TrainerRoad's Big Data

Hire a lawyer? Screen scraping violates the Strava terms of service. And I don’t know about privacy laws across the world, but surely screen scraping has crossed ethical boundaries.

A few reasons why I proposed building an opt-in community of like-minded individuals.

1 Like

Ha ha, I wasn’t suggesting I’ve done it. My point was it could be done if you so chose to. I imagine screen scraping is against every companies terms and conditions, yet there is an entire industry based around selling scraped data.

Anyway, you’re right, an opt-in community is indeed a much better idea.

1 Like

Given that TR has accomplished as much as you have, my guess is that you already have plenty of ideas. But, simple & valuable things would be optimizing the training plans for riders after they’ve logged a given amount of riding. This might rely more on getting ALL of a rider’s power+HR data into the system, assuming that most cyclists don’t exclusively do TR, that they’re riding outside plenty, racing etc. You should be able to see what sort of approaches tend to yield what sort of results.

For instance, I’m sure there are riders who’ve read all the papers lionizing polarized training who select TR workouts and do their own plans. You can likely identify what sort of results (i.e., changes to FTP) they’re getting.

OR, you might even up-end certain training assumptions, like the hegemony of FTP in the first place :wink:

I’d like to know the percentage of users whose TSS drops at the 3rd week of a mesocycle.

I seem to do well with the first 2 weeks then I tend to have trouble meeting the 3rd week TSS before a rest period in the 4th week.


Not a bad idea actually. It would be nice if TR had some plans with a 2/1 rather than a 3/1 or 5/1 rhythm (I don’t think there are any?) Personally, I find the 3/1 rhythm OK, but was hanging by week 4 of my last SSB2!


That’s the beauty of the Calendar. Make those changes and set your own On/Off timing.

Here is one recommendation that I am trying this year.


Here is my personalized SSB1 and SSB2 plan, currently in week 2 of SSB1:


  • SSB1 week1
  • SSB1 week2
  • SSB1 week3
  • SSB1 week6 (RECOVERY)
  • SSB1 week4
  • SSB1 week5
  • SSB2 week1
  • SSB2 week2
  • SSB1 or SSB2 week6 (RECOVERY)
  • SSB2 week3
  • SSB2 week4
  • SSB2 week5
  • SSB2 week6 (RECOVERY)

The calendar makes it easy to push weeks and then copy a recovery week to fill the gap. I played around with that schedule above, and depending on Nov/Dec work travel might still need to modify.


I am most definitely doing this for SSB! Thanks for the info.

1 Like

Some great ones in here.

I guess the one of the key things here for both TR and Users would be what Actually makes people faster. This is a tough one how do you get the actual results for a user and how do you determine if they did well in a race or whatever (accounting for weather etc). Maybe looking at different power buckets and the improvements in those and attribute those to different rider profiles or goals - say if you’re training for 10/25 mile TT then the plan that benefits your FTP would be best or if you’re doing Crits then VO2 Max Power and repeatability would be best

As many people have said the big thing here is to say - my Big Goal or A Race is on X date. It’s Y type of race (TT/Road/Crit/XCO/CX/Tri) it’s Z distance and then based on my training history (+ Swim CSS and Running Cooper Test - if you want to get really fancy) and what TSS I can sustain then using the Global trends TR has calculated from the above schedule out the optimum training plan for me

TR could sell the data to Power Meter makers and Smart Trainer makers to help improve their products.

Also sleep tracking vs performance would be interesting

Problem is that this just kicks up more features for the app to capture the data anyway!

Take the data I already feed into Garmin, My Fitness Pal, and Strava and dump it into the TR data bank with everyone else.

Then ask me my goals (obviously, make my friends envious of my incredible fitness).

Then do magic number crunching.

Then make it as easy as possible for a 40 y/o overweight father of three to meet unrealistic expectations of body fat percentage, speed and endurance.

Should be straightforward.


I know you’re being sarcastic (or at least I think you’re being sarcastic) but for those of us tracking all of these things it is a feasible goal to have this done in the next few years.

Data I have entered into a database that could be correlated to my fitness and fatigue include:
Sleep time (Polar)
Sleep quality (Polar)
Resting HR (Polar)
Activity level (Polar)
Calories consumed (MyFitnessPal)
Training Stress (Strava, TrainerRoad, and Training Peaks have all of this - Garmin has a subset)
Weight (digital scale)
Body composition (digital scale)

All of this paints a fairly complete picture of me as an athlete and, with a post-workout self assessment, could be used to determine more realistic targets.

You saw companies like Xert and Garmin starting to play in this over the past year or two and there is still a long way to go - to me they clearly aren’t there yet.

I think there are a few comparatively easy wins to be had by beginning to integrate activity and sleep tracking metrics but even people who have single source of this (Garmin) really struggle with it across devices (get one of their fancy watches and see how the data doesn’t show up on your bike computer showing fatigue correctly…come on Garmin, get your act together).

For someone like TrainerRoad I’d love to see them do something a bit simpler like analytics on breaking within a given workout and simply giving users the suggestion or option to dial it back for subsequent workouts of a similar type.

Beyond that it is all about integration. You have to be able to consume and process data from a huge variety of sources and get it into consistent structures. As a simple example - every sleep tracking company seems to use a different algorithm and store data in a different way. This needs to be standardized from a messaging standpoint such that the limitations of activity tracker X don’t force everyone who wants to consume their data into a strange data quality gap.

There’s a ton to do here but the data is available and someone will pull it all together. Xert has the ‘first to market’ advantage here but their solution is far from complete and they have not established themselves enough to dominate the market in any way shape or form

1 Like

My comment was only really partly a joke because it seems a little far fetched right now, but I totally agree with you. I think that this is definitely the future.

My guess is that the challenge for those of us on the consumer side will be finding the algorithm that the delivers the outcome we are looking for. If it’s a power to weight goal or even body fat percentage then it may be fairly easy for a company to plug and chug the numbers for us using big data.

Companies racing this tool to the market will probably try to sell it on intangibles like race wins and confidence that are nearly impossible to measure.

My suggestion is twofold, and pretty low-hanging I should think.

  1. Analyze completion rates of individual workouts. This came to me during SS base low vol 1 when doing Palisade, 5x(1-2min) over/unders. I knew probably in the 3rd set that I wouldn’t manage the entire thing as prescribed, while I breezed through all other workouts in the plan. So the 5th set was done at 85% of the intended intensity, making for a “failed” workout.
    I suggest all workouts be analyzed for this, creating statistics for each workout within each plan. This should pinpoint workouts that are too hard for the intended spot in the rider’s training curve.

  2. Create a community - allow users to friend or add each other to track each other’s progress more easily, brag, encourage and even challenge. Even better, make the stats from point 1 public to add a very interesting dimension into bragging rights … “I completed Palisade in SSI which only 8% of all users can do … how about you?”


Hey @Nate_Pearson, any idea how many Kw/Hs are produced by all TR users in a year? What could we collectively power with our legs?


With a quick query it’s 5,002,590,783 kJs or 1,389,608,550 watt/hour.

That means we could roughly power New York City for like 2.5 hours :laughing:. You can think how hard it is to power a 100 watt bulb for 24 hours.

In terms of food though it’s a lot! That’s 8,885,596 McDonalds Big Macs!!

If you take into account all the data we’ve synced to TR (historic indoor and outdoor) that it ballons to 25,861,672 big macs!!


I prefer to have my share in pizza, if possible :sweat_smile:

1 Like

2,908,483 medium Dominos pepperoni pizzas. :pizza: :smiley:


I’m wondering if it wouldn’t be possible to get rid of FTP assessments completely and have TrainerRoad estimate FTP from past rides.

This is what Xert does. Assuming you do a hard interval every so often your FTP will stay up-to-date. I tested it a while back and found their maxmum power available approach to work really well on the first interval when I was fresh. However, it didn’t seem to take into account cumulative fatigue.

1 Like

I would love TR to be able to predict my expected FTP bump based on completing a plan (with a +/- tolerance) and dynamically alter it based on my completion rate of individual work outs etc. Personally I would find that a great motivational tool.

Also if the calendar function could predict FTP across multiple plans based on typical results by age group etc. that would be super useful to see what you could potentially achieve based on which plan/volume you pick.

I’d also be happy to allow wearable metrics to be imported plus data from other apps such as MFP, anything that helps paint a clearer picture of achievement and potential, and thus gives me a more efficient training plan