If there is one non-cycling book worth reading this year, it should probably be Thinking, Fast and Slow by the Nobel Prize-winning behavioural psychologist Daniel Kahneman. To paraphrase the product description from the publisher: “Two systems drive the way we think and make choices, Kahneman explains: System One is fast, intuitive, and emotional; System Two is slower, more deliberative, and more logical. Examining how both systems function within the mind, Kahneman exposes the extraordinary capabilities as well as the biases of fast thinking and the pervasive influence of intuitive impressions on our thoughts and our choices.” Why this is interesting in the context of this blog post will be returned to below.
Talking of books, VeloPress is publishing in North America the fantastic Slaying the Badger by Richard Moore, who argues, eloquently if not entirely convincingly, that 1986 was the “greatest Tour de France”. The obvious rival Tour to this claim is 1989. As we know, Greg LeMond beat Laurent Fignon at this edition because of aerodynamics. In an article in issue 10 of Peloton magazine, John Wilcockson goes into this in some detail. Interestingly, wind tunnel tests after LeMond’s win revealed that his aero bars were only worth an 8 second gain as LeMond’s position on the bike (which he had spent much time perfecting) was already highly aerodynamic. But the bars were not the reason for his narrow win, as his helmet was actually acting like a “sort of parachute” and cost him 12 seconds. Therefore, it was more LeMond himself, not his bike and gear, who beat the ailing Fignon. This is worth keeping in mind as this discussion progresses.
Going faster on a bicycle involves overcoming resistance: wind resistance, rolling resistance from the tyres, bearing friction, and gravity. At speeds below 13 kph, the dominant forces are rolling resistance and bearing friction. But once these are overcome, they increase only slightly with speed – once you’ve gotten rolling there’s not much to hold you back. Gravity is a constant and only changes depending on the gradient of the climb; you can out sprint gravity. Wind resistance, however, increases as the square of speed over 13 kph – it gets harder to go faster (refer to Ed Burke, High-Tech Cycling, for much of the technical information here). The higher your speed, therefore, the greater importance of increasing aerodynamics but the less absolute benefit you will gain.
Weight, so important in climbing, is much less of a factor on flat roads (although it does have some role in acceleration). For example, reducing the drag of a bike by having its cables inside the handlebars and frame (reducing drag by about 10 grams) is equivalent to dropping over 2 lbs in weight over a 40 km time trial, but even then it is only a handful of seconds. For climbing, the key point is that aerodynamics can play a role, but it is going to be a minor one because of the relatively low speeds. On a 10 km climb, the best you might hope for is around 1 minute in time gain in theory: if you can reach speeds of 20-24 kph in some sections and you can maintain an aero position (and still get the same power output) for the duration. Aero equipment, such as wheels, will give you a time advantage somewhere south of this figure. If these are not the conditions then the gains from aerodynamics are going to be much less. But still, because air resistance is lower at lower speeds, there are decent absolute gains to be made relative to your speed. Any time a long climb levels off a little for a decent distance and you can kick it over 20 kph, ‘getting aero’ will be a handful of seconds of advantage (for more on this, see part 3 – coming soon).
Pantani getting aero in the drops while climbing.
One example of the important of aerodynamics in cycling in general is the hour record. In one study, the researchers charted the power outputs of the hour record holders, based in some cases on power meter readings where available but on calculation in others. They were equalized for comparison, which was a difficult process given the difficulty in calculating wind resistance for different shaped riders (the outsize Miguel Indurain, for example, was a problem). When Eddy Merckx posted 49.431 kms, his corrected power was 429 watts on average for the duration. Moser, Obree and Boardman all bested this record but they did so by producing less power – aerodynamics was their advantage. It was not until Indurain that power levels were higher than Merckx; Indurain recorded 436 watts in 1994. Tony Rominger (infamously coached by Michel Ferrari, whatever that might mean) produced 460 watts when he pushed the record over 55 kms, with a standard bike but aero wheels and an aero bar. Chris Boardman’s record of 56.375 km was achieved with less watts than Rominger (but more than Merckx) using an aero bike and the ‘superman’ position.
When Boardman beat Merckx’s traditional record (and only just: 49.441 km) in 2000, on a standard track bike following the UCI regulation change, he produced less power so must have had a slightly better position (not that it was easy – he couldn’t walk for four days afterwards). Overall, since the 1960s, the study estimated that the gains in the record were 40% attributable to riders producing more power and 60% due to aerodynamics. The takeaway message is that aerodynamics matters quite a bit, at least in the rarefied world of adding just a few more metres to the hour record, but so does wattage – and the latter is up to the rider to produce.
Position versus equipment
The drag of your bike is around 25-35% of the total. In other words, you the rider are 65-75% of the drag (some studies suggest that it might be slightly higher). The rider’s position on the bike is the key factor in reducing drag, and studies have concluded that gains can be up to 5-6 minutes over a 40 km flat course. The difference made by aero wheels, in contrast, is only around 1.5 minutes, according to one study; this is with all other factors being equal – if you cannot hold an aero position, for example, you might offset any benefit from the wheels. As well, because your body makes up the majority of the drag, even wearing a skinsuit and riding standard wheels can be as much of an advantage as regular kit and aero wheels.
So, the first step to reducing your drag and overcoming wind resistance is going immediately to your LBS and getting them to set you up correctly on your bike so you can get your back as flat as possible on the hoods and comfortably reach the drops to get even more streamlined. You should then watch Fabio Cancellara and spend time working on it until you can mimic his position on the hoods (refer to 2012′s Milano-Sanremo, for example). You may also wish to include his simulated aero bar position where he rests his forearms on the flat of the bar and emulates an aero bar position (it’s harder to do than it looks, while not weaving from side-to-side and thus negating the drag reducing effect; Cancellara is of course not the only pro to use this position – Tom Boonen did pretty well at Paris-Roubaix this year using it). It all stems from Roger de Vlaeminck, the king of the ‘on the hoods’ aero position. Let the following picture be your guide.
The current focus on aerodynamics means that there are lots of misconceptions about the scope of the gains to be made. For example, in the same issue of Peloton magazine, in a review of the $2,700 Bontrager Aeolus 5 D3 wheels, reviewer Ben Edwards writes: “It’s speed you will feel palpably, on your first ride.” (In the same review, Edwards notes that the 1,550 grams weight of the wheels excludes them from being “a dedicated climbing wheel”, another sort of myth that the next part of this ultimate climbing guide will address.) There are figures of 10 grams of reduced drag over other aero wheels used in the article, which, as Edwards notes, equates to less than 1 watt of energy at 40 kph (or, as is noted in High-Tech Cycling, the drag produced by holding a pencil up in a 30 mile-per-hour wind). Yet Edwards is apparently able to notice this difference, in particular whilst coming out of corners on a fast ride as “they corner as if on proverbial rails… and you will be rewarded with a gap… you will go farther, faster with less energy.” To be able to notice such a performance difference, just 1 watt, is indeed very impressive. But it is entirely fanciful. The limit of aerodynamic gain in a cornering situation, according to a rough calculation at Analytic Cycling, is at most half a wheel length or about 0.02 seconds. Are we really to believe that someone can discern this difference whilst on their bike and attribute it solely to the wheels?
Perception v reality
Let us then see where Kahneman fits into all of this. Our System One way of thinking is basically our instincts. This system, which is active all the time, is constantly making assessments and judgements about all manner of things in our environment. It is very good at its job, but it is also easily distracted. It is particularly bad at complex problems involving numbers, statistics and probabilities. It is also easily fooled by optical illusions (the famous ‘which line is longer’ test) and prone to biases, particularly when there is ‘anchoring’ or ‘framing’ involved. The vulnerabilities of System One is why we have System Two, our considered mode of analytic thinking where we carefully consider a problem before reaching a conclusion, to balance the impulsiveness of System One. This is why Kahneman’s book references fast and slowing thinking – System One and System Two, respectively.
The main conclusion we should be aware of is that our instincts, perceptions and snap judgements can be wrong. They are highly subjective and easily influenced by other factors. It would be very difficult, in an objective sense, to measure the speed difference between two wheel sets on the open road. In a cornering situation, for example, you as the test rider would have to mimic the exact wattage, riding position, and line through the corner on each set of wheels to control the major variables. Given that riding position is the most important aerodynamic factor, it would be nearly impossible to keep it uniform in a meaningful way. But this apparently does not stop reviewers’ instincts from taking over and proclaiming that one set of wheels is noticeably faster than another.
One of the particular biases that influences System One is the so-called anchoring effect, where we place an over reliance on a particular piece of information in making our judgement. In this context, if we are told that a very expensive wheelset is more aerodynamic than another, and that this has been proved by wind tunnel testing, we are – one would argue – more likely to conclude by riding said wheelset that it is faster than a more modest set of wheels. In fact, we might conclude that it is even faster than the tests showed and say that it is worth a bike length in a sprint when the actual objective laws of physics will show that this is not possible (all other factors being equal). This would be a good case of System One (our instincts) versus System Two (the objective application of the laws of nature). The anchoring effect will be stronger if the information comes from an authority source (like a respected bike magazine, for example) or if it confirms strongly-held beliefs or conventional wisdom (aerodynamic wheels make bikes go lots faster, for example).
In all manner of situations, people will often say, “Yes, but what about in the real world” as if there are different rules for how things work in different environments. In the real world, people often pride themselves on their instincts and how good they are at making snap judgements. This is particularly the case if those judgements confirm strongly-held beliefs. In many cases, psychologists in experiments have found that subjects will cling to their beliefs with even more tenacity when they are exposed to contrary evidence. They simply refuse to believe that their judgements are wrong. Our System Two mode of thinking is a powerful analytical tool, but we are prone to ignoring it, under-utilizing it, or refusing to believe its conclusions.
It is extremely improbable, if not impossible, for anyone to be able to perceive the drag difference between different wheelsets coming out of a single corner. Sure, they may feel different or even faster, but that does not mean they are noticeably faster in an objective sense at a specific point in time. The speed gains will be cumulative over time, but not immediately discernible. But how do you refute someone who claims that they “feel palpably” faster or that they got a gap riding the wheels? (Although you could say, “Wow, you can feel the <0.02 seconds in time difference coming out of the corner – that’s incredible!”) Paying $2,700 is a lot of money to “feel palpably” faster. One could almost guarantee that buying a pair of handmade Rapha shoes and drinking two negronis would make you feel faster, too; still a bit pricey, but you’d have change left over.
What it means
Aerodynamics matters. Reducing the drag of the rider matters a lot more than reducing the drag caused by wheels. At least Bicycling magazine, in a test of aero wheels that included the Aeolus 5, says that “measurable differences are pretty small” and includes a quote from Steve Hed: “Early on, our comparison was to a box-section Mavic rim; we’re not saving anyone a minute over 40 km anymore. Now it’s more like seconds.”
So, according to Hed, the drag gain was 60 seconds over 40 km, or 1.5 seconds per kilometre (0r 0.0015 seconds per metre). That is pretty consistent with the figure cited in the study noted above. They are little gains that add up the longer you ride. If you are a pro cyclist, seconds matter. Just one second might be the difference between first and second in a long race, or small aerodynamic gains add up over time – like on Boonen’s Paris-Roubaix breakaway (although how much was down to aerodynamics, particularly as he is a big rider, and how much was his incredible form is a big question). When it is your livelihood, it matters; you will do anything to get a possible advantage. For amateurs, though, should we be obsessing as much over a handful of seconds? Furthermore, as has been argued previously on this blog, controlling all the variables so that a tiny reduction in wheel drag that is measured under controlled conditions does make a quantifiable difference on the open road is very difficult, if not impossible. Does that make $2,700 for a set of wheels worthwhile?
When reading bike and component reviews, we usually have our System Two modes of thinking in full function. We understand that reviewers have to write something about their riding experience and that a simple account of facts and figures would be extremely boring. We ignore highly subjective performance claims and we know that if Freddy Maertens circa 1976 on his bike of that day was transported forward in time to our local crit, he would easily out ride each and every other competitor on their 13.6-pound bikes with aero wheels. We know that it is the rider that matters most. We know that phrases like “cornering on rails” have no objective meaning because there is nothing that can be quantified and compared. We also know from our own experience that cornering has more to do with a rider’s line and ability than equipment – although upgrades can give us more confidence in our our abilities.
So why does it matter what gets written? Because our System Two is lazy, we might fall into the thrall of System One thinking and trust our instincts, particularly if it confirms our biases, especially if that bias is a mantra in the bike industry: that spending a lot more money will make you go a lot faster. We might start thinking that the gains from equipment are more than they actually are. And this is a sloppy way to approach a sport and a pastime. An alternative review, therefore, could read like this:
In controlled conditions these wheels will save you 1 minute in a 40 km time trial over a standard set of wheels. In most cases on the road, you would only save this amount of time if you kept all other drag factors constant, something that is difficult but not impossible to achieve. These wheels may not feel immediately faster when you ride them as their initial speed advantage is small and only accumulates over time. You might find they give you, all other factors being equal, a bike length or so of advantage at your mid-week crit, but the young gun on a borrowed cross bike will still beat you to the line. Conversely, however, you may feel quite a bit faster on these wheels – even if the speed difference is tiny – simply because they are superbly made and presented and appear to be really fast. The thrill of new wheels may prompt you to push yourself a bit harder. If a potential gain of 1 minute over 40 km is important enough to you to spend $2,700 (if you are not a pro rider already, in which case your sponsor has given you these wheels) and replace the $500 wheels you already have, then go for it. However, you may wish to first attend to a number of other factors – such as your ability to ride in the drops or an even more aerodynamic position for long periods of time. In fact, you should do this right now – and throw in some intervals – instead of reading this review and obsessing over carbon wheels. Your abilities as a rider will have more of an impact than your equipment.
The review in Peloton magazine, an otherwise fine and excellent publication well worth reading, has been singled out in this blog post but it is not the exception in the supposedly objective product reviews that we get in bicycling publications. We should be mindful as to how biases influence subjective judgements. We should understand that objective comparisons are not possible outside of controlled situations. And it matters because the cost of bicycles and components are – at the top end – skyrocketing, even while we enjoy the great benefits of trickle-down to the lower end. It matters because at some point we might start believing that $2,700 aero wheels are essential for going faster. And to go faster is what everyone wants, right?
There are a number of good reasons to buy high-end bikes and components. Performance is one of them, and the potential benefits are (mostly) very real. But they are not as much as you will be led to believe by reading magazine reviews.* It is not honesty that is needed but rigour (or rigor, for US readers) in review writing. The facts and figures are all there for anyone to access, and a little System Two thinking will go a long way to confirming or disproving what our System One instincts might be telling us. So, get into your drops more, or do some intervals, or buy a cross bike from your LBS for extra training and give the change to support junior cycling or donate a kids bike to a worthy cause. Do all these things. Think more. Think harder. Ride more. Ride harder.
Freddy Maertens won 13 stages at the Vuelta a Espana in 1977 (AFP photo).
* Rouleur is a very expensive magazine, relative to its competitors, but the interesting editorial choice is that don’t do reviews. The focus is mostly on rides and riders. Whatever you might think of Rouleur’s generally breathless and hagiographic approach, they should be given some credit for their stance on equipment. It’s refreshing.