Sandbagging (human error or Category cheating) still a major problem it seems (Cat C, 5.4 watts/kg?)

[William_Ric-Hansen (William)]

I’ve been off Zwift for 3 weeks due to illness and did a Crit Race today. I was under the impression that there had been changes made to flag people who were in the wrong category. So imagine my disappointment when I see the below results.

Second place with 5.45 watts/kg over 20 minutes. That’s an FTP of 5.18 w/kg. C is up to 3.1 w/kg. Changing that to a 20 minute effort with a 0.2 margain for error gives 3.5 w/kgs which should be the max of a rider in a 20 minute C race.

This means this guy is nearly 2 w/kg out of category and yet he doesn’t get flagged? What is the point of even having a system to remove sandbaggers if someone like this doesn’t get removed?

[Tim_White (Tim)]

This is still a huge problem. Makes cat C and D racing pointless.

[_Troy]

Hi Tim,
this problem I would say will not be sorted for a while but if you change your outlook of the Zwift race you could find it a great way to have a fun hard workout.

Forget the position or the trophy and enjoy the pain. I don’t think I have ever worked so hard outside, as I have indoors, on Zwift, during a race or event.
My favourite rides are the Fondo’s or Tour events, where there are no W/kg Cats to worry about. You can go as hard as you want and there are generally always other Zwifters around you.
As for racing, unless you are in the A, or A+ Cat groups, where there is I believe, a much more stringent verification system of a Zwifters ability, Zwift perhaps should only be viewed, as a fantastically enjoyable/addictive indoor training/fitness platform. This could all change in the future though, fingers crossed.

[Andreas_Traff]

Simple answer:
What is the point in having anti-sandbagging mechanisms? There is no point. Because they don’t solve the underlying problem, which is how the categories are defined. You don’t solve the problem by adding better anti-cheating measures. You solve the problem by reworking the category system into a results-based system instead of a Watt-based system.

So someone doing 5.45 in cat C can’t be right. That’s obvious. But what do you say about this?

This is from a race I participated in the other day. The graphs show the HR distribution over the race which ended in something of a bunched sprint. One of the graphs shows the winner’s effort - the winner as defined by ZwiftPower, not Zwift. The other shows the no 4 guy crossing the finish line a second later, outside the podium, first loser. Now guess which is which?

(Clue: The race was won by someone cruising around in Zone 2 - green energy!)

Before you say “Yeah but… maybe the winner just drafted smartly and maybe 4th place worked hard to bridge to the front group or something”, I can add that the rest of the top finishers look pretty much like 4th. And you can’t normally draft yourself into a Zone 2 win - anyone who ever participated in a Zwift race can attest to that.

And this goes on frequently in race after race in all the lower categories B-D because of the contrived performance ceilings in B-D. You can’t go too hard, so the race favors those who could go way harder but don’t, over those who work hard and push themselves. A is different because it has no performance ceiling. You can’t have a performance ceiling. It creates cheaters.

[Steve_1506 (Steve)]

Andreas:

Your hypothesis has merit. But the contrasting heart rate distributions you provide are less persuasive. We don’t know the true heart rate zones of those two riders, at least not from those two charts from a single race. Maybe the max for the first rider is actually 180, and 210 for the second, and the first was mostly in zone 3 and the second was mostly in zone 2. Whatever the case may be, it’s not a simple as comparing two persons’ heart rates and concluding that A made a greater effort than B did because A’s average heart rate was higher. Heart rates are too individual.

And as a lesser comment on data presentation, when the first chart’s X axis goes from 60 to 210, and the second’s from 60 to 190, an instant visual comparison of the values charted is not possible. The colored sections depicting supposed training zones could provide information, but I don’t know that the training zones depicted in Zwift’s charts are reliable.

[Gerrie_Delport_ODZ (Gerrie)]

Yes it is crazy to how much effort people will go to stay in the D and C cat, from running cycling computers to track w/kg avr to reverse weight doping.

I don’t trust HR data that much. But I bet if you look at their free rides and workouts you will see they ride much harder then.

[Andreas_Traff]

@1506,

You seriously think the respective HR distributions can’t be compared and that you can draw no conclusions from them?

If you would treat each minute of the race as a datapoint (the data being a certain HR), then the difference between the two graphs would be highly statistically significant. There is a clear difference. It’s huge.

The winner is very young, under 20. We already guessed from the suggested theoretical max HR (which he is nowhere near in this race). The other person is a veteran. Given a statistically significant difference in HR’s, i.e. a difference that is highly unlikely to be just random (and very different in the next race), then we need to find an alternative explanation. Something is causing the HR patterns to differ so much.

What could that be? Do you mean to say that the winner perhaps has a heart condition and is only able to produce an abnormally low HR for his age? And that he still wins over several others spending the better part of the race on what would be the functional threshold of the average rider? He’d need an enlarged heart the size of a horse’s to pull that off. And I would be very worried about his health. But somehow I am not worried at all, and next comes why.

Add to the fact that this guy has won 10 out of a total of 22 races in the category (plus 3 other podiums). 10 wins! How many wins did you get over your last 22 races?

Now, that last bit is a tough nut to crack in itself. Is this guy the next Evenepoel, or is there another explanation? If someone would walk up to you in a shady alley and whisper “Hey, teach me how to win about half the cat C races I participate in, at any price! I’ll pay you handsomely!” then what would you tell him if nobody was looking? How could you actually secure a rough 50% win rate in cat C? Any suggestions? I have one. One which I have tried myself just to prove the point. And yes, it works pretty well. Try for yourself and you’ll see.

Yes, this is just one single example. And yes, perhaps we should not draw conclusions too far from that. But would it make a difference to you if we looked at a large number of races and HR graphs like this? Because with large numbers, any peculiarities about people’s HR patterns would even out according to the law of averages, right?

And would it make a difference if it turned out that the pattern persisted, i.e. that winners in cat C did indeed tend to work less hard than the rest of the podium? Maybe not, because we could then come up with rational explanations as to why a winner might naturally be working less hard. I could think of such an explanation.

But if it also turned out that the opposite was true for cat A, where there is no performance ceiling, i.e. that winners in A instead tended to work harder than the rest of the podium? Would that make a difference?

If that wouldn’t be enough arguments for you, then nothing else is going to bite, so I give up. Belief in the merits of the WPK cat system as a religious conviction or something. If, on the other hand, it would make a difference to you, then here you go.

Cruising is real. Cruising works. And Zwift is the only endurance sport on the planet where cruising is a winning strategy, and this is only because of the WPK cat system.

And then you have the sandbaggers doing 5.45 W/kg in cat C on top of that. They are also created by the WPK cat system and don’t exist in any other sport on the planet. Well, they do on occasion, but there their name is something like Eddy Merckx, they get applauded instead of shamed for crushing the field, and then they are moved to the next higher category for winning so much. The cruisers and sandbaggers of Zwift are not going anywhere anytime soon, except to your next race in C.

Last I checked the subscription fee for riders in cat B-D is the same as for a cat A subscriber. But the service offering is not the same, not when it comes to racing. A cat B-D rebate perhaps?

[Twoshihtzu (Ward)]

People are all wired different,
HR could prove the guys cruising
Or
He could be old
Or
He could be on medication that limits his hr
Or
He could have great aerobic capacity from other sports, and only weak cycling ability (cardio isn’t his limiting factor)

Sure there are some who cruise, but there’s plenty of legit folks that rarely go north of 160bpm

[Andreas_Traff]

@Twoshihtzu,

People are all wired different,

Yes. And looking at all the subscribers in B-D, these differences will average out. (That’s what the Coggan HR zones represent, an average.) Still we see the same pattern when looking at the group level over large numbers. Then it’s a tendency. Any tendency has an underlying explanation. And sometimes the simplest and least far fetched explanation is the correct one.

He could be old

No, he is under 20.

Or
He could be on medication that limits his hr

Medications that limit your HR also tend to cap your performance. I couldn’t imagine a true C capped by medications doing so well. So you’re suggesting he is a would-be cat A rider popping massive amounts of beta blockers? He gave it all but simply couldn’t get the HR higher than 176 if he tried? But the beta blocker hypothesis still doesn’t explain why he keeps winning over and over. The cruiser hypothesis provides such an explanation.

Or
He could have great aerobic capacity from other sports, and only weak cycling ability (cardio isn’t his limiting factor)

Do you find him weak, winning half his races? What is he still doing in C? And if you think C is an appropriate category for him, then why not invite the A crowd too? I mean, as long as they stay within limits… (Oh wait, they’ve already signed up.)

Sure there are some who cruise, but there’s plenty of legit folks that rarely go north of 160bpm

And that’s fine. Absolutely fine. Nobody should tell you how hard you’re supposed to race. You listen to your body and you do what is enjoyable. And you should not be punished for that. But neither should you be rewarded for choosing to go easy, while others are about to burst an artery trying to keep up with you. Not in half of your races.

It’s not his fault. The WPK cat system created him. ZP allows him, lets him win even. So why shouldn’t he?

In fact, I would suggest that every A racer signs up for C from now on and try their best to win but stay within limits. Cruising against other cruisers adds an interesting skill element to Zwift racing that I’m sure they would find challenging and fun. Cruising is about getting it juuust right. Not too slow, not too fast. Just a wee bit faster than the other cruisers.

---

I just thought this example was hilarious. Couldn’t stop laughing for some time when I saw it. It was so overly obvious. But then in most cases, such as with a large part of the data in the link I posted above, it is not quite that obvious. And I don’t think that every winner making a smaller effort has to be a scheming cruiser. Far from. But it doesn’t matter, at all. The current cat system ruins the sport just by being there. No wonder no other sport has a categorization such as ours, not even outdoor cycling.

---

You drive your kid to a little league soccer tournament in the weekend. He’s all about soccer. But he will never be the star he dreams of becoming. You can tell. You’re just happy to see him have fun and make friends, watch him grow. Eventually he will grow out of those dreams and dream new dreams. Soccer practice is super important for both of you, just for different reasons.

Coach puts him up as left defender. At one point he manages to make a lucky loooong, almost perfect pass from the goal area over to the team’s striker, who isn’t even offside. He beams a smile over to you. The passing has been a weak spot for him. He is often overly conscious about little mistakes in spite of your best efforts to encourage him not to worry about them.

But then the referee blows the whistle and gives your kid the red card for overperforming. He is warned that he can’t make passes that long in his division. That’s only for the kids in the next division. He is forced to watch the rest of the game from the sideline. The whole team then gets disqualified from the tournament for scoring too many goals in the same game. 7-1. That’s overperformance. The team should have had the common sense to stop at 4-1, the score difference limit in their division according to the rule book.

You mustn’t do too well in a game of soccer as you’re clawing your way up to the Major League. You need to either hold back or simply not be that good. But it’s a great sport, isn’t it? I wouldn’t have it any other way. Would you?

You cannot have performance ceilings in a sport. But you still need categories to involve everyone. So how do you accomodate both? It’s easy. Just do what every other sport does. It’s a tried concept.

[Daniel_Andersson (Daniel)]

Interesting perspective. I’m one of thoose people. I have measured a 179 bpm average on a 21.1k half marathon and a 186 bpm average over a 10k. I can run 1k intervals at up to 200 bpm, but I can’t put out more than just above 160 bpm for an hour on my road bike. My respiratory and cardiovacular systems, built from running, can aquire and transport a hell of a lot more oxygen than my puny legs can make use of yet. My heart and lungs are therefore quite at ease at hard pace. I didn’t even breath hard during my only FTP test so far, I think I maxed out at right above 170 with an average around 165 for the 20 minutes. But my legs were thoroughfully spent.

However, if you “suffer” from untrained cycling legs and a relatively superior cardiovasclar fitness you would surely expect to see a low pulse/ watt ratio, but I seriously doubt that such a person would also be able to dish out 4-5 W/kg for an hour! What could the same person then sustain with legs trained up to the same fitness level as the lungs, an improbable FTP at Grand Tour pro level, 6-7 W/kg? I therefre believe the theory of a fit cardiovasular system and weak legs to be faulty in this context.

Regards
Daniel Andersson
Sweden

[Daniel_Andersson (Daniel)]

Can you look at it from another angle comparing to kids sports? What if your 8 yo kid was playing soccer facing 12 yo rather than 7-9 yo’s, or what if your kids team were 11 yo’s facing 15 yo’s? Would he be happy or would he loose interest? I seriously think the latter would happen. I thought the ceiling was about providing interesting races for most people regrdless of physical capacity on the bike.

[Andreas_Traff]

Kids’ sports is a bad analogy really. But OK. What you describe is exactly what happens in Zwift. 11 yo get beaten by 15 yo again and again and again, because the 15 yo are never moved to a higher category. There is no category enforcement and there is no sense in the system, which allows the 15 yo to join the league of the 11 yo as long as he pretends to play like a 11 yo a little.

As a side note, have you ever raced against real kids in Zwift? I have. This is what happens:

Race Report: 15 SEP 2020, 3R Sundown Race Series, Watopia Figure 8, 1 Lap –...

The whole system is just so unbelievably flawed.

And regarding the poor people with ultra elite cardiovascular systems you talk about, but who ”suffer from” poor legs. Oh how they suffer! I meet them in EVERY race. They defy the research of scientists like Andrew Coggan and Gunnar Borg but they’re everywhere! And they win and they win, if you actually take the time to look at their histories, but those darn legs still never really seem to catch up, do they?

Come on, you’re smarter than that.

All that happened was Zwift divided subscribers into performance bands early on to make it easier to organize and join meaningful group rides. But then a group of morons thought it also served as a race categorization. And here we are.

Name one other sport that divides categories or divisions by performance bands and not results. Just one!

There is a reason you don’t do this in sports. Not even in the Sunday leagues, whose purpose is exactly to make a sport accessible and fun to everyone. And they, as opposed to Zwift, succeed. But you have to base categorization on past results. You can’t base it on how many Watt Johnny, your team mate who lives around the block, can do when he kicks the ball.

A sport that:

1. Has no cat enforcement
2. Has a performance ceiling (you get a DQ if you play too well)
3. Divides players/teams based on past Watt rather than past results

will never work. It just can’t. And that’s why I haven’t participated in one single race this year that didn’t contain cheaters and unbeatable participants with unfair advantages due to system quirks.

Well, ■■■■ happens. But it’s when people stubbornly defend obvious mistakes just because they are afraid of change and progress that you start to lose hope in humanity.

[Milan_Rost (Milan)]

“… you start to lose hope in humanity.”

There are much more important reasons all over the world for losing this hope…

[Andreas_Traff]

Right. So if we can’t fix a silly little broken cat system, then how are we gonna save the world?

No one would ever regret doing these three things to our cat system:

1. Enforce categories
2. Remove the performance ceilings and stop DQ’ing people in races for hitting ceilings
3. Categories based on past results rather than past Watt

No one is ever going to so “oh it was so much better before”. In fact, the people who defend today’s cat system are most likely going to pretend they weren’t, that they were always advocates of a more sensible system. That’s just how people roll.

[Paul_Rayner (Paul)]

Hi Andreas, you make some very good points and I agree with you on the current system not working well at all.

One thing I’d like to challenge is where you said “name one sport which uses performance bands.” Zwift is the only sport I’m aware of which has electronic devices constantly tracking the power output of all participants in every activity, workout and competition they do, so I think Zwift has a much better potential to make something work with performance bands than other sports do.

The only sport I can think of which comes close is road cycling clubs, who often pair themselves into “A”, “B” and other categories for Sunday rides based on the intended pace/speed/power output of a ride. I remember the first time A riders came into a B ride, they did the same they do on Zwift races - pushed too high a pace and blew the pack apart.

A results-based system has problems, such as it requires people to enter races in a particular qualification / league system. This would mean you would always have a constant stream of “new A” riders - or A riders who’ve been off Zwift for a while and lost their positioning - thrashing lower riders as they move up the league.

Another way is that Zwift could use the fact it monitors every one of our rides to make things more fun.

Instead of having rigid categories for all races, Zwift could assign categories automatically based on performance. These wouldn’t have to fit into the rigid A-B-C-D classes - Zwift could create an algorithm which seeds people.
It could rank the 400 people entered into a race based on all their activity data and say “right, you’re the fastest 100 - you’re in group 1, 101st to 200 in grouo 2, 201st to 300th group 3 and 301st to 400 in group 4.” No seeding system will ever be perfect but they could have a go.

This would be a more fun, less rigid way of racing, which wouldn’t suit everyone but would suit those of us who just like to jump into races for fun. And it would mean that sometimes - if the right proportions of riders signed up - you could even have middle-level D class riders being seeded near the front of a race. I’d quite like to have a different seeding in every race!