Am I the only one with RGT issues?

RGT
40

@Robin-RGT-FB-mod sorry but I deleted the email since nothing they offered was in any way helpful. I would be happy to send you the email I use as well as a list of issues.

Just as an aside, I spent many years in software development. I have considerable experience troubleshooting bugs in software. Nothing that support does resembles any troubleshooting process I have ever done. I would be happy to invest the time to work with some one a systmatic (see what I did there) process to see where any flaw(s) may exist.

BTW - I just finished another RGT ride. Please see my next post.

41

New ride today, consistent issues. I have noticed my cadence on many rides have been dropping so I decided to do a long descent on RGT. I find it helpful training to keel up a rapid cadence for 90 minutes. This ride was 99KM.

The good; it did not drop the bike off of the sensors once it connected! YEAH!!! Also the delay was down to just 4 or 5 seconds, not 25-35 seconds.

Issues:

  1. As is VERY VERY common, the Ride Now button would not connect until the time remaining on the five minute setup was under 2 minutes. Does this happen to everyone?? Anyway, it was a long setup from when it finally allowed me to click the button, so by the time it finished, the ride was already a quarter mile down the road. Now I was pedaling the whole time so I was in the hunt but still?

  2. As ALWAYS happens, I am working much harder than the bot for little or no benefit.
    2(a). For the first 30 miles I was riding at a steady 1.7-2.2 W/Kg while EVERY bot rode at a constant 0.9 W/Kg. As I usually do as a check, I geared up and stood on the pedals, raising my power to over 7 W/Kg. I gained exactly ZERO on the bots riding a steady 0.9.
    2(b). At about that half way point, a group of six bots started to pull away. Again going a steady 0.9 W/Kg. So I crank it up to 2.2-2.5 W/Kg, but they continue to pull away never to be seen again, except one.
    2(c). A few miles down the road one of the six drops power to 0.5 W/Kg. I slowly catch her over about 10 miles. When I do catch her, she stays with me for a full 15 miles doing between 0.2-0.5 W/Kg while I am at my usual 1.7-2.0.
    2(d). With about 20 miles to go a group of bots start steadily closing on me. Yes doing NEVER more than 0.9 W/Kg. Yes they caught me at about 5 miles to go and I could never drop them no matter how much power I produced.
    CONCLUSION - the app has a predetermined outcome for the ride and it makes ZERO DIFFERENCE how hard you ride. The bots will finish where the app wants them to finish.

One final note. Why does the app ask for your weight? Is it just to penalize you on climbs? This was a descent ride averaging 7% decline. I go 90 KG. Given the laws of physics, I should run away from most riders with the same power production. In real life, the group I used to ride with consisted of riders typically 20-30 years younger and 30-50 pounds lighter. On EVERY downhill i would close gaps and take leads just because of my weight. Not on RGT.

1 Like
42

Right, in a few quiet minutes between bits of Christmas mayhem…

#1: For Quick Rides, I think it is typical that it is only a few minutes before the ride starts that the Ride Now button becomes available. This is because it takes a while to set up the server for the Magic Road. If you want to give yourself more time then set up and event (Group Ride or Race) on the course instead. You can then choose the start time, and it can be longer in the future. You can also customize the bot settings. There should be instructions on the website if you don’t find it intuitive.

#2: I am confident that there isn’t a pre-determined outcome to the events - I have not experienced or heard any other evidence of that. However, it does sound like something strange is happening. I definitely think that doing a ride with some humans would be instructive, so hopefully that can happen sometime. In the meantime, if you DM me your email address then I’ll see if I can get someone to look at your old tickets.

The app does take your weight into account on hills - both up and down. Overall, the physics is fairly accurate. One main thing missing is that the CdA and therefore air resistance does not vary between riders regardless of their height.

3 Likes
43

I seldom know much in advance when I will be riding. When I find a window I go and ride. Yesterday was a great example. While I generally ride in the evening I have a window open up in the afternoon so I went to ride. I will look in to it though. But not the biggest issue.

With all due respect, what other explanation is there? Computer programs are bound by how the code is written. In this care, what happens on the screen is limited to two inputs. That which is written into the code with respect to the bots and the input it interprets from my bike.

If I am riding with a bot for 15 miles with the bot running between 0.2-0.5 W/Kg while I am at 1.7-2.2 W/Kg, and I do not gain a single foot on the bot, then there is NOTHING I can do on my bike to impact my relationship to the bot on the screen. Therefore what I am seeing is a function of what the app has decided will be the outcome.

Again, if weight is a consideration, and “the physics is fairly accurate”, then a heavier rider should have a decided advantage in a long steady descent. Clearly that was not the case as I was required to maintain a minimum power output of 50%+ more than EVERY bot to maintain an equal position with them.

Robin I sincerely appreciate you wanting to help and I would welcome the opportunity to get this figured out, but as the app currently executes on my computer, the bots are static players that will finish wherever the program decides irrespective of my relative efforts as displayed on the screen

I do wish you a very Merry Christmas…

44

@Critmark Yes - common for me.

3 Likes
45

I’m a Mod on the Facebook support group, and been using RGT since it’s early days, so I’m pretty familiar with some of the issues. My first thought is to ask what your network connection is like.

Because RGT seeks to model the physics of cycling as accurately as possible, calculations are run on the server, rather than on your local hardware. This makes RGT sensitive to network latency (delay in packets) and jitter (change in packet delay). It doesn’t take a lot of bandwidth (megabits / second) but latency and jitter can cause issues.

Also, based on your write-up, I’m guessing that you’re a fairly light rider, just because your complaint is common with light riders in RGT. W/kg matter more in climbing than on the flat, so if the course is flat (or descending) a heavier rider with fewer W/kg can run away from a lighter rider pushing more W/kg, especially in a pack due to the modeling of aerodynamic effects.

What hardware are you running? I’ll be happy to help if I can.

3 Likes
46

I upgraded a while ago to fiber internet. It has been faster, more stable and much better quality that I had previously. Latency & jitter, at least as measured by the sites I use, is much better than I have ever had.

Sorry but exactly the opposite. I am 198# (90Kg).

I agree this is what I would expect. So unless all the bots are 100Kg+, the app is doing exactly the opposite. I am pushing much higher W/Kg and even on a descent, not running away from anyone.

I ride a Kickr bike, a Windows 10 laptop with a 10th Gen i7 processor, 16Gb of RAM, and a 1 Gb SSD. The only programs on the laptop are Systm, RGT & Chrome.

Much appreciated & Merry Christmas.

47

Ok, just checking the details: What is the weight listed in your RGT profile?

Other than that, let me do some checking with the RGT devs. Have you submitted a support ticket within RGT? If you do it within the program, it attaches the logs, which helps a bunch with diagnosis.

48

198# as it has been since RGT was added to Wahoo.

Multiple with not response that has improved anything.

49

Regarding setting up an event, you can set them up whenever you want - well, to an integer number of minutes - so you could give yourself a few more minutes faffing time than Quick Ride, if you wanted.

Note that, as Chris has stated, the physics calculations are performed on the server so, although the app runs locally, the app is not the only factor that determines who goes where on the road. The rides are not predetermined, certainly if any human riders are involved - there are interactions that happen that cannot be predicted beforehand*.

*Let’s avoid a debate about physics, philosophy and free will…

50

The fact that you have experienced a very large delay with sensor inputs and, even now, experienced a significant delay (4-5s), makes me suspect that there is something wrong with how the app and your devices are communicating. There have been some issues with particular combinations of turbo trainers and operation systems - I experienced one a year or so ago.

The things that you are experiencing with power output relative to the bots are baffling to me. I do not doubt that you are accurately describing what you are seeing and doing but I think that the behaviour is down to a bug of some kind rather than it being something deliberately designed by the RGT team. Not only would that be just plain daft of them, but also the experience of lots of users is not the same as you. What you are experiencing seems to be a minority issue. That doesn’t make it any less annoying for you, of course. Perhaps I am just being unnecessarily pedantic about the way things are described!

Once I am back from visiting family them I am happy to set up a quick group ride with you to test things out and to discuss it ‘live’. Also, if you DM me your contact details then I can get an actual RGT dev to look into it (rather than just a ‘random RGT punter’ like me).

2 Likes
51

Early on, my BT connections were intermittent. In my case, I figured out that my mobile phone was often grabbing a BT session/connection and that it interfered with my laptop’s ability to connect with my Tacx Neo and Tickr HRM. Since I have been turning off BT on my phone, no problems.

I am making it sound easy, In fact, I went through a couple of weeks of trial & error.

52

I thought of that. I have been turning my phone off for the past month. Not the problem.

1 Like
53

I am having similar problems with doing a “Just Ride” with my Magic Roads where I can’t click the join button until 2 minutes before it starts, but then it takes 5-10 minutes for the MR to actually process - even ones I’ve ridden before. So I end up joining the event a mile or more after it’s already started. It’s worse with the longer MR’s I’m using that are 40miles+.

54

No, the just ride rides always start up. Then about 2-5 minutes in I get dropped from the server. Doesn’t happen with magic roads.

55

Today was another new experience. At least some of the issues change daily.

I started the app and went to sensors to connect the bike. It finally appeared, I clicked it from white to green and tried to close the sensors tab. Nothing, nothing, nothing…crash. Had to restart the app, and, lather rinse repeat…crash. On my third attempt I just began to load the ride and figured I would see what happened.

The ride finally loaded 17 seconds before the start and NO sensor connection. Get off my bike, go to the sensors tab, click the bike from white to green and it did let me close the page, but the ride had started without me. pedal, pedal, pedal and nothing. Get off the bike again, yes it dropped the bike, so I clicked it over to green again jumped back on the bike. Pedal, pedal, pedal and finally after the usual delay (today was about 15 seconds) I start to move. However, the last riders is a full half mile ahead of me.

Then was the daily confirmation that you have ZERO impact on your relationship to the bots. I rode for a full 10 miles with a power output of 1.5x - 2.0x of the trailing bot group half a mile ahead. I closed the gap exactly zero feet. Then the app decided I could actually start closing the gap even though I did not increase my power output and the bots did not decrease theirs.

At another point later in the ride I was closing in on a bot. It was about 0.3 miles ahead. So I jumped on the pedals and raised my W/Kg from between 3.0 to over 5.0. The entire time the bot held steady at 0.9 W/Kg. Ho much of he gap did I close??? NONE, the bot gained over a tenth of a mile on me while I was doing 3x to 5x the power output. Yes, I was paying close attention to the terrain and no the bot was not on any different decline.

When I did raise this issue with support all I was told was they don’t have any data on the bots so they can’t comment and sent me a link to an explanation of Real Bots. So they have no capability of EVER addressing this issue.

I know everyone at Wahoo believes this is not the case, but EVERY DAY it happens to me.

I have resigned myself to just riding against the clock as much as I can. But when the bots are there, I get competitive even though it is a rigged game. Is there a way I can do a Quick Ride without any bots?

56

You cannot do a Quick Ride without bots but you can set up a Event (Group Ride or Race) without bots, and you can arrange them whatever time you want (eg ~6 minutes in the future). It takes a fairly small number of clicks to do.

I am a bit puzzled why you are convinced that the things that you are observing represent that app and servers behaving correctly. I may be missing something due to child-induced sleep deprivation but it appears that (1) you either don’t believe me that I (and many others) experience a different and less flawed experience or (2) you think that what I and others have experienced is the outcome of a bug or issue and what you experiencing (which sounds bloody annoying) is what the software is supposed to do.

57

It sounds like there are major problems with how the app is handling your sensors, with a big lag in what is being sent to the servers (which is where the calculations of the avatar behaviour are done) but also some other weird stuff going on.

I honestly think that there is a fighting chance that the RGT devs could help figure this out. They are different from the first line support team. If you send me the email address that you registered with then I will see if I can get the tickets escalated. (Edit: You sent it to me so I have kicked that off. It might take a while due to people being on leave).

By the way, what device are you running the app on? Do you have an alternative device that you could try it on? Eg if you are using a Windows laptop then could you run it on an Android phone or iPhone instead?

2 Likes
58

That is an excellent suggestion @Robin-RGT-FB-mod ! Fault finding by substitution is a great test!
This has been a long thread and I’m sorry If I’ve missed this and it’s already been tried, but @Critmark have you totally uninstalled and re-installed RGT?
Just for completeness, can you tell us what the Graphics Spec of your i7 is?
Can you run Ping, a webtool to test the stability of your internet connection (ping test) and post a screenshot of the result here please?
Can you try connecting all sensors through the RGT Remote App (REMOTE BLE).
Are you using a Bluetooth Dongle or Built-in Bluetooth on your PC?
How far is the trainer from that Bluetooth?
Is your PC internet connection WiFi of hard wired? (Use your normal connection method for Ping Can Be Useful)
We have encountered some VERY strange causes for users’ problems over the years and as with my day job, learned a long time ago to NEVER say NEVER & NEVER SAY DEFINITELY. :wink:
I’ve never heard of issues remotely similar to those you are experiencing. PLEASE don’t automatically assume that RGT is flawed - it has it’s niggles, but as your experience is SO unusual I think that RGT would have come across it if it was DEFINATELY an RGT fault- It might yet turn out to be something obscure, but until we know then all avenues must remain open and no conclusions can be made.
RGT unlike Zwift RGT is COMPLETELY real-time - second by second what you see is what you get.
The logs on RGT can be retrieved after a ride as long as RGT hasn’t been restarted (at which point they get overwritten) If necessary we can give you the path to recover those logs :slight_smile:
If you can persevere with RGT and the volunteers here (some of who are incredibly experienced) then if nothing else, we can build up the biggest picture possible to present to the dev’s :slight_smile:

2 Likes
59

Hey Sir @Critmark, just a lurker, checking in and keeping tabs on things for “fun” (nothing must dull the suffering).

As I’ve said before, I have yet to hunker in on RGT this winter but was having a wee fiddle with settings and setting things up etc. I would highly recommend you create an event as well. You’re really just riding “alone” anyway and it is pretty easy to do. You can set up the event as a Group ride (of 1 rider), with a variety of bots and you can do it from the app (and it’ll take you to the website to set up), or you can do it from the website in advance then open the app at your convenience. This also lets you “join” the ride in advance until the clock ticks down. Did a quick set up test and was able to “join” with plenty of time remaining. see pic

Screenshot 2022-12-27 at 6.23.48 AM
Screenshot 2022-12-27 at 6.23.48 AM1920×1187 324 KB

I’d be curious to know if this resolves or eliminates the issues you’ve been having. It’s pretty straight forward and doesn’t take much longer to set up and do than quick ride (though there are a few more steps).

Hang in there Sir! And follow the guidance of the folks here who have stepped up to lend their wheels. @RGT-FB-Moderator, @Robin-RGT-FB-mod I know who I’m gonna tap if I run into issues with RGT over the winter :slight_smile:

2 Likes