Rankings based matchmaking (get rid of cats!)

Really strong new zwifters will dominate lower categories with this system, it’d be unfair on lower ability riders.

Did you read the Zwift Insider article? I pay particular attention to new riders to cover this point, which isn’t explained in the post here.

I may also do a follow-up article to cover some additional points. For example, I think the ranking should be weighted to consider more recent races as more critical, it should cover more than 90 days, and if you haven’t completed 5 races, it should only consider the races that you have completed (not default to 600 for the others).

So, just like real life. Then they get upgraded and spend little time in that cat. The entry level will be constant churn of new people. Just like Cat 5

Or just have races that is mass start that any one can enter, so if you don’t have a ranking you will start there.

Or Zwift can look at your FTP and make a guess where you should be.

Many options to get new racers into the system.

Not sure if my 2 cents is of any worth but. I’ve raced for years as an age group racer both in running and triathlons. And in those races there are pro racers and they carry pro cards. Why not use that system. Age groups based on FTP with no cat. rankings. You progress through that tier and then move up to a tier based on age and cat. So there would be the pro class and then a class based on age and cat. I can say this for myself at age 60 I will never be able to compete at the A cat level so why would I want to move up to and A when I would not have a chance at being in the lead group.

I view it as “I’m racing everyone that enters a race,regardless of category”.

“why would I want to move up to and A when I would not have a chance at being in the lead group.”

So are we turning Zwift into a participation medal environment so everyone wins ?

In competitive sport you get to a point where you rise to being mediocre and winning is rare. If you are winning a lot you are edge case, sand bagging or world class.

Age groups pose the same issue. The proposed is one of the best ideas and allows the best racing experience for all.

For sure. This concept seems very sound. The gap is new riders but that is something that can be solved for. Worst case they start at the bottom and have to move up…just like real life. Start in Cat 5 and move up.

I’ll continue to advocate for the recommendations which take into account those who will remain in the lowest category long term. Some of us will be in the lowest category for months, and some of us will stay there for years, if not forever.

So just like every entry level Cat you will compete with those that are new and just starting out. They could do better or worse than you. Not sure how to get around that.

Be more than an advocate, propose a solution. It is easy to poke holes in an idea, a lot harder to actually solve for an issue.

Good solutions were posted earlier in this thread. Solutions better than putting new riders of every ability in the bottom category with riders of lower ability. As in, for instance:

Auto assign is going to have the same issue that we see now with it being off based on zero results or riders not having an FTP entered, not knowing or it being wrong. Auto assign assumes there is sufficient data to place them in a cat, so does not solve for anything for new riders. This solves for people with data.

Most new riders have zero idea their FTP. FTP is also a flawed metric to use as is w/kg. That is why we are where we are with people struggling in the existing system.

Mass start will have the same issue. Existing lower riders will not place. Or are you suggesting you have to enter an open race before you can enter a categorized race? That is removing options to new riders and that is backwards. We want participation. Possibly a A-D + W + E? That is slicing and dicing a lot for a lot of races.

The Open (E) would end up being only new racers and would end up being really small (as would a lot of cats) in a lot of the races. They would not get the experience of being in a bigger group and TBH, would end up under-performing to their potential and then being slotted in a lower group…there by negating the benefits. How many new racers are we really seeing? Are we solving for an issue that really isn’t large?

Obviously auto assign does not work for riders who enter a race as their first Zwift ride. Obviously many Zwifters have a Zwift history before they enter their first race. Obviously all Zwifters have a Zwift history before they enter their second, third, or twentieth race. Auto assign would be a vast improvement over the current laissez faire situation. Zwift knows each Zwifter’s demonstrated capabilities. To implement a rankings based matchmaking system, Zwift does not have to put every first-time racer into the same category with established riders who are in the bottom category because that is the level of their ability.

They have a zwift history that will be them riding around in most cases. So…then dropping them in the bottom category…so not helping your bottom level at all.

Auto assign will work, and is the way to go, but will still not solve your issue. I am fully onboard with match making/auto cat, I just dont see how it is going to solve for your concern. Let’s also be honest, you will have people gaming the system or sand bagging no matter how you cut it.

Are you proposing putting new riders into an intermediate category where they will get spat out the back? That is not an ideal beginning and will lead to them just not racing or just riding around in the race…with low results/power and again ending up in the bottom cat.

No. What I wrote is that one way forward would be to use their recorded Zwift history – which could include workouts, group rides, fondos, as well as just riding around – to seed their entry into a match making system. If their history to date puts them at the 25th percentile of Zwift racers, they start at the 25th percentile of the matchmaking system, for instance – not at the bottom, 0th percentile.

edit: even “just riding around” includes anyone going for personal records on climbs outside of an event. And catching up to the Zwifters ahead. Just riding around definitely can be competitive.

If they have done fondo’s or group rides they will not be new to events. I just don’t think it will solve for your concern. I think it is a valid concern, but I also think a reality is not everyone can win, just like real racing.

We can agree to disagree.

A potential solution to this has also been posted before; 1) Automatically link a Zwiftpower account to each Zwift account and keep early-quitters in the results as DNF. OK this point is off-topic, but its just weird to finish 60/99 entrees on Zwift, 60/82 on Zwift CA results and 51/65 on ZP; 2) people that have a race history are only shown appropriate race cats for sign-up (I.e. a B rider can only select the A, B or E penn); 3) All cats start at the same time, but only your cat is visible and draftable, all others are hidden (this is required for the open-cat to work); 4) The E-Penn open-cat has all cats visible and can draft off all cats. That way this Penn can be used to push yourself and find out your correct cat, but also to guide friends, as a recovery ride or whatever your motivation to ride with lower-cat riders, without affecting the race experience for those riders.

It’d be fine to roll it out as another BETA test series or race-mode option that can be toggled on/off to see if it has potential.

A concern is that Zwifters whose ability is at the 5th, 10th or 15th percentile get matched primarily with other Zwifters with the same range of ability – excluding new Zwifter racers whose power curve, as recorded by Zwift, already shows that they ride at a higher range.

I think the truth is somewhere in the middle… Riding data before someone tries a race will probably be available most of the time, but not always. People who are new to Zwift and really want to win that first race can learn of the rank-based cat system and stroll around a few rides to establish their “D” credentials and then crush the D field in their first race. Perhaps requiring HR data could make it more likely their pre-first-race rides were genuine efforts, but I doubt Zwift wants to enforce that.

So, at worst, some riders (but not all) will come into their first race at a level way below their real ability. However, when they win that first race, their ranking will reflect that and they will be promoted quickly.

Its true that some people will tend to be towards the bottom of their quartile (assuming we stick with 4 cats, but given the number of people in many/most races I think we could easily go to 5 or 6 cats and still have enough riders for an interesting race), but I think the way many people are envisioning this, in practice those people will flip/flop between being the weakest in one race, and the best in the next one. This is because nobody will have a fixed category, races will be set up on-the-fly, splitting the entire group who entered into 4/5/6 groups based on their relative rankings. So in practice its only the “middle of category” riders who will tend to be placed in the same quartile in almost all races they enter.

Having said that, one benefit of expanding the number of splits (ie: generating, say, six races from each set of entrants rather than an “ABCD” four) is that the number of riders in each race is reduced, putting more riders near the boundaries between cats, which means more people will get a chance at a race win, at least some of the time.

I think some of the legacy issues stem from when ZP was not controlled by zwift. In reality, now that they own it, we need to see more full integration.

Fully agree with DNF’s. It is common in all cats. I race A and it is really frustrating to get popped off and then you see people get popped and just leave. I think a min distance needs to be covered to show the didnt just miss the start of had a starting tech issue.

The mixed fields I am hit and miss on and really depends on the event. For most, I 100% agree, hidden is best. Raced last night on Crit city with 450+ people and it was a mess!.

The real concern is that something has to change and I have not seen any progress. I love zwift but the racing/events needs to progress soon or people will migrate. Even the anti-sand bagging is meh in Beta. People just intentionally ride under to the last bit. I had big hopes when they took over the ZP site but they just seem to keep breaking things that worked and I have not seen anything improved.