adaptations) to be amassed that is much larger than would be possible when training at higher intensities. This is an ideal way to build a potentially huge fitness base.

Furthermore, in addition to being built upon and benefiting from aerobic adaptations, many anaerobic adaptations can develop over a much shorter time frame than aerobic ones (partly because they actually have less development potential). Understanding why this is true necessarily involves examining the specific physiological changes that we're talking about in some detail, which is a large topic that we'll look at in future, across multiple articles.

The big caveat for amateur riders when applying this to ourselves, of course, is that most of us don't have the time and resources necessary to emulate this kind of training. So we must make compromises. This also is a huge topic that I'll repeatedly address in future articles. A clue for now: if we can't match the volume, we must modify the other term of the training equation i.e., intensity. But exactly how we do this is of great importance; some ways are much better than others.

1 Note also that, prior to the Team Pursuit, two of the four riders on the team finished first and second in the Individual Pursuit event.

#training #intensity #olympics #worldrecord

#cycling #research #science