TrainerRoad's ML Adaptive Training successes will lead to a significant increase in pushback from coaches

That sounds pretty obvious. I just didn’t realise how fevered it was already getting. The Dylan Youtuber wasn’t really that interesting other than young influencers are better than the older gens at using social media to generate attention. But I’ve now seen more serious coaches on social media this morning sniping at Adaptive Training, coaches who’re into data science and AI. And around them a small crowd also buying into the anti-TR narrative.

I think TR have done a great job with the podcast at addressing the criticisms, but I also think they need to bang it home in shorter, punchier promotion on their site and external blog interviews etc, as they’re still being criticised loosely on being ‘sweet spot’ based. Many simply won’t sit through the whole podcast, and misinformation spreads fast these days. Social media is an information war, incredible how fast it can spiral into a PR nightmare once the herd picks up an idea and starts a hate-train. Everyone just piles in.

1 Like

Kind of hard to really say anything at this point. Adaptive training is a concept or theory at this point. 99% of trainer road customers do not have access to it yet. Over time I believe the data will show that it works. We’re not there yet. If it was me I would just ignore the chatter and concentrate on the product. As they say this is what keeps the lights on and food on the table. No matter how good it is people with say negative about it when it threatens their income.

1 Like

I think that’s also an important point they could drive home with PR, that online training is at it’s infancy and not an end point ready to have a final judgement made on it.

Not sure I’d stick my head in the sand from a PR perspective though. Social media is too powerful now and a lot of coaches are already feeling threatened (I’m sure TR team have been reading the comments elsewhere to).

I’m saying this from the perspective of someone who was in the middle an information war waged against diagnostic companies during the pandemic, where a a handful of scientists and a lot of sophistry created mis-information campaigns about PCR testing and the pandemic being fake etc.

There will be people here on this forum who have ideas in their heads about the pandemic that are false, and they will consider it factual, and those ideas were baked to be extremely convincing. Catchy dismissive arguments fly on social media - really good debunking is much more difficult and doesn’t suit Twitter’s format.

Personally, I’d get ahead of the incoming coachmageddon. Ultimately you’re threatening incomes streams, that’s the bottom line, and it was the same as the pandemic - the reality of it threatened income streams, hence free-market antagonists tried to make it appear to be fake, then failing that, less serious.

The best thing TR can do is give their strength to the product, AT. Let the product speak for itself. Coaches are needed and have their place (I used one for 4 years) but are costly. TR has been the product of its users versus a ramped marketing media. The podcasts and users have spread TR’s success. I hope it continues to be. Marketing has its place “on a leash” but it is terrible uncontrolled.

2 Likes

I guess all these coaches forgot about the Streisand effect

All they are doing is helping more people hear about AT that haven’t heart of it or to see more information about it, or question to themselves why their coach would be worried about losing a client to it.

I’m pretty sure I know you are talking about and if it is who I think it is, said individual knows a fraction about ML and AI than he thinks he does.

If your clients are suddenly questioning your monthly fee because of a $20 a month product that says it can pick workouts better than a human, well then you have bigger problems as a coach.

I doubt that Bill Gates’s wealth manager was worried when roboadvisors came onto the market.

6 Likes

Yes!

There’s a lot of truth in that, especially now TrainerRoad has a forum full of people who I think will shortly be saying, ‘oh this AT is great, my FTP went up by X% or I won X event or I’m KOM’ing everything…’

And then they see it costs in a year what these coaches charge per month.

You’re probably right about the coach, I saw him linked in another thread myself. Don’t want to give him more space, he hasn’t even watched the latest podcast judging by some of his comments.

1 Like

Never trust anyone’s take who hasn’t listened to the source material and has an opinion about it.

8 Likes

Usually, but if a source is known to always or at least almost always be a bad source of info. I mean I dismiss Q without reading what Q claims

It’s funny how people who charge to largely train adult amateurs with disposable income would get offended by products attempting to do 95% of what they do at a fraction of the cost. I know if I approached my wife about getting a coach vs what I do with TR I’d get laughed out of the room

1 Like

3 Likes

I wonder if tr is taking away from coaches or enlarging the market of people following training plans.

You should see the coach on the xert Facebook group. Complains heavily about how some people are just leeches trying to get something for nothing by not paying for a coach

1 Like

If coach C (not Chad :grin:) is coaching rider A then rider B asks for the plans and workouts then I’d say that’s leeching - rider B is getting much/most of the benefit of A’s association with the coach without paying for it.

For many/most(?) people, do we really need a coach? I’m not aiming at the Olympics/Nationals/whatever so I have to decide whether the extra cost (over that of something like TR) of a coach is worth it to gain those last few percent of potential gains. I.e. using TR plus a bit of (critical) reading might get me to 85%* of my potential, a coach might get me to 90%. Of course I might have lots of disposable income so the cost-benefit analysis would change.

So for me, N=1 and all that, TR’s AT system isn’t taking away from the pool of available clients as I’m simply not in the market to hire a personal coach. One might as well claim that a book like Chris Carmichael’s Time Crunched Cyclist is taking clients away.

*totally made up figures :roll_eyes:

4 Likes

If rider B is indeed getting most of the benefits of A’s, then I’d argue they are both fooled by the coach. I thought the whole reason to get a coach was to have someone who 1) has experience and skills 2) knows your goals, strengths & weaknesses 3) creates and guides you through the process to hit your targets. In other words, it is supposed to be highly individualized with feedback loops built-in.

If a coach sells plans, how is it better than using TR plans, or any other non-individualized static plans, really? You don’t need a coach for that.

Also, AT will work well for people for whom existing TR plans worked well too. For everyone else - there is no guarantee or good reason to believe it will, until we see a proof to the contrary.

In summary, I’d love to see AT weeding out the population of coaches, where bad coaches lose clients (because simple ML model does your job better than you do) - it would make it easier for people like myself to find a good coach in the future.

2 Likes

I wasn’t meaning that the coach “sold” a plan, more that rider A would be given a set of workouts as part of their agreement with the coach, maybe a week’s worth, and rider B copied those, obviously there’s an assumption that riders A & B have similar goals. Other than maybe an FTP test I can’t see any coach going with handing less than a week’s worth of workouts to their athletes, it would just be too laborious. This is for remote coaching, I’m not meaning those coaches who are present at training sessions, that’s another level again.

Until a coach knows your response to various workouts the initial parts of any plan will have to be somewhat generic. Not as generic as one of the TR plans, more along the lines of “This works for most in your position and goals”, but I’d expect some adjustments pretty quickly once the coach and rider got in tune.

Definitely agree with your point about a coach selling plans - see lots on TP for example. If they are a starting point for the above coach-rider discussion then maybe not so bad otherwise just cookie-cutter trading on a name.

1 Like

Right Ivang.

Great coaches who have worth will be working with top athletes who need that nth percentile edge.

But that is a niche within a niche. The attempt lately to mass export coaching via services like TrainingPeaks and easy to attain coaching ‘certificates’ and coaching affiliate programs needs to get in the sea. We have seen this cookie-cutter mugging of amateurs who simply haven’t found sensible people around them yet to give basic advice or point towards a great plan they can customise to suit their lifestyle with a few clicks.

Bet we all know riders who aren’t even enthusiast level yet they’ve paid 100s a month for a coach at some point. They simply didn’t have a mate to point them in the right direction.

:joy:
Sure thing man

Hmm, I got faster following principles from that book :thinking:

1 Like

Many of us benefitted! Used to be the goto recommendation, forums used to be littered with amazon links to it.

Now you see TrainerRoad and Zwift recommendations. But tangent…

Could argue that some of his work was the beginning of the democratisation of training knowledge and helped open up the industry.

Frank Overton from FasCat made at least three comments against AI/ML-based training plans on their latest podcast. Nothing overt, click bait, etc - just comments.

He also went on a “rant” (his words, not mine) last week about sweet spot versus polarized. Again, not targeting or criticizing anyone or any company.

Point being, lots of other coaches and plan providers took notice. It will be interesting to see how the community responds as AT rolls out to the TR masses.

Also, in another thread, someone posted a link to a great video about WKO5 and simplifying metrics / trying to tease out relationships between all of the data. I would guess that Training Peaks has a substantial dataset based on all of the coaches that use TP and / or WKO. They could be well positioned to embark on AI/ML, too, if they choose to augment their plans and coaches that way.

2 Likes

Just had a listen through now. Yes no knives out, but subtle dig there and he’s definitely been paying attention to TR’s development…

1 Like