Australian Cycling Forums - Bicycles Network Australia

g-boaf wrote: ↑

Sat Jan 18, 2025 6:12 am

I got a decent ADZ lap yesterday - 45:38, I paced it very well and my power seems to be getting better.

Nice time!

Second all time one hour power PB of 252W in today's 'Tour de Zwift' event.

The 'Snowman' course is nasty at race pace - half of 'the grade' (2.11km/8.7%) stacked immediately on top of the Mayan and Itza KOMs (6.34km/2.2%). Didn't look too much on paper but when you've just spent 10 minutes doing 5W/kg to try and maintain 5W/kg for another six minutes becomes pretty difficult.

5th of 240. Interestingly of the four riders who beat me (all A+ on Zwiftpower) three of them were obvious cheats. 20 minute in excess of 5W/kg, no linked Strava accounts, racing scores of 600 each - which indicates that despite racing multiple events per week but none of them were actually races, just group rides. 600 is the minimum racing score, and exactly where you'd sit if you were avoiding events that may put you under any scrutiny. Was good to see the actual winner was genuine though.

To beat my all time one hour power PB twice in a week suggests something's working!

Interestingly, ran into a very simple way of seeing if someone's genuine on Zwiftpower.

There's a hidden metric called 'HRR' which is literally the ratio of their heart rate to their power. (Power/HR).

Because I'm a small mammal mine is always (and I mean always) between 1.41 and 1.58. So at 1.41 in an event I'd expect 254W average at 180bpm.

Everyone's is clearly different, and it varies on the time it's maintained. But the consistency of the figure can tell you if someone's for real or not. If it only varies a maximum of 0.2-0.3 (mine's 0.17) their power is relative to their effort and they're for real. If (for example, checking a bunch of names from my event today) it swings wildly between 1.3 and 3 their power output has no relation to their heart rate and they're FOS.

Going by this metric at least half the event leaderboards are consistently bogus. Interesting that!

Sort of surprising it hasn't figured in any of the anti-cheating algorithms. Apparently some of the event organisers use it.

Stopping cheating means less money coming in when the cheats are banned.

So not going to happen.

Also heart rate information can be falsified, as can power data.

IRL racing is at least not affected by the exploits and cheating that is on Zwift. Also not just open cheating but the use of exploits of either power measurement from trainers or power meters or finding exploits in the Zwift pace dynamics.

g-boaf wrote: ↑

Sat Jan 18, 2025 3:29 pm

Stopping cheating means less money coming in when the cheats are banned.

Yes, this is it, isn't it?

I find the whole process of working them out interesting, though by and large don't care enough to actually flag anyone (unless I keep running into them repeatedly).

It is strange how many people are perfectly happy to cheat to win an online race with no prize where half the field can clearly see what's going on. It would be pretty demoralising being in the half that can't see that though!

As it stands the important thing is it's good motivation and I'm hitting numbers I've literally never hit before. Outside racing is a bit of a no-go to me now I know how much it hurts to hit the ground at 48km/hr.

I'm on ADZ most days and the cheating is rampant - it's even more obvious than in racing. I do flag them and provide detailed info, but I want to also do my lap and get a reasonable time also...

My multi-tasking abilities (eg, riding and typing) are somewhat limited.

Anyhow, another ADZ done, 53:30 today.

Have rediscovered there's absolutely no point pushing until the last minute or so in a Zwift race.

Towed around a combined 'A' and 'B' grade field on the hilly 'Tour of Tewit Well' course today (two laps, 21km/409m elevation) to get dumped in the final sprint which is after a long downhill descent. Led for about 95% of it.

Unfortunately there's no way of holding off anyone heavier on Zwift in a downhill sprint finish. Particularly when they averaged 3.6W/kg versus your 4.5W/kg for the rest of the event.

Mr Purple wrote: ↑

Tue Jan 21, 2025 11:53 am

Have rediscovered there's absolutely no point pushing until the last minute or so in a Zwift race.

Towed around a combined 'A' and 'B' grade field on the hilly 'Tour of Tewit Well' course today (two laps, 21km/409m elevation) to get dumped in the final sprint which is after a long downhill descent. Led for about 95% of it.

Unfortunately there's no way of holding off anyone heavier on Zwift in a downhill sprint finish. Particularly when they averaged 3.6W/kg versus your 4.5W/kg for the rest of the event.

Ahh, but Pace Dynamics 5 is so great, it will stop this from happening, racing will be so exciting, blah blah blah. So they say. But it's still the same result.

It's no use when the people can sit on at comfortable watts and then dump more than 1000w average in the final sprints.

Majority of Zwift racing is not for light weight riders.

The pack dynamics are appalling. I absolutely can't stick a long break in A grade, though could manage it in B and C.

You literally need an extra 2W/kg to hold it from any pack chasing you bigger than 2 riders. In this one was up 13 seconds after the initial KOM which was about 2km in. Lost it instantly on the descent. They just chase you down without any effort.

Mr Purple wrote: ↑

Tue Jan 21, 2025 2:41 pm

The pack dynamics are appalling. I absolutely can't stick a long break in A grade, though could manage it in B and C.

You literally need an extra 2W/kg to hold it from any pack chasing you bigger than 2 riders. In this one was up 13 seconds after the initial KOM which was about 2km in. Lost it instantly on the descent. They just chase you down without any effort.

A grade you need FTP of 500w+. Some of those people apparently really are at that power level and appear to ride IRL.

Difference in IRL is you have cornering and braking. In Zwift on downhills you just fly through corners at 80km/h no bother.

If anything they could put braking in that forces the bigger riders to slow down earlier for the corners, since in real life, they take longer to stop on 10% downhill.


For fun I’d make everyone have precise steering and brakes for races and then enable collisions and crashes. That would also make it harder for bunches to chase breaks. They would have to worry about crashing as well as chasing.

They all talk about realism and wanting more of it, but they shy away from the above.

It would be like sim racing but with pedal power.

g-boaf wrote: ↑

Tue Jan 21, 2025 4:10 pm

A grade you need FTP of 500w+. Some of those people apparently really are at that power level and appear to ride IRL.

There are a remarkable number of people with FTPs in the high 400's, but I've never seen anyone 500W+.

I did read a thread where someone claimed to have a 1 hour power (which we all know is not actually FTP) of 460W. He was shut down pretty quickly when someone pointed out that was more power than the 440W Bradley Wiggins hit for his hour record. There are any number of Zwift racers approaching that, many of which are extremely suspect.

All you need to win in 'A' grade is the ability to hold 3-4W/kg for about an hour and then finish with 1000W+ sprint for 15 seconds. I hit every metric except the 1000W sprint.

There's a good article on it here:

https://zwiftinsider.com/top-a/

And this graph is relevant.



My 20 minute is 4.72W/kg, 5 minute 5.34W/kg, 1 minute 8.34W/kg but sprint only 11.18W/kg. So I'm great at getting podiums in 'A' grade races but not actually winning.

Annoyingly my outside 1 minute is 10.2W/kg, and outside sprint 15W/kg. It's just impossible to replicate those on an indoor trainer, especially after an hour at FTP.

Mind you 'size' is a bigger factor in most Zwift events because they're almost invariably dead flat. And it's possible to draft for all but the last 15 seconds due to the pack dynamics.

There is a guy on Zwift forums I saw who reckons he’s at 500w if I remember right.

I don’t believe all of those, could be trainers and power meters being a bit optimistic, etc.

Hard to tell.

https://rouvy.com/press/ROUVY-acquires-FulGaz

DC rainmaker on the low down
https://www.dcrainmaker.com/2025/01/iro ... lenty.html

jasonc wrote: ↑

Wed Jan 22, 2025 8:20 am

https://rouvy.com/press/ROUVY-acquires-FulGaz

DC rainmaker on the low down
https://www.dcrainmaker.com/2025/01/iro ... lenty.html

Thanks for the DC Rainmaker link.

I bailed on Fulgaz a few months ago because I felt there was little interest in maintaining fair competition for the events. This explains why, there was probably little interest in anything because they were understaffed and I'm sure some of them knew what was coming.

From reading that article Fulgaz is effectively now dead. Wonder if any of my filmed rides will make the top 100? I doubt it for some reason.

Mind you that's a whole bunch of virtual KOMs I hold that will never be challenged. Because no-one will ever be able to ride them again.

Feel for the crew that's been laid off, I knew a few of them. There were literally about 5-6 guys holding the whole thing together.

jasonc wrote: ↑

Wed Jan 22, 2025 8:20 am

https://rouvy.com/press/ROUVY-acquires-FulGaz

DC rainmaker on the low down
https://www.dcrainmaker.com/2025/01/iro ... lenty.html

Just, (couple of months), when I was getting into it (FulGaz), really liked it, don't care about "other" riders or AR, was churning through lots of rides, I have apparently done hundreds !! in over 30 countries and was actually improving !!! Guess I will enjoy it while I can, yes doubt a lot of FG rides will be ported, pity, some were pretty average but some had obviously had a lot of planning and good execution.

I guess that's the way of the market.

Well, it shouldn't really be the way of the market.

What Rouvy have done is essentially bought a competitor to close it down. They'll get minimal value from adding the filmed rides (several of which are mine), the user base is only 20,000 and most will simply cancel, and they've sacked all the technical team so won't benefit from their expertise.

That's just buying the competition to stop it providing competition. Capitalism!

jasonc wrote: ↑

Wed Jan 22, 2025 8:20 am

https://rouvy.com/press/ROUVY-acquires-FulGaz

DC rainmaker on the low down
https://www.dcrainmaker.com/2025/01/iro ... lenty.html

Awful.

Feel sorry for those employees. Fulgaz was great, it had some very high quality videos of real world climbs I do.

I just cancelled my subscription. I got no email messages regarding the change/sale, apparently one was supposed to be sent. So that's it, gone.

Once Alpe du Zwift loses its appeal it might be time to fire up old PerfPro Studio again - no subscriptions needed for that, one time payment, a 5 trainer license.

Made a mistake today. Saw a Zwift 'verified' rider in my race and figured they were a pro so marked them to stay with for an express ride to the finish.

He made me tow him the whole way then quit at the bottom of the final climb. Absolutely no help chasing the suspiciously fast Polish rider off the front.

Checked the account afterwards and he's a professional, but a triathlete. A very good one by the looks of it, too! Going by his power curve (6.5W/kg 15 second, FTP 4W/kg) shows how clearly they need to prioritise steady state long distance power over climbing and sprinting. Plenty of KOMs at 60km/hr, you just don't want to rely on one to tow you to the end of a Zwift race.

Quitting with 600m to go was a bit low though! I was winding up for the sprint finish and then all of a sudden was riding solo.

Sounds like racing tactics in the brave new world of racing scores.

g-boaf wrote: ↑

Thu Jan 23, 2025 3:41 pm

Sounds like racing tactics in the brave new world of racing scores.

Probably more Zwiftpower racing score.

I think we're all happy if our Zwift racing score goes down and drops us back into 'B' grade where easy wins are on the cards.

Mine actually went down this week from 773 to 758 despite a second and third in 'A' grade. Small fields knock it around a bit.

The annoying thing is I end up doing pretty much the same three races every week because they're the only times I'm on the trainer (though the courses obviously vary). You'd think an international application with hundreds of events per day would shuffle the order occasionally, but no. Pretty sure I was doing the same three races when I was last doing this two years ago.

ADZ today KOM was 42:19.9 at 204w, 5.1w/kg, 92rpm and 140bpm.

That power and w/kg means 40kg. And he was “B” grade 764 racing score and 64yo…

His avatar looked nothing like 40kg.

And another one at 36kg.

Zwift is just as bad for cheating as Fulgaz.

g-boaf wrote: ↑

Thu Jan 23, 2025 6:37 pm

Zwift is just as bad for cheating as Fulgaz.

The difference is at least on Zwift someone cares. There's nothing they can do about it, but at least it can be commented on.

One of my Fulgaz KOMs was taken out today by someone doing 28.4km/hr on a 13km 2.2% climb with 143W. Bike weight + rider weight for that one would have to be 12kg combined.

I didn't even bother saying anything. They inevitably whine 'oh, I just use it for training, I don't care about my times'. Then why is your weight set at 5kg?

I've been at it for many years trying to get the cheating stopped, and it's the same old thing.

Thought I had a shot at getting upgraded to A+ today on the Ventoux TDZ race, but the humidity got to me and absolutely exploded and quit about a third of the way in. The guy I was duelling with ended up sustaining 4.9W/kg for an hour including a 5.4W/kg 20 minute so don't think I was ever in with a chance.

Think A+ will have to wait until winter. Though managed to finish 2nd of 485 on the shorter event an hour later. Probably would have won that if I hadn't been so dumb earlier on! It's all character building.

Update on the M4 Pro Mac Mini - so far it has proven quick and stable with Zwift.

It does get warm but never has any problems.

I haven’t yet moved on through old Mac Pro 2010. If I do it will get an RX580 GPU put back in it and probably be reverted to MacOS 11.2.3, the newest it will run without needing OpenCore.

I am up to 176 Alpe du Zwift laps.