Who’s combines Trainer Road and Xert, I’m a trainer road user for a while but since 2016 I mix both platforms in order to build my fitness. Xert advice me which workout I should do according to my rider profile and then I look in trainer road workout to match and perform. I try 3 times to cheat the xert system and I can’t. In my opinion the system is reliable. Number are equals!For me, it has been useful.
I tried Xert for 5 months last winter and found it too confusing. It was nice to have it confirm my FTP, but that was about all I got out of it. Measures and recommendations were TOO dynamic and it was difficult to understand the concepts they promoted. I don’t miss it. Keeping things simple helps reduce the stress of thinking about what you need to be doing.
After digging into the background on Xert’s power modeling, I came to the conclusion that it’s not a sound power model for FTP and for training.
Everyone is entitled to their own opinion, but I would encourage you to dig into the research on the differences between Xert’s model, Critical Power, mFTP/iLevels and other power profiling and come to your own conclusion.
There’s no need to do a research when you improve you own numbers, I’m like a guinea pig!
I use Xert & Trainer Road. I use Xert to keep track of my fitness and fatigue. I like the way it separates the different energy systems and points towards a polarised method of training which I like and which I think suits me. Like the OP my TR FTP and my Xert TP align very closely. The smart workouts are great when you can get them to work! I don’t seem to be able to get on with the very structured training plans and prefer the more dynamic and adjustable Xert system.
I use TR for the workout library and the simple fact that it just works - no messing around. I love the podcasts and this forum. I had a free Training Peaks account but since TR made the changes a few months ago I’ve not looked at it. All in all I’m happy to subscribe to both.
Just finishing my 4 week free trial on Xert. I like the concept and I like the “Fitness Breakthrough” analysis / concept. Prefer the TrainerRoad workouts though - there is something about the instructional text that makes the workouts pass quicker.
Xert’s workout player is fantastic. I wish TR would have a Garmin IQ app that downloads the workout and allows you to perform the workout outside. Dynamically adjusting the workout as you ride would be a bonus, but just having the option of riding outdoors would be great.
I also agree with keeping track of fitness and fatigue. It’s the only reason why I’m keeping my Strava premium subscription. I wish TR would include it in their interface.
The whole system working without an FTP test is great. I hate the tests.
I wish I could feed TR’s training plans and workouts into Xert, that would be a killer combo.
On the other hand, I dropped my Xert subscription for now:
- Their UI is horrible. It’s a bag of usability fails.
- Their training plan library is no match when comparing to TR.
- All the dynamic functions with everything constantly updating and changing, combined with their confusing UI are really tough for someone who’s new to structured training - there just doesn’t seem to be a plan and structure behind it (I know it is, I’ve read all the docs, but in the end it’s all just seemingly random)
I cannot comment on the science behind it. I wish TR would pick up where Xert has clearly outpaced the competition so that next summer I don’t have to renew my Xert subscription.
When in TR sweet spot base plan, I watched my Xert estimated ftp drop day by day. Xert assumes you are doing maximal efforts, and while there is a blog entry on reconfiguring it was too much hassle.
I didn’t use the workout recommendations. Post ride analysis - I did like seeing MPA drop and sure enough it was pretty accurate when I was dropped by a stronger rider (MPA dropped below 5 sec average power).
During base phase you not going to attempt breakthrough so you should change settings to NO decay!
its not just setting to No Decay, from the blog entry:
To properly enable this setting:
- Identify your last good breakthrough activity: Go back through your history and find the most recent activity whose signature accurately reflects your signature at that time. Ideally, it should be within that last few weeks or months.
- Ensure you have good and complete power data after that point.
- Unlock any signatures that have been manually set against activities after the identified breakthrough activity.
- Enter in all activities with missing XSS for your entire history if possible. The more gaps you fill, the better the results.
- Change your Signature Decay Method from Original to New – No Decay .
- Delete the activity with the identified breakthrough and add/sync it back in to force a trickle-down process from that activity forward.
Yeah. Xert doesn’t deal with lots of sweet work particularly well. It’s probably due to the fact it’s based on the polarised model training.
I’m not so sure about that – thought the key assumption is that you have some rides each week with maximal efforts? Regardless, I can be on TR SSB and if I also go on Wed night hammerfest rides then Xert worked fine.
yes you have to do that!
If you do lots of sweet spot then , because you are below threshold, Xert never seems to think you are tired. Doing a Hammerfest once a week probably is enough to mark you as tired due to high & peak intensity you do in it.
@stevemz I would be interested in your take on Xert MPA. Would it be useful to see W’ Balance for rides and workouts in TR?
There are several issues outstanding with any sort of live anaerobic energy balance:
- Physiological recharge rate has not be established, modeled, or validated
- WBal and MPA are essentially using a curve a cherry picked maximums
- The cherry picked maximums may or may not contain valid maximal data
- The model itself may have a low confidence variable for the data set it contains
The biggest issue is #1 and until that is solved, there isn’t much point in going any further.
Philosophically about #2, #3, and #4:
- The model has to be good (scientifically validated with a large enough structured data set)
- The model has to have good data (true maximal data not just local maximums)
- The model has to have recent data (within the last 90 days generally, but the last 30 is more useful)
- The model has to have diverse data (enough data points across a broad spectrum of energy systems i.e. at varying durations)
If one of these breaks down, the margin of error becomes so wide that you are mostly firing in the dark.
WKO4 is the only product on the market that I’ve used that checks all of the boxes, as well as shows you the margin of error for your data set, but the learning curve is very high, the UI is awful, and frankly most people would just be better off doing some stretching, eating some vegetables, doing their workouts, or practicing bike skills which will have a much higher return on time investment.
I subscribe to Xert to see the estimation of my FTP and to view some of the metrics but I have not done their workouts yet.
Thanks for your answer and sharing those insights. Interesting.
During my times with zwift and GC I used 20min FTP test or race performances like short TT to set WBal by adjusting so that the graph decreases to 0 at the end on exhaustion.
Don’t know if that is a valid approach but the WBal graphs of the workouts and rides matched quite good to my perception. So I’d really like to see that in TR and I don’t know if it’s just a matter of priority or if it’s the concerns you pointed out that @chad doesn’t think it’s beneficial?
What do you think you are getting out of it?
For the most part it would be in retrospect to look at outdoor rides and races. How deep did I go on that segment or surge or whatever.
Additionally the graph helped when reviewing or modeling a workout in GC and you could better see or define the surges.
It was highly requested on the wahoo bolt forum. The reviews on the Garmin plugin showing MPA for example when attacking a Strava segment were quite positive.