If you think about sports and games and their rule sets, you can always discern some sort of ethics and ideals in them. 

Take chess for example. There is nothing random about chess, except maybe picking sides. In theory, you could foresee just about everything in a chess game, as permutations or possibilities, and thus you have a fair chance to avoid disasters and to push your opponent into a disadvantageous position. The side who wins has done the calculating ahead better, and this is also the ethics of the game. The player who is the best at calculating and visualizing position deserves to win. If they didn’t deserve to win, then there would be some handicap rule that would kick in once a player becomes too dominant on the board. No, no holds barred. The smartest guy deserves to win, every time.

In football (either kind), the team that scores the most goals wins. They deserve to win because they scored the most goals. It makes them better, more ideal. Scoring more goals than the opponent is going to take both individual and team skill and efforts. The winning team was better at those skills and in those efforts, we say, and so they deserved to win. Sometimes, if it’s not our favorite team winning, we say that the opponents were just lucky or that they played unfairly or we blame it all on the referee. But all else equal, we’d say that the team that scored the most goals wins according to the rules and that they also deserved to win. The rules are aligned with our sense of ethics.

In cycling, real-life cycling, the rider who crosses the finish line first deserves to win, unless he doped or raced unfairly or dangerously. Why? Well… he was the fastest and umm… fast is good, isn’t it? Why was he the fastest? Because he was better? Better in what way? Well, he was either more talented or fitter, or he worked harder than the opponents. So talented and fit is good? Yes, that is good. The ideal rider is talented and fit. What about working hard then? Yes, that is also good. The ideal rider works hard in a race. If he is up against an equally talented and fit opponent, then he will win if he works harder and that is an ideal rider, one willing to dig deeper than the others regardless of the pain and misery it puts him in. And getting fit takes a lot of pain and misery to begin with, so someone did their homework. OK…

Talent is needed at the top level but talent alone won’t usually win a race, because a rider might be up against an equal talent, and then it all comes down to, all else equal, who dug the deepest. And we like that. We admire that. When people go through a lot of suffering without giving up to achieve a goal. A person like that deserves to win, we think.

Now, look at Zwift racing. What is the ethics there. We bring our tarmac values to the OLED screen, but if we consider the rule set of Zwift racing, what it actually implies, we arrive at different ideals, ones that conflict with out tarmac values.

There isn’t one single rule set in Zwift racing. There are at least two broad main versions and then some subsets that the organizer can choose between. We have the Zwift Proper rules and then the ZwiftPower rules, and they are not quite the same. 

The Zwift Proper rules say that the fastest rider deserves to win, period. They do not make any assumptions whatsoever about the rider and why he was the fastest, regardless of category. The Beta Crit City races are a little different and approach the ZwiftPower ethics, but they are experimental and a separate case.

The ZwiftPower rules for cat A is similar to Zwift Proper rules, but the rules set for cat B-D says that the fastest rider deserves to win too, but it does make some assumptions about the rider and stipulate some preconditions. More specifically, they say that the fastest rider up to a certain point deserves to win. You can’t be too fast. Sometimes you also can’t go too hard. Then you don’t deserve to win. You have to nail it just right, depending on your capabilities but in a way also regardless of your capabilities. 

The rider that pinpoints this ideal speed the best, which translates into pinpointing ideal estimates of physical dimensions (matching Watt, weight and height in some ideal combination, and add draft and powerups to that) is also the ideal rider, the one we should admire. 

The ZwiftPower rules do not, however, make any assumptions about subjective effort. If a rider barely wins the sprint in a race after digging horribly deep, but not so deep as to go over cat limits, whereas the runner-up cruised his way through the race, also respecting cat limits, then the winner deserves to win, but not because he worked harder than the other rider. He deserves to win because he optimized certain physical estimates slightly better, thus getting closer to the category ideal than the runner-up. Correspondingly, if the cruiser had won the sprint, then he would be the deserving winner.

So the winner was the smartest and therefore deserved to win, like the winner in chess who calculated the whole game better than the opponent? No, the rules don’t put a value on smarts. This optimizing of physics estimates could have happened for any reason – careful calculation or just dumb luck going flat out – it doesn’t matter. It’s not why you got it just right that makes you admirable and deserving, it’s simply that you did.

This is Zwift racing ethics, like it or not. It’s a quirky sport, isn’t it? 

As a side note, did you for one second think that someone sat at a drawing board and thought it all up just like this, the way it became? 

“Hey, I know, I’m going to design this cool new sport which is about making guesstimates about physical estimates like, you know, when you have to guess how many marbles are in a jar at a market, only you have to put the marbles in there yourself and they’re heavy-like and umm…”

If you, like me, don’t think that Zwift racing, as it turned out to be, was ever truly planned, then ask yourself this: Why exactly do you keep hugging the W/kg system? Why exactly would it be better than a results-based system that promotes doing your best rather than hitting a target W/kg? Is it, perhaps, just because you’re a hopeless conservative conformist in general, one who dislikes change just because change makes you feel uneasy, and who wouldn’t know ‘improvement’ even if it hit you in the face?

Tagged : Cheating Theory

Think about W/kg categories for a second. It’s the bane of fair Zwift racing. We have talked at length about that already. But let’s think in broader terms. What could W/kg possibly be good for at all? How did Zwift come up with these categories to begin with? Let’s do some guess work.

Zwift gives us an understanding of our own personal physiology that surpasses even that of the most expensive sports watches. It’s all these numbers and zones and whatnot. Confusing at first but they tell us how we work on a bike and given some time we start to get it. What we can and cannot do, what we might be able to improve, how to approach certain types of efforts and challenges.

At the center of all this sports science is our functional threshold, arguably the most important of all the numbers. We arrive at our functional threshold power through an FTP test or ramp test in Zwift or sometimes by just going hard yet somewhat consistent in a race.

What do we need the FTP or the W/kg for? It will tell us our max sustainable effort for an hour of work on the bike, which incidentally happens to be a very common time span in the activities on offer in Zwift on a daily basis and for good reason. It will also help us pick a suitable group ride in Zwift. 

Often the organizers of group rides will be quite specific. Such and such a ride will aim for an average of, say, 1.8-2.0 W/kg, given the ride leaders weight of so and so many kg. And then, since you know your FTP and your W/kg, you can quite easily decide if the ride is for you or not. You will have at least a rough idea of how the ride will feel in your body, whether you can cope and whether the ride fits into your training regimen if you have one.

These things were at the core of Zwift early on. This was likely what Zwift had in mind when introducing the categories. Racing was underdeveloped but caught on more and more as time went by because… well, racing is fun! The community wanted it, more than Zwift could foresee. And Zwift provided the means to race but did not meddle too much with how races were organized. They left that to the community. ZwiftPower did the meddling instead.

Have you ever participated in a Zwift race with non-standard race categories? One common example would be masters/veteran races with age interval categories. The organizer uses the A-D categories for convenience but the meaning of those categories is not the standard W/kg one. And the Zwift race mechanics are crude and flexible enough to let you do that. It works just fine. You could organize a race where cat A was male riders on US virtual bikes and cat D female riders on European bikes. Or rider length-based categories. It’s up to your imagination, more or less. At least there are no technical limits to what you can write in your race presentation about what the categories are supposed to mean.

In other words, ZwiftPower could have worked for any type of race category system. They were never tied to W/kg. And, in fact, an embryo of something different can be found on the far right of the ZwiftPower race reports or rider profiles, a kind of rank score that could have been developed further into a results-based categorization.

Moving towards a results-based race category system would have required ZwiftPower to get the clubs and other race organizers onboard. Not necessarily an easy thing. But it would have been possible. And then we wouldn’t have ended up in the W/kg mess we are in right now. 

ZwiftPower is as much a culprit as Zwift, I dare say even more so, when it comes to unintentionally promoting the far most common forms of cheating in Zwift – sandbagging and cruising. Then they set out to chase down the cheaters they have created themselves by building racing on an unsound foundation, through ever more complicated means of catching sandbaggers. Even Zwift have started to help out with that lately. But the cruisers are untouched so far. You can cruise all you want. And I argue you should. 

This is all so backwards if you think about it. The road to hell is paved with good intentions, they say. And Zwift racing went straight to hell, I say, as an addendum to that. You cannot have an influential third party working against reason in a platform of yours. ZwiftPower must go. Well played, Zwift, and I mean it. This is potentially a new beginning. Not a day too soon.

This sounds very harsh, I know. But we just have to kill a few darlings now. The subscribers will benefit in the long run.

Tagged : Cheating Theory

I concluded in the previous post, on cruising in the Virtual Tour de France, that Zwift races are necessarily brutal in their very nature.

That said, the men’s stage 5 up the Ven-Top route, the virtual Mont Ventoux, was indeed brutal. Perhaps a little too brutal… 

I am not saying that there necessarily had to be something fishy about the front trio, but those W/kg numbers they produced were very high. 

The break-way effort that lasted all the way to the finish line for all three of them (althought they got separated amongst themselves towards the end) didn’t last a full hour. Let’s keep that in mind. 

Rather, the attack came 18 min into the race and they got to the finish line in about 45 min. That’s a 27 min breakaway. Roughly. At least it went on for more than 20 min. Let’s keep that in mind too. 

And the pace of the break-away trio never settled to match that of the chasing peloton since the three of them were duking it out amongst themselves all the way to the finish. Let’s also note here that a large part of the break-away was spent pushing pedals at 7+ W/kg. Impressive!

Now over to something completely different. I don’t know if you have thought about this, especially now with the Olympics being postponed, but aren’t the track and fields world records coming in more rarely these days? How come?

Sports science and medicine have an explanation. Top athletes in sports that have you push Watts, whether cycling or running or skiing, and whether sprinting or going long distances, are close or even at the physiological limits of the human body. That’s the official explanation. Hence we are not to expect the 100m dash world record to be beaten easily, and if it is, then it won’t be by much.

I’m sure you have seen a power curve. If you haven’t, then here is one I st0led from the interwebz:

It’s someone’s curve, I hope they don’t mind. They look like this. A little different from person to person, some individual weaknesses and strenghts, but you will always see this downward-sloping curve with a bit of a hockey stick tendency. 

On the Y-axis there is the Watt output of the rider. On the X-axis is a logarithmic time scale. The curve is like a scaling FTP report. Look at any time frame and you can find what is the maximum Watt that rider can produce over that time frame. The Watt for a 1h time frame is what we normally call a person’s FTP. 

In Zwift the 5 min power is also very important, and you can find that too. On the far left is the peak power when sprinting, and it drops off quite fast over time. And if you remember, the way to calculate your FTP is by doing a 20 min max power test in e.g. Zwift and then multiply that number by 0.95 to get the max sustainable 1 hr power. It doesn’t always hold true, though, but at any rate your 1 hr max power will be lower than your 20 min max power by some factor. 

Now, by looking at world records in sports such as track and fields or cycling (mainly track cycling), what sports science says is that you could infer an ideal power curve for mankind. The human body can only move so fast over a 100m dash. Likewise, the human body can only go so fast over the 1 hr track cycling world record attempts. I.e. unless we alter our genetics. Are top athletes at this limit already? Like stated above, sports science claims if they aren’t, then they are at least very close. Which could explain why e.g. Pantani’s record up Alpe d’Huez still stands since 1997 – a both fantastic and terrible year for cycling.

If you want to read more about what science says about cycling on the topic of max performance or perhaps about the physics of cycling in general, the physics Zwift’s computerized model most likely is based on, then I recommend this e-book. You can get it from Amazon or similar. It’s a really interesting read. A wee bit of maths in it – you can’t get away from that in physics – but explained in the simplest possible of terms.

Anyway, sports science claim that the upper human limit for the 1 hr FTP is 6.4 W/kg for men and 5.7 W/kg for women. It seems to check out. It does for track and fields. And no 1 hr world record attempt in cycling has ever crossed that line. Check for yourself! Anything above those numbers and you could fairly become suspicions. It would warrant a closer look. And an explanation of some kind.

All WT pros are genetic freaks. You cannot get to that level with determination alone, or you would run up against someone with the same determination as you but a better genetic disposition and you’d lose. WT pros have both the determination and the genetic underpinnings for performance at that level. But couldn’t it be possible that there are freak-of-freaks too? Guys that stand out even among the best due to some extremely unique genetics, one in a billion? It’s not impossible. It’s just much more likely there is another explanation to the results of such an individual, one such guy that really stands out. Or three.

Let’s return to the break-away in stage 5 in VTdF. I’m not sure what the ideal power curve would have to say about a 27 min effort, but it should at least not be higher than a 20 min effort. Turning the standard calculation to arrive at the 1 hr FTP from a 20 min test on its head, we could say that no rider should push higher Watts than 6.4 / 0.95 = 6.7 W/kg during a 20 min effort. I don’t have data on the average W/kg output for the trio during those 27 min. I sure would like to see it though. Maybe WADA and ZADA should take a look too. Just to be sure.

There are many ways to cheat in Zwift, like we have discussed in an earlier post. But there is also something to be said for Zwift racing. It brings visibility! You can easily hide EPO shots in coke cans in the fridge (in reference to a certain notorious rider in the past), but you cannot easily hide what EPO or blood bags or whatever bring you. Not in Zwift. 

Well, not unless you cruise WT Zwift races…

UPDATE: During the later broadcast of stage 6 it was mentioned that the winner’s W/kg average during the last 30 min of stage 5 was 6.592. 

Tagged : Cheating Theory

With the Tour For All previously as the debut and now with five of six stages completed in the Virtual Tour de France, the World Tour pro riders are beginning to settle in into a world the rest of us are already quite familiar with – the world of Zwift racing. So how are they doing? And are they cruising?

That’s two questions begging answers. Let’s start with the first question. A brief answer would be that the pros seem to be doing fairly well, considering the circumstances. What circumstances? Quite a few things. 

First, many of them are still quite new or even completely new to the platform. Remember your first race? Right. That sort of confusion. The ‘Imma fire off this truck powerup and then I’ll hit ’em like a truck with this early breakaway’ confusion. You know the rest of that story.

Second, we have all met these non-digitalized roadies that have negative opinions on platforms such as Zwift. ‘Zwift miles don’t count!’ That type of guy. They take some convincing and first-hand experience, and that takes time. You can probably find guys like that among the WT pros too. Or should we perhaps say, you could even spot them when watching the broadcasts.

Third, while the Tour For All ran the pros still didn’t know whether there would be any normal races at all this season. At the announcement of VTdF the UCI had already presented the preliminary schedule for the autumn of 2020. Zwift races will likely not make or break a pro contract, but a real-life race might. Obviously, any rider will put priority on the real-life events that will hopefully follow starting in August. Their hearts are into those (and not Zwift races) and it has to be that way since their world and careers are built around them.

Fourth, in preparation for the the real-life events many of the teams take a break from scheduled activities to create a little sponsor publicity by participating in the VTdF. But in the hunt for marginal gains, which is the only thing you can throw at your competitors in the very level playing field of a drug-free WT pro racing scene, would you be willing to risk screwing up the training plan your expert coach designed for you so close to the events you are gunning for, by racing hard in Zwift, now that your team finally managed to get away to some training location in spite of everything this year? Probably not. Not if you won the TdF last year, aim to win in September 2020 too, and happen to be scheduled for participation in VTdF stage 4. You are most likely going to treat the stage as a recovery ride and your directeur sportif is going to tell you that it’s all right, as long as you show up on the screen for the fans.

Fifth, even if you were willing to go flat out for an hour during your high-altitude training camp as a WT pro, you might not want to. Not if you’re a big contender. You may want to hide your cards just a little longer. Or if you are participating and decide to make at least a half-decent effort, you may want to rip off that HR monitor that’s been growing into your skin over the last few years. Why reveal your strengths and weaknesses to your opponents when you don’t have to? It’s all so visible in Zwift, all those well-kept secrets!

So, considering all these circumstances and more, I’d like to think the pros are doing quite alright already. And Zwift is doing quite alright with them. It’s a joy to watch it all. And already at the start of the VTdF there were few if any draft powerups flying off the front, so things are definitely picking up.

So far so good. But next question, are the pros cruising? 

Technically speaking, no. You can’t cruise if you’re racing in cat A+. And you certainly can’t as a WT pro. But it’s an interesting question nevertheless because it reveals things about the nature of Zwift racing. We’ll come to that.

In a way, you could say that Alaphilippe was cruising stage 5, for whatever reason, and I’m sure he had a good one. He didn’t cruise to cheat though. Falling minutes behind the leader is a terrible way to cheat if nothing else. Whether you still felt cheated on as an Alaphilippe fan, I leave to you to decide.

But then there is this other cruising tendency that you see sometimes in the WT Zwift races. It’s the ‘Let’s do this like we do it outdoors in this team sport’ tendency. Look at the women’s stage 3 for example. Parts of it looked like the average Tuesday night cat B race if you look at the W/kg numbers. Surely they could have gone harder, but for whatever reason the peloton settled for a moderate pace mid-stage. Bear in mind that nothing prevented the riders from sitting at threshold for the entire stage since none of the stages lasts more than an hour. It’s still not cheating of course, although we like pain faces on TV.

It was a bit surprising, though, since the female riders showed such aggression and willingness to suffer overall in Tour For All compared to the male riders. Why did they? It’s what happens if you have riders participate who can’t rest on any laurels, riders whose 15 minutes of fame only happen here and now. Like the women in cycling who struggle so hard to get a sliver of the attention the male events get. Or like your average Zwift racer.

If you want more WT pro pain faces on TV, I have what I believe to be a fool-proof recipe. Throw in a bunch of random, lone cat A+ riders in the mix and it will all sort itself out. Not too many, though, because that would only scare the pros away from participating. The pros may have the genetic advantage and more, but they sure aren’t 1-hour-puncheur specialists, any of them. In Zwift races, fear the zwifters.

Note that I am not suggesting that amateurs should have been invited to the VTdF. That is not the point. The point is rather to capture the essence of Zwift racing by contrast. Zwift racing will always be brutal at the front. It’s in the format, it’s racing on the breaking point, it’s in its very nature. And if the racing isn’t brutal, then something funny is going on. And then you may want to take a closer look at it to understand why.

Tagged : Cheating Theory

You have probably already heard about weight doping. That by entering a deflated weight in your Zwift profile you will become faster since it will make your W/kg increase. And there is no one around to catch you red-handed if you do this smartly and only participate in everyday races. Only suckers who suddenly drop a nice even chunk, like 10 kg from one day to the next, will look suspicious. ZwiftPower will react to those and so will fellow racers who might be studying their ZwiftPower profiles. Zwift, however, will not react. You can enter any weight you like at any time. And if you race in the lower categories and make sure to decrease your weight in the profile gradually over time, as if you were on a diet, who would dare to call you a cheater? You were just getting fitter and slimmer!

Weight doping, I would assume, could prove to be a big problem in the higher categories. In cat A you only stand to gain if you drop weight, one way or the other, as long as you don’t get caught stating a false weight before a major event where a weigh-in might be required. 

Speaking of weigh-ins, has it ever occured to you that real-life racers, as opposed to Zwift racers, don’t weigh in? Think about that for a second or two. “It’s just because we all race alone on top of electronic devices trying to simulate outdoor riding”, you might say. But is it really? Meditate on that for the next 30 min.

Now, here’s the thing. Did you know that the above is not the only form of weight doping in Zwift? Cruisers commonly use another form of weight doping, although I suspect that the general lack of knowledge of the cruiser phenomenon makes it far less conspicuous. We can call what cruisers like to do to stay in cat reverse weight doping. I will explain. But first a lesson in demographics.

What is, realistically, the weight of a Zwifter? Or rather, what is the most common weight in Zwift for, let’s say, a male rider (sorry ladies, but cheating is far more common among male Zwifters). This is where you probably google ‘avg adult male weight’ or something like that. And then you find that in North America, and northern Europe might be similar, it is something like a little over 80 kg or 180 pounds or so. So then the most common weight in Zwift would be 80 kg, right? Let’s pause for a bit and consider.

Weight, like most other human characteristics, are normally distributed, a statistician would say. Our weights will fit under a bell-shaped curve, like the badly drawn one I made below.

Imagine you squeeze in all adult males in the Western world under this curve. As you can see, 80 kg is the most ‘roomy’ area under the curve. More guys will fit in there than under other weights, with weights just above or below 80 kg coming pretty close. In the two tails of the curve are the people that either weigh very little or quite a lot. There is far less room for them, meaning there are far fewer of them. So yes, out in the real world something like 80 kg will be the most common adult male weight, with weights around that number coming in close.

But this is not what the Zwift weight distribution curve would look like, especially not in bottom cat D. First of all, in D it might be somewhat skewed to the right (whereas A and even B might be skewed to the left). Since category is defined by W/kg and since weight affects that ratio, it is only natural that there is an over-representation of the heavy weighters. But if we disregard the skewness, the below curve is what you would find in Zwift cat D (maybe a little exaggerated).

“What’s this spike at 100 kg”, you ask. Well, those are a very special type of cruisers, one that deserves a special mention. Don’t get me wrong, some people do actually weigh in at 100 kg and there is nothing wrong with that. But there are only as many as would fit under a smooth bell curve, skewed or not. The spiky bit would be these special cruisers. What’s so special about them? Why, they are just the cruisers who are bad at maths! (I have seen entire cat D time trial teams consisting of guys each weighing exactly 100 kg for a period!)

It is not uncommon for cruisers to inflate their weight. The reason is that these cruisers are way too strong for their category and need something to hold them down a little so they don’t get green coned or upgraded so easily. It could be any weight really. But why then is 100 kg so common? Because it makes it easier for someone who can’t handle a pocket calculator to keep within cat limits. And it makes running a time trial team much easier. To stay in cat D you must not surpass 2.5 W/kg. If you weigh exactly 100 kg (real or not), the highest Watt number you can average is a nice and even 250W. Now, if weights differ within a team participating in a time trial league, imagine if you at 100 kg would push 250W. Then a team-mate at 87 kg would have to go well over cat limits to stay on your wheel. Your team would fall apart long before the season was over since half the team would get upgraded to cat C! And this is why WTRL is running an esoteric categorization with mixed teams from various categories. It would be impossible to run a TTT league with the standard categories, and you would get massive cheating like in early days of TTT.

There you have it. Reverse weight doping. It’s real. And it’s coming to you courtesy of your friendly neighbor cruiser (not from me though, I don’t cheat with weight). And Zwift made it happen and will allow it. Even ZP will, if you play it right. ZP will try to fight it, but really they are just glueing extra wings to an ostrich. It won’t work, because they are trying to save a category system that was completely whacked from the start. It just won’t work.

Tagged : Cheating Theory