Opinion: Results-Based Categorization (Race Categories, Part 3)

Opinion: Results-Based Categorization (Race Categories, Part 3)

In part 1 of this series, I laid out the current Zwift race categorization situation. This included a discussion of the standard FTP-based categories, how enforcing rules post-race falls short, and a look at some actual participation numbers to show how currently-published race results are far from ideal. We ended that post with some ideas for improving the race experience for everyone, including the simple solution of using a rider’s saved FTP to determine their race category. Read part 1 here >

Part 2 followed, where we looked at some ideas for further improving race categorization. This included basing rider categories on more complete rider phenotypes, enabling real-time detection of category violations, and more. The big idea here was an iterative improvement on race categories and enforcement, so when the final phase rolls out, we’ll already have many of the tools in place to make it work. Read part 2 here >

Now let’s look at the final phase: results-based categorization.

Why Results-Based Categories Are Needed

The vast majority of Zwift races currently use the same categorization scheme, which is based on your FTP. Back in part 1 I said: “While some may disagree, I think ZwiftPower’s categories do a good job of breaking riders into groups that can compete well against each other.” And I honestly believe that. The biggest problem with our current w/kg-based categories isn’t the category setup, it’s the sandbaggers who race with impunity below their category, negatively affecting the race experience for legit riders.

But here’s the thing: the sandbagger issue is going to be resolved. That’s what my first two posts were all about, and if Zwift was to implement just a few of the suggestions found there (and it sounds like they are) most of the sandbagging problem will go away.

Once that happens, we’ll have a functional FTP-based categorization system on our hands. And that’s when we’ll start to notice its shortfalls!

Here are a few of the common complaints regarding FTP-based race categories:

  • Favors heavier riders: a 100kg rider’s 3.2w/kg (320 watts) is much faster on flat ground than a 75kg rider’s 3.2w/kg (240 watts).
  • No skill component: a highly-skilled rider could dominate a category, winning every race without ever bumping against the w/kg ceiling. Conversely, an unskilled rider near the middle or bottom of their category’s power limits won’t stand a chance of winning, but they are forced to race in that category anyway.
  • Forever losing: if a rider happens to have power which places them at the bottom of a category (for example, a B with an FTP of 3.3w/kg), they stand very little chance of ever winning a race, yet can never drop to the C category.
  • Hard to Understand: “I just want to race, I don’t know which category I’m in!” The current system presents a barrier to entry because it requires riders to know their FTP w/kg: a number which many (most?) cyclists don’t know.

A results-based system, if properly implemented, would eliminate each of these complaints. It really is the way forward for Zwift racing, which is why I’m sure it will happen eventually. Let’s look at how such a system could work.

Functionality “Must Haves” and “Should Haves”

The biggest challenge in creating a proper results-based system is that it’s complex. No ranking system will be perfect, and while we have good models in real-world cycling and gaming, Zwift presents some unique challenges that aren’t accounted for in those systems.

So the way to begin is to outline what a workable categorization system would accomplish. Here are my proposed requirements – the must-haves which would be required in order for the system to function on a basic level:

  1. Simple Startup: this is Zwift. Newbies should be able to hop into their first race, have a great time, and get a result on their very first try.
  2. No Sandbagging: controls must be in place so sandbaggers don’t spoil the race experience for others.
  3. Immediate Results: as soon as I cross the line, I should be able to see my result, and how it has affected my overall ranking. All disqualifications should be automatic and already in place before the race ends.
  4. Strength of Field Included: points earned or lost must take into account both the rider’s finishing position and the strength of the overall field. It’s not as simple as receiving X points for placing in Xth place. Example: an average B racer who gets 4th place in a race that includes 50 riders, many who are stronger than her, should see more of a ranking improvement from her result than when she wins a race against 5 riders who are all weaker than her.

And here are my additional wishes – the “should haves” which would really make the system great, but aren’t required in order for it to function.

  1. Easy Downgrades: a rider who is coming back from injury or sickness should be able to race in an easier category, provided their involvement doesn’t ruin the experience for others.
  2. Race Up if Desired: riders should be able to race in a more challenging category, if that’s what they want to do.
  3. Flexible Categories: when we have races ranging from 5 riders in a category to several hundred, it makes sense to develop a system that allows for flexibility in terms of the number of categories and their breakpoints. When rider counts are high, race organizers could choose to have more/narrower categories, making the competition even fiercer.
  4. Racer Dashboard: racers need easy access to a few key stats which show where they rank in the Zwift universe and how their latest results have affected their rankings. Ideally this would be available in-game, in Companion, and on zwift.com.

Structuring the Categorization System

Spoiler alert: I’m not going to attempt to lay out the details of a categorization system in this post. Even if I was an expert on categorization systems (I’m not), it’s too complex of an idea to detail in a single blog post. I will, however, reference a few systems and what makes them interesting.

USAC Points

The system used by USA Cycling has obviously been proven in the real world. ZwiftPower uses this system for its ranking points (with a minor modification for small race fields), so this system is actually already in use by Zwifters. Read how it works on ZwiftPower >

ELO Rankings

The ELO Rating System was originally designed for chess players, but is now commonly used in modified forms for ranking players in major sports and video games. This seems to be the most commonly-used ranking system in gaming. Read more about ELO >

TrueSkill Ranking System

Developed by Microsoft Research and used for game matchmaking in Xbox live, TrueSkill has been used to rank players in many different big-name games from Halo to Forza Motorsport. It works well in multiplayer and team scenarios and includes a number quantifying the degree of uncertainty the system has about a gamer’s skill. This would seem to be an important metric in a sport like cycling, where newbies arrive often and one’s ability can change based on fitness level, injury, etc. Read more about TrueSkill >

Cycligent Virtual Rankings

This was the first true results-based race ranking system for Zwifters. Rolled out in January 2017, it ceased operations thanks to GDPR and not long afterward, CVR decided to create its own virtual cycling platform (CVRCade). This system’s structure was very complex, but how it all worked actually made sense if you could wrap your head around it. Read more about it in this ZwiftInsider post.

Dynamic Categories

Things could get pretty wild if Zwift allowed for dynamic categories – that is, categories which aren’t locked in place. Think about it: what if the number of categories in an event could dynamically change based on the number of participants signed up, with the system automatically grouping riders by rank into categories of, say, 40+ riders?

Nobody wants to race against just a few riders, but in an event with 500 signups, why not have (for example) 6 categories instead of 4? This allows riders in each category to more closely compete with each other so more riders feel they’ve “got a chance” and therefore will push harder for the chance of glory.

It could work in the other direction as well: if a race only had 20 participants, maybe it would only have 1 category.

This is all just brainstorming, of course. But I do think there are many interesting possibilities here which could let Zwift racing stand out from outdoor racing in meaningful ways. Perhaps we’ll end up with some races being based on “stable” categories, and others being dynamic. Why not?

When

Based on the recent Minterview, we know Zwift is close to implementing some controls to reduce the negative effects of sandbaggers. That’s great news! Eric Min made it sound like those changes would roll out in the next month or two.

What about results-based categorization? When can we realistically expect that to happen? I can only guess at a timeline, but my guess is that it’s not going to happen any time soon. ZwiftHQ doesn’t seem to have this as a high priority item, so I’d say we’re looking at 6-18 months before such a system rolls out.

Your Thoughts

How important is it for Zwift racing to switch to results-based categories? Got any great ideas for how such a system would work? Share your thoughts below!

< Read Part 2 of this Race Category Enforcement Series

About The Author

Eric Schlange

Eric runs Zwift Insider in his spare time when he isn't on the bike or managing various business interests. He lives in Northern California with his beautiful wife, two kids and dog. Follow on Strava

Subscribe
Notify of
guest
108 Comments
oldest
newest most voted
Inline Feedbacks
View all comments
PATRICK MURPHY
PATRICK MURPHY
1 year ago

Something that is missing from the categories and would make an excellent series is age group racing. I am of the firm belief this would be a super popular series to bring in. (assuming here this does not exist!)

Gerrie Delport
Gerrie Delport (@gerriedelport)
1 year ago
Reply to  PATRICK MURPHY

Patric, look for the ZHR masters races, it is age based and a lot of fun.

Matt
Matt
1 year ago
Reply to  PATRICK MURPHY

Sure agree. But most races don’t have enough to get 4 “power” based categories. You probably need 200+ to have proper competition in age, categories and gender.

Anders Larsson
Anders Larsson
1 year ago
Reply to  PATRICK MURPHY

There is a masters series on ZP that categorizes people in (I think) 5 year brackets. Ride those races but be prepared to have you ass handed to you by some old fart. There are also clydesdale races that categorize you on weight. Ride those and get destroyed by some gorilla much, much stronger than you. The easiest solution to the sandbagging problem is to have everyone start in A and then have z-power sorting everyone out. A bunch of KISS races did that last year and it worked great. Only problem was that the A-category seemed to scare a… Read more »

Matthias
Matthias
1 year ago
Reply to  Anders Larsson

Races for C and B are way different if there are stronger A rides in the mix, to pull some to the finishline….

Derek Redel
Derek Redel (@djr_stl)
1 year ago
Reply to  PATRICK MURPHY

Count me in!

Kevin Worley
Kevin Worley
1 year ago

I love the idea of results based categories. Though a Zwift newbie, I have 35+ years of road, MTB and track racing as well as race organizing experience. When I first started on Zwift and saw how the w/kg categories were set up, I thought it was kind of weird for many of the reasons Eric has discussed. I think a couple of things might help this work. 1.Automatic upgrades – if a racer accumulates enough points based on placings they will be automatically upgraded to the next higher category. 2.Downgrades – a racer may request to be downgraded. The… Read more »

matthias
matthias
1 year ago
Reply to  Kevin Worley

optional downgrading is a perfect idea. Anyone in A or B will think twice before downgrading if he/she has to start all the way from D again. But if they really want to because they didn´t race for a long time they can.

Wouter
Wouter
1 year ago
Reply to  matthias

I think optimally there should be a system that avoids placing strong riders in low categories, because with large numbers, there will always be a few and it takes away the chances of less strong riders to compete for podium places.

Matt
Matt
1 year ago

I wonder if any of this brings new racers to zwift

So many of us have endured years of great racing fun, albeit w flaws / sandbaggers (many whom have been gaming the system for years too)

Does zwift get more or less subscribers when they put more emphasis on racing / structure vs the free for all we have now

matthias
matthias
1 year ago
Reply to  Matt

at least it would be easier for newbies. Start in D and see what happens. For your first race you wouldn´t even have to worry about categories. Zwift would automatically put you in D.

Matt K.
Matt K.
1 year ago
Reply to  Matt

Just one new Zwifter’s opinion, but it would get me to start racing. I’ve been on Zwift since Nov. last year and still haven’t raced. Based on the current w/kg. ranking, I’m a B, but never having raced (not IRL, either) I don’t really want to start in B.

James
James (@jeastwood)
1 year ago

If this isn’t high priority for Zwift they are completely missing the mark with their eSports focus. Guys at the top end have an excellent race experience, but for anyone further down, we can only dream about a fair race. The experience of sticking in the lead group and then launching a sprint in the final 200m, even to come 5th, simply doesn’t happen for cat B and below, because a load of sandbaggers have finished further up the road. The key to eSports success is being able to replicate the experience at lower ranks.

Dave
Dave
1 year ago
Reply to  James

I completely agree with you but I’m pretty sure Zwift has said they want to be the e-racing platform used but don’t want to create the rules. I was reluctant to try racing because I didn’t think I was good enough (C now a B) but I was surprised how much fun it was to be “in the hunt” on any given race. I’ve never been on the podium but it hasn’t mattered. It was fun. That’s the huge challenge Zwift has, not to give every participant a chance to win, but feeling the enjoyment of actually COMPETING not just… Read more »

Esko Lius
Esko Lius (@esko)
1 year ago
Reply to  Dave

Glad to find out that I’m not the only one. I prefer to have those 600 Eric mentioned riding/racing together. Gradually I lose the sight of the backs of the fastest ones, and same happens behind me. If there are enough riders, groups form and every position is worth fighting for. Of course other peeps have other kinds of goals, but for me it’s more about how I’ve been able to develop my 5s / 30s / 5min / 20 min power and by tactical skills during the race. Maybe this approach comes as a necessity with 3.75 W/kg and… Read more »

Sara Lance
Sara Lance
1 year ago
Reply to  James

James, after listening to Eric Min Zwiftcast, I get the impression their goal is to deliver the infrastructure (roads, platform,tools) to race organizers and let them deal with the sandbaggers etc and making sure that they have the right people participating. Its a novel approach and gets them out of the business of being the bad guy . Presumably they will offer some API integration for Race Organizers who would implement some classification model. Just looking at the work that goes into the TrueSkill model, makes me wonder whether economically speaking this is too much of a stretch for their… Read more »

Will Chambers
Will Chambers
1 year ago
Reply to  Sara Lance

It should work fine if Zwift wants the race organizers to enforce race categories. But Zwift has to provide the tools for the organizers, and as far as I know, none exist. Whatever system is used – or perhaps there could be a choice of several – the race organizers need to have the power to ensure that all racers are in the correct category before the race begins.

S B
S B
1 year ago
Reply to  James

I am cat A and my races are constantly ruined by racers who are using ZP instead of a reliable power source. Also, lower category racers joining category A is annoying, because in race you have no clue if you’re battling a real competitor or a C who just pretends to be an A.

Gerrie Delport
Gerrie Delport (@gerriedelport)
1 year ago

I am a big supporter of DYNAMIC CATEGORIES. One thing that I was thinking about was on the fly racing like the battle royale style races. How would that work? You would sign up for a battle royale race lets say you have 3 options short,long and hilly. You would ride around in the game while zwift collect all the entries and group them, once there are 10 riders in your group Zwift you would go the the start pen and off you go. These will just be fun races at a time that suite you.

Aaron S
Aaron S
1 year ago
Reply to  Gerrie Delport

I really like the battle royale idea for casual racing. No planning, no scanning the event list for a suitable race. Just click a Race button, and get grouped with riders of a similar ability. I think groups larger than 10 would be better if there are enough people racing. Zwift could cap the races at X riders or max 5 minute wait or whatever.

El Debarge
El Debarge
1 year ago

Seeing as Zwiftpower is the only reliable source of race results. It seems logical to me that we use their current ranking system to set categories A+ less than 200, A 200-300, B 300-400, C 400-500, D greater than 500 or something similar. Maybe a race organizer could run a test series to see how everything shakes out with the new system and have Zwiftpower enforce the new cats on the back end in the results page. This seems like the path of least resistance and would be the easiest to transition too. Obviously everyone is reluctant to make a… Read more »

matthias
matthias
1 year ago
Reply to  El Debarge

why does everybody think zwiftpower is a “reliable source”?? They only show half the field that raced…. The only result for me for todays races is on zwift.

Wouter
Wouter
1 year ago
Reply to  matthias

Mathias, I ignore riders that aren’t on zwiftpower. But it would be great to have this digital “blood passport” functionality in Zwift as well.

matthias
matthias
1 year ago
Reply to  Wouter

Hmm, I ignore Zwiftpower. Deadlock…. :O/

Chris Cleeland
Chris Cleeland
1 year ago

Another very viable ranking system you missed is something like http://crossresults.com. It’s frequently used for seeding start order in local and regional cyclocross races, and has its basis in the points system used for ranking skiers, and is philosophically similar to chess rankings (you’re ranking based on the strength of your competition–race higher-ranked riders, and your ranking goes up).

Josh
Josh
1 year ago
Reply to  Chris Cleeland

Crossresults and their race predictor are so accurate it is scary. If it says I will get 10th in my race, it is usually correct +/- 1 position, in real life. Could definitely work for zwift. And USAC ” borrowed” their ranking algorithm directly from cross result from what I’ve heard

Henrik Porthin
Henrik Porthin
1 year ago

Performance based categories don’t make sense anyway. If you win or lose is just up to where you put the line. Who can run 100m the fastest but faster than 20 seconds not allowed?🤔
Also you could weight dope by INCREASING your weight if you put out too much power to stay in cat C. Or put out 400 Watts and set weight to 160+ to dominate in cat D.

Categories are only useful for pacing group rides imo

Hybrid Noob
Hybrid Noob
1 year ago
Reply to  Henrik Porthin

Agreed, the current categorisation is totally arbitrary. I understand people want to race against others of similar ability (and please note I am not knocking this desire nor racing in whatever existing cat) but ultimately if you are lucky enough to have an FTP at the top end of a lower cat then you will be in with a good chance of “winning” through this random correlation/setting of the bar. Unfortunately I’m not sure what the alternative answer is to give people a good, competitive race experience. In real racing amateurs are categorized by age (unenforceable in the remote sense)… Read more »

Hybrid Noob
Hybrid Noob
1 year ago
Reply to  Hybrid Noob

Certainly the points based systems you mention are a very good starting point though Eric!

Jaym
Jaym
1 year ago

I was very disappointed to hear Eric Min say that this is not a priority for Zwift, and a demand he thinks comes from a loud minority. I think a very easy measure that would improve things dramatically is just not to let people sign up for events outside their category. Just like men cannot sign up on women’s-only races. Simple, easy, cheap, and a lot less controversial than disqualifications, sanctions, etc.

Wouter
Wouter
1 year ago
Reply to  Jaym

Eric Min forgets that it’s this loud minority that’s equally loudly giving him free mouth-to-mouth publicity. Give me an alternative software platform that focuses more on honest racing and proper categorisation and I will switch in no time.

jcg
jcg
1 year ago

“Here are a few of the common complaints regarding FTP-based race categories:

Favors heavier riders: a 100kg rider’s 3.2w/kg (320 watts) is much faster on flat ground than a 75kg rider’s 3.2w/kg (240 watts).“

Great suggestions. The majority of races are still hopelessly flat, or have a few bumps. Racing is fun, but it’s discouraging as a lighter rider more suited to the climbs. Racing on all the typical flat circuits I’ll need to be a category above my watt/kg to maintain the flat speed of heavier riders. If I’m reading correctly, this is a bigger obstacle than sand baggers.

Steve
Steve
1 year ago
Reply to  jcg

As a heavier rider, I COMPLETELY disagree with this. The advantage heavier riders have on the flats is very small – smaller than IRL. And apparently we also only get half the grade on downhills so a heavier rider’s advantage there is also reduced. And there are very few flat race courses. Almost everything has at least a short climb in it where the lighter riders attack. It doesn’t take much to create a gap which is insurmountable for anyone who can’t stay with the group on the climb. This happens even on the short climb up Whitehall on the… Read more »

Hybrid Noob
Hybrid Noob
1 year ago
Reply to  Steve

Just to correct: ” apparently we also only get half the grade on downhills so a heavier rider’s advantage there is also reduced” If you are referring to the article on ZwiftInsider/Shane Miller video on trainer “difficulty” (I have to put the “” in because it’s such a misleading term), just because smart trainers only simulate half the downhill grade it doesn’t mean your avatar isn’t fully benefiting from that extra weight downhill – YOU ARE, it just doesn’t affect your resistance as much on the trainer. That’s so people can continue to pedal downhill more easily without spinning out,… Read more »

Vince Bossi
Vince Bossi
1 year ago
Reply to  Steve

Heavy riders have a big advantage in a w/kg system. Going uphill on steep climbs everyone who is right at the top of their Cat say 3.2 w/kg for Cat C will perform similar. It will be a tie. Everywhere else heavier guys have the advantage because the pure watts dominate. False flats, flats, downhill. Where do you believe a lighter rider exactly has the advantage?

Reverz
Reverz
1 year ago
Reply to  Eric Schlange

As a heavier top end cat C racer I KNOW that I have an advantage over lighter racers in every race.
I think we’ve had this discussion before and the fact of the matter is. If I ride my 85kg ass up the Alpe at 272W (3.2wkg), you could be 30kg… it won’t matter, as soon as you go over 96W you’ll be over cat limits an I still win. However next week on every other course I’ll crush the 30kg guy like a bulldozer.
The only racers at a disadvantage in the current system are the lightweights.

Wouter
Wouter
1 year ago
Reply to  Reverz

Reverz, wkg goes up more than proportionally for lighter riders. For that reason, they have an advantage on big hills (IRL and on zwift). Go and look at the results of the A cat on hilly races with a good riders field. You wil never find a heavy weight on the top. Even if you are putting out 272 watts, you will just get passed by lighter riders who will have a higher wkg (ceteris paribus factors like VO2max etc). Add to that the fact that WKG matter much more on flats in Zwift, than it does IRL (something I… Read more »

Reverz
Reverz
1 year ago
Reply to  Wouter

Wouter, I was speaking for cat C (and B) riders only as that is what this series of posts is mostly based on.
Of course I understand that once you get to cat A basically all wkg restrictions go out the door and from that point on the less you weigh, the better you’ll preform. However I do stand by my point and almost everyone with a ton of race experience on Zwift will agree with me on this: a 90kg 360W rider will preform way better then the 70kg 280W rider he’s racing against.

Daniel Andrews
Member
Daniel Andrews (@sprenten)
9 months ago
Reply to  Reverz

This isn’t true, heavier riders with bigger anaerobic engines will always be DQ’d in races with CEVAZ limits on short and flat courses if their FTP is close to the WKG limit for their class. On an easy ride I can crush 1200+ watts for a second or two if I want to sprint, in harder rides my 5 and 15s sits between 700-800w but i struggle as CAT D rider in most races over 25 minutes in length. I rode a race today which took me 19:49 to complete in the Kirchmair series, it was my best race, but… Read more »

Stuart Lynne
Trusted Member
Stuart Lynne (@sl)
1 year ago
Reply to  Eric Schlange

Speaking for the middleweight riders I think that we are at a disadvantage compared to the lightweights and heavyweights. We’ll need some special hand-holding to address our concerns as well. 🙂

James Daley
James Daley
1 year ago
Reply to  jcg

W/kg is relevant for climbing, pure wattage is relevant for flats. Obviously, categorisation should reflect this. But, of course, most courses are a mix of the two terrains so with the exception of the Alpe and the likes of Tempus Fugit there is no ideal answer when using one or the other. A more complex system involving the grading of courses would be needed – for example, the Richmond full circuit could be graded, say, 70% flat, 30% climbs, in which case a categorisation using 70% pure wattage and 30% w/kg would be used. Can’t be that difficult to use… Read more »

Ole Kristian
Ole Kristian
1 year ago
Reply to  James Daley

I was thinking the same when I read the article. I have a pretty good score in ZwiftPower mainly because I have a decent sprint and have done good in some shorter flat races. My score as a racer just reflects that. It does not say that I’m a bad climber. I think this is where machine learning could be useful to categorize riders based on performance in previous races. It could take the length of the race, %flat, and %climb, grade and length of climbs, etc into account and find the category a rider should start in for each… Read more »

Wouter
Wouter
1 year ago
Reply to  jcg

As a medium weight rider, I think it’s pretty clear light weights have an unrealistic advantage. IRL they would never stand a chance to win sprints against big power hitters. In Zwift, they are almost on equal footing 🙂

Roark Herron
Roark Herron (@roark-herron)
1 year ago

With respect to categories, where does age fit in? Might it be possible to break down the larger races into age groups as a sub-category. As a 70 year old, I have some physical constraints that a 30 or 40 year old wouldn’t have.

K Denton
K Denton
1 year ago

Why can people just have fun racing .Most riders will never win a race on Zwift or in real life so why is this such a big deal. I have great races with the groups on the screen. hanging on, making attacks trying to see who is weak and who is strong. A group of 10 can be whittled down to 6 then 4 for a sprint to the line. Then look at the results 15th ave wkg 3.2 in C cat. I don’t even bother to look at who won just where riders were in the group I was… Read more »

Kevin Worley
Kevin Worley
1 year ago
Reply to  Eric Schlange

Way more fun. I was always kind of jealous of those riders who “said” they raced for fun only. I do think they were/are a minority of folks who actually raced in organized races.

jack
jack
1 year ago
Reply to  Kevin Worley

Or they have never won anything in the first place to know that euphoric high!

Steve
Steve
1 year ago

I listened to the “Race Categorisation/Sandbagging” section of the podcast. I don’t think Eric understands the extent of the problem. This is not a few newbies with mis-calibrated trainers. This is dozens, or hundreds, of riders who habitually enter races in lower categories than the “rules” state.

The problem was bad last year. It’s ridiculous this year. I won’t enter another Zwift race until some sort of category enforcement is implemented by Zwift.

jcg
jcg
1 year ago
Reply to  Steve

Eric did mention that machine learning could be employed to better create and monitor categories. Presumably, Zwift has all of our data. Creating a power profile for riders, that could supersede single 20 min threshold test should be possible.

Stuart Lynne
Trusted Member
Stuart Lynne (@sl)
1 year ago
Reply to  jcg

Zwift is already monitoring power data to estimate FTP. As we all know occasionally it will pop up and show a dialog box with the new FTP.

Tim
Tim (@iamtimhanson)
1 year ago

Great series Eric. I’ve enjoyed reading it. I listened to the Minterview and like the idea of “ghosting” sandbaggers as a quick win. As for a long term solution, I think Zwift should look at how other online gaming platforms solve a similar problem. I think a system with only four categories is too broad. A game like Rocket League has 6 categories and there are 3 ranks of each. Your ranking adjusts after each match based on the difficulty of your opponents.

Stuart Lynne
Trusted Member
Stuart Lynne (@sl)
1 year ago
Reply to  Eric Schlange

Dynamic categories based on ZwiftPower USAC would be great. Possibly have a minimum and maximum number of categories and participants. E.g. always have a minimum of two categories, otherwise the number of categories is participants divided by 20 for participants up to 100, 30 up to 200 and never have more than six categories.

test
test
1 year ago

To add to the list of common complaints, sprinters have an advantage since w/kg catagorization doesn’t take sprinting power into account. I think Zwift should take inspiration from iRacing, a car racing sim which I think uses a modified form of ELO for catagorization. Everyone is assigned a score (called iRating) which starts at 1600 for new accounts. After every race, you take iRating from people who finished behind you, and everyone who finished ahead of you takes points from you. The number of iRating given/taken depends on the difference in finishing position and the difference in iRating. If you… Read more »

James
James (@jeastwood)
1 year ago
Reply to  test

The majority of IRL bike races are won by sprinters. This can be redressed by not using double draft and giving more chances for breakaways to stick. Also more hill climb finishes. If you are not a sprinter, not a climber, and cant sustain good 5 min power for a breakaway, it’s kinda tough. Enter a longer distance TT or something.

Wookie
Wookie
1 year ago
Reply to  James

Double draft actually helps break-aways stay away. If you’re not a sprinter you’ve got to create a gap, and hope a few other non-sprinters come with you.

bandini
bandini
1 year ago
Reply to  Eric Schlange

I think real draft, I suppose it’s more or less doble-double draft, maybe draft X 5, could do the trick. When you are in front of the group, like irl, you think twice to pull at 5 w/kg knowing the group could be on the wheels at 2 w/kg. Without an appropriate pull in front of the bunch the breakaways could be successful. Like irl.

Bobby Mac
Bobby Mac (@rmacdowell1)
1 year ago

As a Zwift racing addict, I gotta say: Another Great Article! Unfortunately, I think us Zwift racing zealots are of the minority of Zwifters. I read where you did some sample statistics on how many Zwift racers are also registered on Zwift Power — and the percentages are abysmal less than 50% to 60% in some races. But how many of the total Zwifting subscription base actually Zwift Race regularly? And you have to exclude the fondo’s and Zwift Tour events what have categories based on distance. Because the topic of properly managing and policing Zwift Race Categorization applies to… Read more »

Chris S
Chris S
1 year ago
Reply to  Bobby Mac

I think you’re right that only a small percentage of Zwifters care about racing, but I think the main reason is Zwift doesn’t seem to care about it. There’s really no push to get people into it, and it’s incredibly intimidating/confusing for newcomers. My personal experience with it: I started using Zwift in 2018 and only did my first race during this year’s Tour de Zwift. Before that I had never even done a group ride on Zwift. There were a lot of reasons for this. I didn’t want to have to stick with a schedule, I wanted to ride… Read more »

PC Carlin
Trusted Member
PC Carlin (@cabopc)
1 year ago

This would be great. I race as a cat C and often the guys on the podium win most of the races they enter. I’m like dude why don’t you upgrade and give someone else a chance! Now if Zwift would just do something. They clearly don’t seem to care. I only use Zwift for racing but I must be in the minority.

Umsiukas
Umsiukas
1 year ago

I believe Zwift only needs to implement one new category to make everybody happy: The One Where I Can Beat All The Other Guys.

zach
zach
1 year ago

Im not sure what the soulition is. Im a low level A racer and the other day I was pounding it out in the front of the pack with a 17 year old kid, two 50 year old men and a woman. I thought to myself how cool it is to be in a sport where nothing matters except ability. I want to continue to race diverse fields of gender and age.

Jack
Jack
1 year ago

Incredibly well thought out. I find the current structure to be difficult for a non-racer to understand where to place myself. The other part to this is I am a big guy and I can crank out some watts this all regulates well I think from a k/w standpoint, the thing it doesn’t take into account is that I have absolutely horrible form (I don’t stay in an aero position) I am all over the place when I am pushing 100+rpm. All you see is that I am pushing out the watts. Not an easy solve but I must say… Read more »

Patrick
Patrick
1 year ago

W/KG works… add in points for results and force the upgrade. Most races are won by riders at the groups upper limit… no dropping into lower group until points expire… if you are a B then race B, injured or returning from break should not matter if you are carrying points…
Seems unlikely that Zwift will police anything to seriously as users pay a decent fee to be here…

Jeff Bloom
Jeff Bloom
1 year ago

I am not sure how well Eric Min’s proposal for eliminating sandbagging will work. The idea is to detect riders who are sandbagging during the race, and then ghost them, at which point they will be invisible, you can’t draft them, and they are not counted in the race. The idea of ghosting if fine, but these riders will still influence the race up until the point at which they are ghosted. A better solution would be not allow a higher category rider to enter a lower category race. In addition, you could allow a higher category rider choose to… Read more »

Ole Kristian
Ole Kristian
1 year ago
Reply to  Jeff Bloom

Yes, It is not obvious how Zwift is going to ghost riders. I’m thinking of races like in Crit City where many races only last for about 20 minutes. How can a rider be ghosted if a 20 minutes W/Kg can’t be calculated before the race is over?
The idea of giving riders an option to race down as a ghost is good. I have read that some race down if they have a bad day and still want to race. With your suggestion, this could be done without affecting the race.

Stuart Lynne
Trusted Member
Stuart Lynne (@sl)
1 year ago
Reply to  Ole Kristian

Properly done Zwift would need to establish maximum power numbers for shorter periods of time for each category. E.g. 1, 2, 5, 10 and 20 minutes. Then maintain a rolling average for observed power for each of those durations. Exceed any durations maximum power and you get ghosted. There would need to be different sets of numbers for male versus female.

Ole Kristian
Ole Kristian
1 year ago
Reply to  Stuart Lynne

I’m afraid that would make the races boring.
Let’s say that I have an over average 5 min power that I try to use to go in a breakaway. The rest of my powercurve is on the average or low. 5 minutes into the breakaway I get ghosted for having a too high 5 min power.
If I want to split the field on a climb, l have to monitor my power so I don’t get ghosted etc.
It might be that I’m overlooking something obvious here, but I an afraid we will get a lot of false positives.

Aaron S
Aaron S
1 year ago
Reply to  Ole Kristian

If you are getting a lot of “false positives” in this instance, you would be racing in the wrong category. I don’t think the rider themselves should be aware that they have been ghosted. Instead at the end of the race, instead of the results popping up, they should get a huge CONGRATULATIONS! You have been upgraded!

Ole Kristian
Ole Kristian
1 year ago
Reply to  Aaron S

Aaron S: It can’t be that you are in the wrong category only because you have a strong 5 min power or a 1 min power if the rest of your power curve is normal or low for the category? We can’t expect every rider in a category to all have a power curve within the same boundaries. Some are climbers, others sprinters. They will have a very different power curve.

Aaron S
Aaron S
1 year ago
Reply to  Ole Kristian

I think this would make the racing interesting, because you would have riders of comparable ability, but different strengths, racing against each other. So the FTP riders could hammer away to try to drop the sprinters, who would by definition have a lower FTP, and riders with a relatively strong 5 min power could launch a late attack to drop the FTP riders, etc.

Dave
Dave
1 year ago

Eric, what also could be a cool feature in the future is “predictive seeding” the participants of a race, either in the pen or once it starts, based on recent race results, power phenotypes, course, and most importantly who you’re competing against. That is, Zwift calculates what rank you SHOULD finish. Obviously just for fun but could be another marker to establish did I outperform or underperform in this race?

Stuart Lynne
Trusted Member
Stuart Lynne (@sl)
1 year ago
Reply to  Dave

ZwiftPowers USAC ratings are a predictive statistic. All things being equal two riders should finish in the same order as their ZwiftPower race rank.

Of course, the rankings don’t take into account all of the factors in a race, e.g. distance, climbing, flat, etc, etc. So comparing the ranks of two riders that may specialize in different types of races may be difficult. It would be possible to maintain different ranks for different types of races (Cycligent’s ranking did that.)

Wouter
Wouter
1 year ago
Reply to  Stuart Lynne

predictive seeding based on initial performance could still be useful if a rider hasn’t yet finished sufficient races to have a meaningful USAC ranking!

Alistair Flack
Alistair Flack
1 year ago

Or…get rid of all the categories DURING the race and then show comparative stats afterwards (i.e. w/kg as a % of your PB over that time, age ranked, split by gender) I know the drafting dynamic changes things, but my comparator us running races where it’s a simple time over the line from when the gun goes. No need for categories during, but post race you can see how you compare to all other runners, plus filter by age/gender. Run Britain ranking website applies a handicap system so you can see how your performances compare by relative effort. It’s the… Read more »

Reverz
Reverz
1 year ago

I am of firm belief additional category’s would be a quick an easy fix for most of the problem. Eric admited it several times in the comments… racing might be fun but winning (or at least having a good shot) is even more fun. I mean why do we race? For the competition… but is there really any fun in being spat out after 4 minutes en then riding a 1 hour time trail? In my opinion it would help a lot if we had a B+/B- and C+/C- cat, cutting both cats in half. All those 3.2-3.4 riders now… Read more »

Daniel Andrews
Member
Daniel Andrews (@sprenten)
9 months ago
Reply to  Reverz

I could go for a C- classification, you will have a hard time convincing me I am anything but a solid Cat D rider with above average anaerobic power for Cat D and just about average for Cat C. For short flat courses of 15km or less I can hang with 2.8 to 3.1 riders and still have a chance to beat most of them in a sprint. However, I still suffer on longer flat courses like Tick Tock. My wheelhouse is that 12km or less range and I would have no trouble locking into a category with a top… Read more »

Mark Larson
Super Member
Mark Larson (@marklarson44)
1 year ago

For Tour De Zwift events we were getting “D” fields in the 120-150 range. So perhaps there should be 6-8 categories instead of the 4ish they have? Probably less easy in the random weekday races. I’ve never been in the top 10 of a D race but I’ve podiumed and even won 1 race according to Zwift Power. People see D and equate to Cat 4 and think “well I’m a Cat 3 but this is just a training race so D should be fine”. And that’s just the US way of looking at it, the systems might be different… Read more »

Reverz
Reverz
1 year ago
Reply to  Mark Larson

You really should think about installing the Companion app and registering for races (or any event) from there… it eliminates having to click the very small cat buttons in game. It’s also something you can do a few hours or days even before the event starts and is one less thing to do just before the race start because if you’re anything like me you’ll probably the rushing around the house like a headless chicken 2 minutes before race start 🙂 PS: “training races” are the worst excuses ever. IF you race… you race hard, at the right level or… Read more »

Jean Jeanie
Jean Jeanie
1 year ago

Not too fussed myself. If riders have more all-out power and are faster on the flat, so be it. I just have to use skill as in any race – draft, and try and open a gap on the climbs. Maybe I need to get more competitive! But one category, or at least comparison, you don’t mention that is useful and available real-world, is age group categorisation. Would think it’s easy to enable, like all-Zwift rankings. Like Zwift Insider, Eric. Thanks mate.

Wouter
Wouter
1 year ago

What would make sense is to have dynamic USAC point thresholds that determine, say at least 5 categories, who divide the field of current racers in 5 equally populated categories (hence at each 20th percentiles of ranking distribution), taking some initial USAC ranking / point distribution as given. A race ‘season’ then would consist of 1 pre-defined period (2 months?) where the riders within each category race among each others, starting with their initial USAC ranking. At the end of the period, the 10% riders at the top of their within-category USAC point distribution get promoted, the bottom 10% riders… Read more »

Iangus
Iangus
1 year ago

I started reading this article thinking that the suggestions made sense and in some part they do, but what is racing if not to see who is the fastest? So you sort out the sandbagging and then you start on things like weight and skill, that’s part of racing, I weigh what I weigh so I choose the races accordingly, Peter Sagan doesn’ t race for GC because of his weight, he chooses his races it’s real life, as for skill well that’s what we all strive for, learning from my mistakes, timing the sprint better or going in the… Read more »

Adrian Amos
Active Member
Adrian Amos (@ahamos)
1 year ago

I would NOT be in support of dynamic categories, unless that were a selectable option for race organizers. It would be nearly impossible to race for points in a series if suddenly you found yourself bumped to an overflow/combined field. That almost happened to me in our state championship road race last summer, when only 7 riders showed up for cat 3. All 6 other guys were willing to jump into the 1/2/3 field, but I was competing for the year-long championship, and needed the points from an individual cat 3 race to stand any chance of winning the overall.… Read more »

Todd
Todd (@bombertodd)
1 year ago

Dynamic catergories sound great. I often wonder if a fortnite style would work. Join a race, go warm up while others join, when a certain number of riders are ready you get 3 or so minutes before the start. You could slice that number into 4 to 6 groups. Rank all the riders and place them in groups. The upper limit if each category will change. Sometimes you would be most likely to win and other times it will be difficult.

Jason
Jason
1 year ago

I wonder if Zwift are missing the mark with average Joe/Jane cyclists who would like to race but are intimidated. I won’t go as far to say they’re missing subscribers who opt for another platform because they’re noob cyclists or pure recreationalists, but that’s also a possibility. Plenty of fun group rides, training and just cruising to be had…but the more ‘racier’ it gets the more some may feel isolated. As a solid D who only has the faintest glimmer of C rides, I find Zwift ‘can’ appear more geared towards hard core cyclists (count me in as one, but… Read more »

Dustin Mealor
Member
Dustin Mealor (@dartingd)
1 year ago

So I play league of legends. I think their elo system is great and would work really well. Basically you start as a Bronze. There’s bronze 5-1 then you can move to silver. Same structure goes through bronze, silver, gold, platinum, diamond, master, and challenger. Zwift would of course have its own name system. So each win gets you an amount of points. For zwift it would be based on total number of riders in the race as well as finish position. You go from zero to 100 points and you move up from say bronze 5 to 4. When… Read more »

Marco
Marco
1 year ago

Sorry @Eric Schlange I have read all the parts and everything is correct and insteresting but it seems to me that it is getting out of hand, maybe we did not realize that the division by categories could make sense in a perfect world but certainly not in a world of remote, uncontrolled, sly and thirsty for wins users like on Zwift. I’ve seen it all since the very first beta of Zwift. There are too many variables such as the declared weight, the used trainer, various technological stratagems, etc. etc. At this point the solution seems to look back… Read more »

Ole Kristian
Ole Kristian
1 year ago
Reply to  Marco

Let’s just lie down and die! This was a very passive approach, I think. If we want to find solutions for the future, then we must dare to think a little new and rather take a step back to find solutions that propel us in the right direction.

Marco
Marco
1 year ago
Reply to  Ole Kristian

What you’re saying is not correct, of course we could stay here making great plans for the future and it would also be very funny but honestly I prefer to be frank and realistic so I’ve mentioned my vision.
But if you prefer to stay here talking about how Zwift WAS – IS and WILL BE I’ve a lot of stories to tell…. 😉

Ole Kristian
Ole Kristian
1 year ago
Reply to  Marco

I think it is possible to both make great plans for the future and at the same time be realistic.
What does age have to do with it? If you are 20 years old or 50 years old it does not matter if you are on the same level. Isn’t the most important thing that you can compete against riders at your level so you get tight fights and exiting races?

Marco
Marco
1 year ago

The whole idea of categories based on w/kg needs to be scrapped. It makes absolutely no sense.

Ole Kristian
Ole Kristian
1 year ago
Reply to  Marco

The title of this article: “RESULTS-BASED CATEGORIZATION”

Marco (#33453)
Marco (#33453)
1 year ago
Reply to  Ole Kristian

ops, it seems that someone here used my name….

Lee Redfern
Lee Redfern
1 year ago

Results based categorization is the only way to move forward…I do agree that limiting Zwift to only 4 categories is a bad idea. At least 6 categories are needed to get racers exactly where they need to be racing, skill, power and endurance wise. At 55 Kg and classified as a “B” based on w/kg for me to keep up in the B races, I’m putting out huge efforts, usually getting dropped, never making to the line with the group. Yet in “C” races, I’m putting out just as hard an effort to make it to the line with the… Read more »

Gerrie Delport
Gerrie Delport (@gerriedelport)
1 year ago
Reply to  Lee Redfern

The only reason you have to put out the same power in C is because the front of the C group are all B category riders, if they all move up to B then you will have a more enjoyable ride in the bottom of B, and the top of the B will move to A.

Henry Ashman
Trusted Member
Henry Ashman (@h_j_ashman)
1 year ago

I’d certainly like the idea of results based, however I’d like there to be some sort of “newbie pathway” so that riders can go in at a semi-suitable level, so the bottom rung of competition isn’t full of Cat A riders starting out. Also opens up some options for seasons with promotion/relegation & rank resets, and some sort of competition-only rewards, thinking mostly jersey’s, shades, maybe special paint jobs as race only bikes might be a bit mean to those not interested in racing. Of course I’d be intrigued to see whether you’d end up with the “smurfing” problem some… Read more »

Markus
Markus
1 year ago

Personally I think that Zwift should adopt a controlled categorization system like Zwiftpower. It would be a big step if Zwift just adopts the concept of Zwiftpower. In addition, Zwift can still leave racing alternatives for people who don’t want to follow the rules of the categorization system. This would be the place for all the sandbaggers. I really think it is a shame that Zwift is considering ERacing as an olimpic sport but they don’t do nothing about implementing a fair categorization system.

Galen Kehler
Galen Kehler
1 year ago

I read though all these and it answers the “what” and “how” of categories pretty well. Eric can you do a part 4 where you answer the “why” of category racin in zwift? Thanks

Raoul
Raoul
1 year ago

Interesting article, thanks ! I did my first race today ( 3R Alpe Du Zwift KOM Race) and I’m pretty sure I signed up in the B category but during the race I noticed pretty early in the race that I was in A. After searching high and low I couldn’t find any info about this so I’m thinking I maybe made a mistake when I signed up ? If Zwift do automatic upgrades what is it based on ? Thanks for any help ! In Zwiftpower I am indeed a B (296watts / 3.9wkg) but I got ranked in… Read more »

S B
S B
1 year ago

Excellent article! Zwift must make this their number one priority. Too many cheaters these days on Zwift and it’s really ruining the races. I’m about to quit Zwift, because of this.

M J
M J
1 year ago

Until and unless Zwift goes to a results (points) based system it will be an eSports joke. Or perhaps just a joke. For all the reasons you mentioned above, categories based on WKG are absurd. Also absurd is the idea that you DQ someone after the event for being over WKG for power by over 0.1 WKG. If they entered the correct category to begin with they should get a result. If they entered too low a category why should Zwift allow that? I remember once, I went to an IRL crit and the winner was DQ’d for producing too… Read more »

Hennie Du Toit
Hennie Du Toit
11 months ago

Nice reading, relevant and all relative in the end.
There are many different factors, w/kg probably the biggest contributor, however I would like to see age playing a role in the classification of riders?
Surely this could make racing even more competitive and fair?

108
0
Would love your thoughts, please comment.x
()
x