Some pros show up on Zwift to get a paycheck but do not actually race, which is why their scores tend to be low. One standout racer is Alex Dowsett, who has a score of 570, matching mine from when the ZRS system first started. He last raced in May, so he is past the 90-day window, but he recently did a ramp test that gave him an FTP estimate of 345W (4.6 W/kg). If ZRS relied on AI to assess his abilities, that ramp test itself would be enough to bump him up to a more competitive group.
Organizer has to ask for custom Ranges…we did a week earlier 
Now if they could just fix the damn 30s issue with Seed…and stop multiply with CP600 cause if the true effort is missing the formula is failing.
1 Like
Yeah I know, my tongue was firmly in my cheek when I shared that.
The formula is always going to fail. To many people are already considering gaming the system so they can remain in the lower categories.
Actual Race Score movement needs to be the significant influence on the score and actually move people to the right place.
Currently, if someone sandbags their seed formula, the system cant move them to their correct place based on results only. Basing this on results is the only way to ensure people end up in their correct category.
if people want to crash their score, they wont be winning the races they are using to crash their score and there should be some mitigation against huge drops in score from poor performance.
Edit to add - Nothing has been done to improve the culture of racing, the ethos of always trying to punch down has not been challenged. Zwift need to reward users for improving not give them incentive to punch down.
2 Likes
Yes if people would be racing multiple times a week…the problems are those that don’t and the decay is bigger than the score movement.
Having A’s at 400 Range is screaming for failure.
1 Like
Try looking at this ‘problem’ from my point of view.
I’m a lightweight rider ‘blessed’ with a low 30s power. I wish I wasn’t ‘blessed’ in this way. I’ve improved my sprint, but it’s still a pile of crap. However, my light weight has seen me spend most of the last 4 years getting beaten up in the bottom quarter of B cat races.
There are many C cat racers, including a regular poster here, whose FTP and sprint figures are many watts higher than mine, but they are blessed with being heavier than me. Their raw watts mean they are comfortable in the pack on flat and rolling courses - most races in Zwift - and able to comfortably outsprint me.
These guys aren’t doing anything wrong. That’s the way the CE system has operated. As a former ZRL Team Manager, I lost about 5 riders to B act upgrades - including myself - who had never even come close to winning a C cat race. One finished 29th in a field of 76. I got upgraded after finishing a race in 15th place.
My experience of racing with ZRS, and zwiftracing.app, has been entirely positive. I haven’t got close to winning any races, but I now find I’m competitive to much later in races. I’ll still lose out to better sprinters, but I’m finding I’m much more closely matched with other riders in my pen.
This doesn’t mean there are no outliers, or that the seeding formula can’t be improved, but what I see coming from Zwift says to me that intend to keep looking at these issues to improve the ZRS system.
I do agree with your statement that this could lead to more specialisation by riders. Maybe that could mean race series that have to look at a wider mix of course types. Not every rider can be competitive on every course type IRL. Cav was never going to win a yellow jersey. Froome wasn’t going to win the green jersey. So why do people of Zwift expect they’re going ot be at the front of the pack regardless of course type?
As long as I get a good workout, enjoy the competition and get the opportunity to stretch and improve myself, I’m ok. Last week’s Zwift World Series race on the Peak Performance course was a perfect example. I got over Titan’s Grove with the lead group and along the coastal route to the bottom of the Grade, which split the lead group into pieces. I was dropped fairly early on, but then had a great battle with a smaller group that formed on the climb. The racing between was was good all the way over the KoM to the finish line.
I finished 16 from 40, but it was so much more enjoyable than riding this course as a B cat event, where I’d have been dumped out the back before we got over the Titan’s Grove KoM. Groups of riders formed up and there were time gaps between groups, but again this is closer to what you’d expect on a course like this in real life.
My ZRS score increased and my zwiftracing.app score went up significantly due to a zrCS change. That only happened because the matching of riders and the quality of the race saw me dig far deeper than being forced to do a solo TT in an equivalent B cat race.
My experience so far says to me that ZRS will improve racing on Zwift. There are issues to resolve, but directionally this feels like the right way to go. The principle of trying to initially seed riders and then let results dictate rider’s scores has to be the right way to go.
17 Likes
Whatever system is used and the parameters to work out ranking is shared there will always be racers who will game the system that’s why I said previously I don’t see any benefit of broadcasting it.
The several ZRS events I have raced in I have been much more competitive than CE A pen 4-6 w/kg.
As a light rider the explanations above probably are the reason my 30 second power would be pathetic 
but my 10 min would be decent.
No system will ever be perfect I’m not praising ZRS because for myself it is better than CE what is required is constant analysis / tweaking of the system to improve it.
Once ZRS is live on the 7th ZHQ will have vastly more data to analyse and there will no longer be a choice of different racing systems IMO as it should be.
1 Like
Nice post. It’s good to get a bit of balance. 
2 Likes
Then you know what it feels like when the category system is flawed.
Giving weight to 30 sec power does the same to sprinters as W/kg does with light weight riders under CE.
You can’t really compare ZRApp and ZRS. ZRApp takes into consideration your phenotype and adjust your points accuracy. The race type fits you - you get less gains when doing well, and visa versa.
The main part that concerns me is that the issues with CE, and indeed ZP categories before that, were well understood years ago yet we saw the barest effort from zwift to fix any of them.
While zwift have said they will continue to work on ZRS going forward, we’ve heard the same rhetoric in the past and it has meant nothing.
At least with ZRS the scores should improve over time without intervention, though it has already been established and demonstrated that ZRS can be manipulated should an individual wish to do so.
i sincerely appreciate you taking the time to share this viewpoint, thank you
1 Like
Ok, so I did my second ZRS race today. I am coming from a ZRS that does not have a 30s or a 10min true max effort, it contains power from threshold workouts, interval workouts between 3-5 mins, pace partners and a couple of races.
In my first race my ZRS went up +4 from my seed at 358 to 362 even though I beat a lot of 400-500 riders and nobody above me had a lower ZRS.
Before this second race my score degraded to 360, and the following placement bumped me +8 to 368:
As you can see, the average of the people behind me and ahead of me is well ahead of my new ZRS score of 368.
The average ZRS of the 9 folks behind me (the most I can show in this screenshot) is 473.
This performance at 3.8w/kg for 20mins bumped me to CE B cat, but really didn’t move the needle on my racing score.
My guess is it will again degrade 2 or 3 points before I race again.
My belief is the ZRS I should be at is probably mid 400s or so, but I wont get there with one race a week this season from what I can see.
Edit: For the record it really doesn’t matter either way for me personally because I’m not gonna drop a category and I shouldn’t be in the next category, so I’m probably just going to sit in this category middle of the pack for a long time until I gain or lose significant fitness.
Edit 2: For the second record this bumped my zFTP to 280W at 77.8kg.
4 Likes
My zFTP is 276w at about 77kg and my ZRS is 608 
Still haven’t bothered with a ZRS race.
1 Like
Interesting to note the 7th and 9th place finishers were boosted to scores that promote themselves out of this cat, though not apparent in your screenshot that they performed better than other riders who possessed higher scores than they had. Isn’t that partly how its been said ZRS is supposed to work?
1 Like
Oh crap, maybe it is possible for me to get into the next ZRS cat, ha!
I don’t have a sprint at all, so it’s unlikely I’ll get any podium bonuses, and my seed will pretty much never be bumped up due to any of my in-race sprints since they are so weak, and I’m not planning on doing any max 30s efforts outside of races (not because I’m worried about ZRS going up, just not what I normally do) - I tend to keep my intervals above 2mins in length normally during workouts, so only place I naturally will try to sprint is end of races when I have no sprint left.
That’s unfortunate for the 9th place rider because they were dropped in an earlier lap of this crit race - they didn’t finish with the front group, so they are not going to have fun in the next cat up.
Edit: Maybe they just got dropped on the last climb of the last lap (they were 19s back and not in the pack of folks that sprinted the last 400m or so). Either way they were not at all in contention for a podium at the end of the race given they were 19s back and not there for the final sprint.
Edit 2: For what it’s worth, the Zwift Racing App predicted the 9th place rider (in ZRA it was actually 8th due to an HRM missing in 3rd place) to end up exactly where they did (predicted 8th, result 8th), and ZRA did not move their vELO score up at all (actually down -2.6). While ZRS now moved them up into an entirely new category.
Edit 3: Also I got a “Major vELO rating upgrade” achievement on ZRA for this race (they predicted me 10th, when I ended 5th) when ZRS moved +8 points. I still think ZRA’s logic is working better than ZRS.
1 Like
Now let’s not confuse 600+ with “big boys”! I’m 627, and know of plenty other not-at-the-top-of-B riders in the same boat. We’ll get absolutely annihilated by the A+ guys!
I do like the idea of dynamic pens. The main issue I foresee is GC for things like the monthly zRacing. Not ever knowing what pen you’ll start in would make it awfully hard to compete for the GC (not that I ever worry about it since I can rarely make the “prime time” race, but I know plenty of others care).
1 Like
You don’t necessarily need dynamic pens for every race, you could for instance have an option for any race organizer to use dynamic pens with some parameters and just have it as an option for them, you could have it apply only to races with more than ‘x’ participants, but not multi-race series’ with GCs etc. You could just try it for a specific Zwift event to see how people feel about it.
They could also shift around the race score category boundaries all the time by x% one way or the other etc, and just not have a static boundary that keeps people away from a singular target to hit to move down a category.
3 Likes
Why FRR and chasing tour do their own cats not zwift power ftp based or zrs. Cannot do a week plus stage race with cats changing every day, a three week stage race would be disastrous with ZRS. ZRS is good for most zwift races that solo 20- 45 min long or tiny races and maybe some longer one day if not part of series. If ZRS used for a series or stage race only use the GC for ZRS not daily/weekly race results.