TR makes you this much faster – 185 data points and a model

I just experienced this for my first time this year. Started SSB MV1 after a good ramp test result and workouts just got harder and harder. Looking at my CTL trend, it was falling very fast and I concluded the lack of volume was dropping my FTP, making the workouts harder and harder. I switched to a HV plan that I opted for 1.5 hour weekday workouts just to close the gap in TSS jump from MV to HV. Things seem to be going well now.

1 Like

Very interested to check this out in practice :sunglasses: Just this Monday I started SSB MV1 with 302 watts as a start point. According to the graph I can only expect (on average) a 2,81% increase after finishing the first phase. That would be a bump up to 310 watts (+8W) in 6 weeks time. Very curious if that will be the case.

1 Like

Cool! How did you get to the 2.81%? Taking the linear fit from the second graph for MV, the average percentage gain would be -0.00075 x 302 + 0.26 = 3.4% or ~10W. All within the error bars of course, but just wondering how you got the very precise percentage number?

1 Like

This is going to be weird but… :stuck_out_tongue: professionally I work with 3d engineering and graphic software which I can also use to analyze and extrapolate mathematical and statistical data from 2d images. In a nutshell I took all the ratios and relevant pixel points of the chart image and extrapolated that for my 302 watt ftp the line crosses at roughly the 2,81% point of the vertical axis. Its not hyper accurate but it gives me a rough estimate. But then again it all comes down to the quality of the image and pixel density, the bigger the image the more accurate I can make it. But it’s not science, so don’t quote me on it :stuck_out_tongue:

2 Likes

This is cool! Must have taken a lot of time! I just have one comment. I think you may have switched around the labels of the lines in the graph. Because the equations don’t seem to match the lines that they are color coded to (The smallest slope value is the steepest line). I had to quickly mock it up in excel to make sure I wasn’t going crazy before I mentioned it.

I used the same colors from the legend to make this one…don’t know if that’s correct though
image

1 Like

What happens if you throw those three factors into a multiple regression? How accurately can you predict someone’s improvement?

Would be fascinating to see some analysis like this from TrainerRoad themselves. Particularly since they don’t have to rely on self-reported data and could do the analysis to show improvements by specific plan and could control for plan compliance.

1 Like

Very good data. I just completed the High Volume Oly Tri Base and my FTP went up 4%. About what I expected since I’ve been using TR heavily for 5 yrs (so I’m not expecting any huge leaps in FTP). Thanks for posting this!

1 Like

You are correct, my bad! Will update the colors.

I think you may have switched around the labels of the lines in the graph

i could be one! i “only” saw a ~4.5% increase after doing SSBLV1 with a starting ftp of 212. I had a pretty strict adherence to the plan (i think i missed a single workout but also I had supplemented with additional rides)

I’m not complaining or saying the plans don’t work btw, there’s still a lot that i’m figuring out as a relatively new cyclist (like rest, diet, working in strength training), but my results are about half of what people at my FTP tend to see.

and fwiw i didn’t post in that thread because i didn’t think my results were all that interesting and i didn’t feel like there was that much to say about them.

Had not seen spreadsheet pretty cool but only works on pc

On the spreadsheet, what is your assumption about the average TR user in terms of weight and age? Believe that is used to calibrate the model

Yes, wrestled with that. The weight is based on 230FTP average and 3.0 W/kg from the Bell Curve discussion. So that leads to ~77kg or 170lb on average. But not much known about average age and distribution of gender of the TR user community. If anyone knows…

I might be missing something but to me someone selecting HV might have already done a lot of training coming into Trainerroad which means the FTP gains will be much smaller. That to me doesn’t reflect what a high volume versus a low volume plan offer but much more dependent on the rider history.

i.e. I was close to my peak (ever) before I started TR SSBMV I. My ftp gain might in the end can be like 5-10%
But a beginner cyclist might pick up TR, start SSBLV I and gain 20-30%

This doesn’t mean LV makes better FTP improvements for everyone. If anything If I train 12 hours a week and then pick up LV and do no other riding, my FTP might even drop.

But like I said I might be missing something here.

The best proxy we have for training history is FTP (or W/kg). A beginner cyclist will have a lower FTP, and the spreadsheet/model shows these riders will have much bigger gains. The model also suggests there is a maximum FTP that can be reached with LV which is lower than with MV or HV. If your current FTP is higher than this level one would indeed expect a drop as you suggest.

Some people use LV plans so the can freeride, add Z2 long rides, do outside strenght intervals.

I’m having better results with the LV than the MV I followed last year.

3 Likes

Thanks for putting this together @Janhein!

1 Like

Added a W/kg Bell curve to the spreadsheet based on the forum data. Lines up nicely with the Friel blog in terms of age-correction, but the average gender-correction seems to be more like ~15% (as opposed to 10% used by Friel). Also interesting to see that female-cyclists seem to be able to maintain fitness better as function of age.

3 Likes

This was exceptional work, and a great read. Thanks!

1 Like

Fascinating!

1 Like