Could you expand on this, as previously Zwift staff have acknowledged it is an issue - There is fairly lengthy thread on here whereby people tested this and whilst not outright confirmed, many people saw an unrealistic increase in speed in using the spiky watts output.
A few people have mentioned fairness in racing, can you confirm how this would be applied fairly to all if getting boxed in and the like with no steering etc
Further question, is this change a small way to try and nuke blobs & sprinters winning too many races? I.e. 1 min power is in someway nuked as users need to back off and think about power output over just raw power?
I think introducing greater racing tactics would be beneficial - I just think its needs to impact all riders fairly & equallyâŚ
Iâm quite certain (and belive my earlier arguments show) that auto-braking in its current guise would be a quite significant detriment to race tactics, as youâre robbed of control in critical moments of the game by having the game system force your hand (or legs, as it were) to a greater extent than before, even to the detriment of breakaways.
@DavidP, Iâm currently riding with Coco using Sauce and noticed a draft behavior which Auto braking may be able to correct. My draft bonus is dependent mostly on my speed, and not the speed of the riders around me. Pack is a chill 41kph, and my bonus is 115. If I increase my speed above the pack, say to 50kph, my bonus is closer to 200. Perhaps instead of braking to stay behind a rider, maybe the brake could be proportional to the amount your speed exceeds the rider in front. In a sense, negating a draft benefit which shouldnât exist. This could slow the churn and the pack.
I wonder what you mean by âin itâs current guiseâ seeing as it has already changed significantly from the tiny races and will continue to be changed as already explained by David. Pointing to the Tiny Races and saying âit doesnât workâ is counter-productive.
Itâs still badâŚAutobraking is too much. It feels like you need to push 100% more power not just 10% to move up. There is no REST it feels like iTT and Downhill is really bad.
The Group speed didnât drop at all IMO with no yoyoing.
Last 2km i couldnât move up or close the gap to the rider infrontâŚit just holds you back.
Didnât have the legs to race today so I tried to wheelsuck my way in C instead. The pack got veeeeeery strung out right from the start (basically everybody single file) and generally drafting was very difficult (balancing between not getting dropped, not ending in front vs. getting pushed to the side and further back) much of the time. Drafting per se may not have changed, but obviously the kind of pack formations the other aspects of the pack dynamics produce and affect have a huge impact on the experience of drafting. I did manage to jump solo from one group to another fairly easily and even got away from my group on the descent with a tired little sprint.
Overall verdict: PD 4.2 worse than 3.0, see youse on RGT for now.
Yeah, itâs difficult to discuss and reference auto-brake accurately, as we donât know which exact criteria actually activate the auto-brake across the different versions and races, other than what David has indicated in broad terms, which is that it happens in situations with high draft, many riders and not uphill.
While more testing and tweaking is definitely needed, in the post that youâre replying to I had all versions of the auto-brake in mind that use a set arbitrary limit to disengage it, such as upping your power above 10% of your last 10 sec. average.
I would likely also object to a lot of the exact criteria for activating the auto-brake to begin with, but we regrettably donât know these yet and theyâre thus hard to discuss and give feedback on.
I think I can agree on there maybe needing to be a draft cap, where if you move faster than those around you, you are virtually âin the windâ. Even when an avatar slides through the middle on screen. Another point Iâm contemplating is how things may be different if we were to consider a thicker air density or indeed a minute virtual headwind. The main trouble is to account for all different scenarios though. Its hard to over see and rationalise it all.
Btw I suspect this autobrake will not work great on rollers. Nr1 goal that needs to be met is that a rider feels in control of what happens to the avatar. If that connection is lost many will lose interest.
My thoughts it that 4.2 i a step forward but the downhill feeling is strange. I was trying to split group a couple of times but no one was willing to pull. On the âhillyâ part the group got stretched out though.
âPack Dynamics v4.2 Test Race on Watopiaâs Waistband in Watopiaâ Cat.Câs experience.
The way the pack stretched and cut off when the pace increased seemed more fair. Before, when I picked up the pace, the back of the pack followed me like an âauto-cruiseâ.
I also like packs of around 10 because it takes more concentration to keep up. It is brain-numbing, though. I feel that there is a lot of âloose racingâ in Category C, where people free ride each other for most of the race after the start, which brings me to a good impression of this test, which requires more concentration to maintain the pack.
I havenât noticed this much but i will pay more attention from now onâŚiâm searching the forum for some tips how to fix it.
Maybe @Dave_ZPCMR can help
Edit: I notice rubberbanding camera effect everytime my speed dropsâŚmight be something to do with Autobraking
Sauce (requires windowed mode), plus capturing, on a 60Hz display. CPU load is dipping the frame rate all the time, and windowed mode is cack for smoothness anyway.
Ok first of all thank you for all the feedback and sorry If I canât answer everybody, this thread is getting a bit out of hand
Iâll try to leave some closing statements before getting back to the code
@DejanPresen thank you for your video, it does highlight probably the current major problem to the new system where you got some braking at one point where you shouldnât for no apparent reason.
On the other hand it also seemed that you were pressing much harder than needed at certain points. That is a common thing Iâve been seeing. It seems that a lot of folks spend much more energy than needed and donât trust the draft enough, making any needed strong acceleration much harder.
Future iterations
One change Iâm going to propose is instead of using the 10% (or whatever value) above last 10 sec. average power to disengage any braking, I think it makes more sense to do the opposite. Only if your power drops below a certain level under the last 10 sec. average you might get some braking because you are signalling that you are slowing down or ârelaxingâ. Remember that braking is very gentle, this is not full stop brakes.
Other good news
I found a way to disable double draft in the current version without the need to update the game. That way we can run a couple of more tests without double draft using the current system. I do agree that speeds are still too high maybe due to double draft (?), however I still think that increased draft is the way to go in the future.
So is that on the v4.2, with or without the 10% threshold? My thought would be that if draft bonus is capped to the speed of the rider in front of you, then you wouldnât have to meet a threshold to pass, wouldnât get an added benefit of going faster through a pack, but could also jump on a wheel going faster and get their draft bonus.
