Sunk costs and the long distance cyclist

July 24, 2015

20150718_145414Cycling’s most potent mythology is best signified by the death of Tom Simpson: continuation past the point where the returns diminish to zero or less is admired by many even where it is frankly pathological. In order to meet the demands of this myth, it is no surprise that some resort to doping, or make other physically or mentally disastrous decisions. As riders prepare hopefully for Paris Brest Paris, the Race Across America or the Transcontinental (starting tonight in Belgium) they probably hope that they will not have to go beyond the bounds of rationality to finish, but by the very nature of even the shortest of these challenges (no longer a race) demands a suspension of belief even for repeat entrant. The exposure of the body to such sustained repetitive action is unpredictable but always extreme, and even the least imaginative of riders will know that they will have to proceed beyond any normal definitions of plausibility. Such prospective irrationality only gets worse as a rider accumulates time and distance leading to bizarre and sometimes catastrophic failures which could be minimised by stopping hours or days earlier. However, it is almost impossible to judge whether continuing is rational or not: when even starting goes against common sense, how can one decide when to stop? Riders’ accounts demonstrate that successful completion can come despite all the signals to stop, and for supported rides the rider’s team are often better judges, making their decisions based upon more rational bases. Even support teams, however, can suffer from the sunk loss fallacy. If continuing past an obstacle brings failure there is no advantage in continuing, yet riders continue until they fall asleep whilst riding (and crash), ignore injuries that will eventually lead to abandoning, or carry on rising despite the unlikely average speed required to meet a time limit. As the distance increases so does the investment, and the magnitude of the potential loss. Of course, the paradox is that it is incredibly hard to tell where the threshold between a reasonable decision to continue and abandonment lies. Viewed from outside the world of the long distance racer Josh Ibbet’s decision-making on the way to his second place in last year’s Transcontinental looks foolhardy. Judged purely on outcome, however, his decision to ride through pain and exhaustion was successful (if costly). If such decisions were made on irrational grounds, and merely to avoid discarding sunk costs, that hardly matters unless you permanently injure yourself… as long as you make your goals. As an aside, the result of abandoning the investment one has made can be a transfer of that energy into surprising alternate goals. Martin Cox’s extraordinary decisions to transfer his energies from racing to cleaning up the Stelvio and helping out an injured companion of the road are examples of constructive ways of dealing with what might otherwise look like losses.

What does any of this have to do with me? I sank time, money, effort and spirit into my attempt to ride Paris Brest Paris this year. I prepared well, and pre-qualified last year to get an early entry by completing my first Super Randonneur series in 2014. And yet I gave up on my final qualifying ride of 2015, unlike @fabiorandonneur, who endured many challenges and qualified last minute by completing a 400 under extremely painful circumstances. I overslept in the night after a very bad run from Castleford to Mytholmroyd and completely lost my will to continue. Although officially out of time I could have tried to continue with the hope of catching up on the rather flatter final 225 km of the East and West Coasts 600, but after about 10 km of grovelling into a headwind I returned home and slept for about 18 hours on and off. It wasn’t supposed to end this way, but after20150720_113944 tears came resignation and the memory of an enjoyable first section before I collapsed in the night.

The story didn’t end there, however. Unlike Martin Cox I didn’t manage to sublimate my drive into anything selfless. I did, however, complete a similar route last weekend in 37 hours (validated by GPS), an hour quicker than last year, to complete my SR series for 2015. It was alone, unsupported and beautiful, leading to no glory in Paris, but the return on my investments was just right, thank you, including fish and chips at 500 km.

Follow the much more invested riders of the Transcontinental, including Martin Cox on his second attempt, here:

Feeling the pressure

June 11, 2015

I tend to avoid fixed, singular goals. I spread risk to mitigate disappointment. A lot of academics find solace in the short term shifting of attention from one priority to another, and become frustrated when they are managed too directly or become unable to do this due to sheer volume of work. Last year I committed to completing a Super Randonneur series – and when I failed to start my 300 I filled the gap with a late-season DIY by GPS ride. This season has thus far proceeded in a more orderly fashion, pointing threateningly but inexorably at the start (or end) of Paris Brest Paris in August. Although each ride on the way is an achievement in itself there is a tendency to subordinate these successes to the larger and more singular goal. Anything apart from success starts to look like failure.

640px-No-DozUnderstanding cheating on a personal, rather than an academic level, requires immersion in the high stakes of goal oriented behaviour: cheats become blinded to the larger consequences of their actions because their focus on outcome and often seem surprised by the impact of their dishonesty on others. For them, doping is a personal thing: for others it is an attack on the order of things, calling into question the assumed truths of competition. So, am I becoming susceptible to the temptations of assuring myself through the abuse of medication? And would it matter? On my last qualifier I took a few caffeine gels with me and attempted to deploy them in my battle with my prior lack of sleep. In retrospect I should simply have prioritised my sleep before the event but I have a busy job and family life – it’s easier to pack a few gels than manage my sleep. Of course, there are in fact no anti-doping regulations for PBP (EDIT – see John’s comment below, I am mistaken) AND no testing (although doping is illegal in France) but for me caffeine gels crossed a line (over which are caffeine tablets and then almost anything goes…) regardless of any written rules or laws (I have written elsewhere about doping and medication in randonneur events). I should have taken my own advice from last year:

I finished my first 600km Brevet on two cups of sugary tea (the only caffeine I ingested), water, caffeine-free sports drinks, gels and energy bars; I may have had a can of coke, but I’m pretty sure I didn’t. Real food: cheese sandwiches, fish and chips with mushy peas, flavoured milk… and a vegetable samosa. After the ride, I took some ibuprofen to calm the inflammation affecting my shoulders and hands, but not for long, because it makes me feel vile. I am still in pain four days after finishing, and suffering from altered sensation in both hands, although I am now much improved. It may be that my approach is unnecessarily puritan, or not puritan enough (get rid of the space food) but it wasn’t really considered: like the lovely guy riding a Pashley roadster I just did my own thing. You may choose a different path…

More importantly, my recent experience of sleeplessness has led me to question what I am willing to sacrifice. On the Monday morning after my 600 at 0900 I will be serving on a disciplinary panel – I’m not willing to let this affect my ride but that’s going to be a challenge. I start PBP (if I qualify) straight after a torrid week at work but that’s not yet led me to back out. I am proud of my new ability to avoid self-sabotage but there’s some danger of heading the other way and risking my health or safety (or my competence as parent or manager) by being too goal-focused. I will try to employ common sense and prepare to work around adversity, but as I noticed in my last qualifier the lack of stress can be a factor in my performance – I need a fair amount to be optimally aroused and alert and a few disasters on a ride seem to perk me up. With that in mind, my 600 is on the very last weekend for PBP qualification. No second attempts. I could really do with a load of support to ensure I get round, and if I fail, to deal with my disappointment. Fortunately I have that at home but every extra bit helps.

And I may not even like it if I qualify….

Here’s the link to the AUK calendar page for the East and West Coasts BRM600 which I will be riding on the 20th and 21st of June.

Sleepless in Goole

May 29, 2015

The Marie Celeste at Leven: still light but getting late.

I never thought I had a problem with sleep – or should I say its lack. I’ve ridden a fair few overnight rides, and spent about 6 years chronically sleep deprived as a parent. On 400 km audaxes I often feel a little sleepy around dawn, and I managed my first 600 on just one hour. I had planned for about 2-3 hours on that one, and although I became a little anxious when my plans unravelled I felt pretty awake for most of the ride (despite audax fury trying to find a working cashpoint as proof of passage in the night). Read the rest of this entry »

Land of My Fathers (what do you want to go back to that shit hole for boyo…)

April 21, 2015

My first foray into cycle camping was a mixed experience: a week of scorching weather, fine company and challenging terrain followed by the twin insults of biblical weather and a stupid crash. Mix in a heroic rescue and a comic British Rail cameo and it sounds a story worth retelling one day. Caersws plays a minor role, a tiny settlement close to the route of the Plains 300 Audax: site of our dishevelled rail-transported return to London. Having arrived by bike, crossing into Wales via the Forest of Dean and having traversed Wales via Brecon, returning by train felt quite deflating. Not that we could have been much more punctured by fate. Read the rest of this entry »

Fortune favours the brave: Northwest – Passage BRM 200… again

February 24, 2015



I got lost again – but that’s not the story. Instead it’s about being ill 48 hours before the event, but being ready on the day. It’s about favourable winds, lack of ice or rain, the snow stopping five minutes before the start. It’s about checking and replacing worn wheels, and preparing all the equipment and food even before the illness is definitely gone.

It’s not about being lost. It’s not even about finishing. It’s about starting.


Allure Libre

January 29, 2015

Last year’s 600 Brevet back from the organiser (who has been a little busy as President of AUK). Reading the control details reveals that I had completely forgotten a whole section of grim night riding in the Mirfield area. Bonkers.One of the things I like about cycling is its relationship with improvisation. As a younger rider I rode long distances solo with lightweight camping equipment and no real plan. The day was spent riding from one camp site to another, with only wind and terrain governing my path. The two traditions of Audax, one where all ride together led by a captain (Euraudax), the other allure libre (where one might only meet other rider at controls or choose to ride as a group), are different distances from these roots. Although the “free” riding of the latter allows me to embrace uncertainty to an extent, there is a part of my psyche that craves a rejection of the safety in rules that allure libre audax presents. Audax riding is a pragmatic compromise for me, however: it provides a temporally and spatially contained opportunity to ride with or without others which is coherent with family and work constraints.

gibier3000:</p> <p>Paris Brest Paris 1931.

Four recent stimuli have made me reconsider my dedication to the collection of brevets: the first is the change that a clear goal (Paris-Brest-Paris 2015) has brought. This was the stimulus that drove me to finally become a Super Randonneur in 2014, an achievement I have avoided for many years. It is also the impetus that is currently overcoming the inertia I might otherwise have experienced after finishing my final brevet of last year, and drives me towards another SR and PBP. That drive, however has an opposing force: PBP itself exerts an influence on long distance cycling. The Audax Club Parisien designed the SR series as a way of training for PBP, and ensuring that entrants were suitably prepared. That doesn’t mean that there are no other valid motivations for riding an audax: but it does mean that PBP is an implicit as well as explicit goal whether we as riders acknowledge this or not. Although I have yet to experience it, all accounts I have read suggest that PBP is about as far from improvisation (and the x-rated Audax UK rides I have come to love) as one can get whilst having a card stamped. Its roots in racing are still apparent, and there is a huge mismatch between the versions of self-reliance it presents and those of local ACP-validated qualifying rides. That segues neatly into the second stimulus, which has been incubating for about three months: my final brevet of 2014 was a DIY event as I had missed my calendar 300 due to illness. The route, although validated by AUK, was designed by the lone rider, and evidenced purely through two gps files. Although I found my other SR rides enormously satisfying in their different ways, this was a completely different level of self-inflicted challenge, and although driven by the desire to complete my SR, actually reminded me of the role self-direction plays in my love of cycling. The final stimulus was encountering a number of articles about the trip taken by Gus and Lachlan Morton across the Australian Outback (Thereabouts). Although supported by a car and one-person support team/driver, there were aspects of their approach that spoke to me and reminded me of one of the books that haunts my cycling (Journey to the Centre of the Earth): the journeys are very different but there is an attention to people, places and the interaction of inner life and environment that resonated. Lastly, the Transcontinental Race provides a different way of negotiation freedom and rules, resulting in some extraordinary stories (especially Martin Cox’s 2014 ride) where the constraints of the race battle with an authentic sense of ethical purpose (see especially this and this).

So, although I enjoy the balance between security and self-reliance of Allure Libre I also long for the bottom-up joy of “doing it myself”. In 2016 I think I might aim to ride more DIY Audaxes, and maybe think about some more radically individual challenges. But first, there are many kilometres to ride to get to Brest and back, most of which will be ridden on fairly familiar roads.

Beating the cheating: plagiarism and doping

October 14, 2014

Many have argued that anti-doping requires increased funding to improve its impact. This argument, although tempting, may not present value for money, even if it is valid. Although recent empirical studies suggest that the prevalence of doping in sport far exceeds current detection-levels (see e.g., Striegel, Ulrich and Simon, 2010), and an increase in the funding for (and frequency of) testing might increase the hit-rate, I have yet to see a business case to demonstrate the cost of raising the hit-rate to an acceptable level, or even a clear statement on what such a rate might be.

WADA and the NADOs have other tools at their disposal, such as reliance on law-enforcement approaches (which they do not directly fund), and athlete education, but the majority of anti-doping effort (and expenditure) still goes on analytical approaches and creating the legal and scientific basis for a testing and punishment regime. Take for example, as an example of a NADO which has been applauded for its educational programme, UKAD: the budget from 2011-2015 was set at £29m: out of this only around 2m was assigned to educational activity, as opposed to 7m on athlete testing and a further 7m on results management. Similarly poorly off is the scientific activity supported by UKAD, set at 1m; even intelligence efforts are pegged at only 2m. Just under 50% of the UKAD budget (14m) goes on testing and results management (although the latter would also include management of non-analytical positives).

In order to better understand what an increased spend on anti-doping might look like, and why things are not that simple, imagine that all athletes covered by WADA are tested for blood and urine across a range of banned substances every time they compete, and at some agreed level of random out of competition testing. In addition all other efforts ramp up proportionally. Now imagine this has a much lower level of false negatives (i.e. approaching 8x as effective in detecting doping). Leaving aside the enormous cost and practical issues this might bring it is certain that this would not eliminate doping, merely reduce it, and would bring with it an increase in false positives. Would this be an end worth aiming for?

Of course, testing and punishment are not designed to catch everyone, they are intended to be sufficiently sensitive and draconian to offset the potential gains of doping. Anti-doping is intended to scare cheats into compliance, or at least to reduce incidence to such low levels as to be acceptable. I would speculate that current levels of sensitivity are insufficient to achieve this, but it is not clear that greater sensitivity would ever eradicate doping, even with the penalties currently imposed.

It is instructive to turn to another area of applied ethics to better understand these issues. In some universities it is expected that every piece of assessed work by every student be checked for plagiarism, both using automated and human techniques. Estimates of the incidence of plagiarism vary wildly, but Park (2003) argues that around half of a given student population might be expected to cheat in some way during their studies, and plagiarism is the most common form of cheating. There is a large investment in software to help detect plagiarism, and additional and substantial investment in ensuring that plagiarism is not a result of any misunderstanding of good academic practice. Penalties for plagiarism vary between institutions and are often nuanced to reflect the perceived seriousness of an offense: a first offense might receive a warning, but a second offense or aggravated example might to lead immediate expulsion. These measures demonstrably reduce the incidence of plagiarism.

Despite all this, students still plagiarise, sometimes with full conscious awareness of what they are doing and the probable consequences if caught, sometimes out of ignorance. This is probably because whereas for most students the decision to either intentionally cheat or fail to attend to instruction is driven by a fairly rational cost-benefit analysis, there is a minority that can both be identified as fitting a particular type of personality and will probably persist in repeating similar offences even if caught and given a minor penalty, additional academic support and education about the consequence of re-offending (Caruna, Ramaseshan & Ewing, 2000; also see Park, 2003 for a broader consideration of personality considerations). If this is true, then the most intent on cheating simply need to be identified and removed from a social context to which they are unsuited. This does not mean that the efforts to identify the remaining ‘strategic’ cheats are best replaced simply by better pro-active education. The fear of being caught and punished is for this group a significant motivation to comply as long as the detection rate is sufficiently high. Moreover, such detection techniques can themselves be used to educate students about good academic practice.

Even with all this effort, the detection rate does not nearly approach a zero rate of false negatives. This is why many institutions have actually chosen to downplay efforts to directly combat plagiarism and instead work towards education about ethical practices, embedding this within all educational activities. Students sign up to agreements which positively identify work as their own, and more general agreements on good conduct (honour codes, student contracts, partnership agreements). Such approaches seem to be effective in reducing the incidence of academic misconduct of all kinds (Park, 2003) especially where such efforts are the result of a partnership between students and faculty, but there is an implicit assumption that there are two types of offender – one that will respond to pro-active or reactive influences, incentives or penalties; and one that will wilfully continue to offend.

I am becoming persuaded that although detection and deterrence are important, and can play a role alongside education, pro-active positive identification and reinforcement of ethical behaviour is more fundamental. It is for this reason that we teach students about good practices in attribution before we teach them about malpractice, and I hope that this is why I see few cases in the work I assess. Of course, as with doping, it is near impossible to judge how good our efforts are without accurately judging the incidence of academic misconduct, and this is as yet unclear, partly because the incidence of plagiarism is related to an institution’s efforts to combat it.

How does this help us with competitive cycling: I wonder whether personality-screening should play a bigger role in anti-doping. There are ethical issues to overcome, but a voluntary approach such as that proposed by the Clean Protocol might go some way towards thinking outside what seems to be an unwinnable war of post-hoc detection and punishment. Alongside education about ethical values there is an opportunity here to work with the psychological aspects of cheating long before someone first dopes or tests positive.


Get every new post delivered to your Inbox.

Join 1,510 other followers

%d bloggers like this: