Might consider it if I ever train again for a big race.
$16 per month, look back through training data Strava /Garmin etc and then uses a trained machine learning model to suggest a training plan. At least, that’s what I understand from the blurb???
Would be interested to know from anyone who has tried.
Been trying Tridot training system over the 2 weeks. It looks pretty good to be fair, but just don’t think a structured training programme is for me. Life just doesn’t seem to work out in a way that today I can easily do a 70 minute run with threshold repeats, and tomorrow a 60 minute swim + 30 minutes intervals on the bike. I prefefer coming at it the other way round - these are the things in life I want or need to do today, like work, shopping, a bit of coding, some reading - how can I fit in some exercise around them? Don’t really want to change that at this stage in life.
But the daily training plans actually did seem very good, and just the right level of difficult based on goals and recent performance data. Supsect that someone who has not previously followed a structured plan, and could commit to following a plan like this, might see some big improvements.
The only other thing that put me off was that the marketing seemed a bit disingenuous. The original link was “you have been invited to apply to the TriDot project for 2 months free training”. After applying, got an email “You have been selected” and got some daily training plans. Then, after 2 weeks of this, another email saying “activate your 2 free months of TriDot today”. When I followed this link, there was a screen to enter credit card details and pay £139, discounted by £139, for a 12 month subscription. Afraid I lost interest at that point. This was the first time anyone had mentioned taking out a 12 month subscription. It all felt a bit opaque, and as if “being selected for the project” actually meant signing up as a subscriber and recurring payments and all that jazz… who would have thought
So I have switched it off. But might be of interest to some.
I actually question how intelligent much of this is. I guarantee much of it is algorithms that compares little more than CTL v ATL v current hours and has a bank of sessions stored that it changes based on pace/power/HR. I would almost guarantee they are not “learning” from other athletes results and replicating real coaching.
I agree. They could look at a lot of data from say Garmin or TrainingPeaks or similar and look at sessions and its related data. But how they learn that then translates to fitness increases I cannot understand. Not as a true AI, it has to be algorithms.
Also, how they could gather enough actual coaching data, session data, life stress, etc. to then learn how to coach? Just cannot see it as feasible.
I would say it will get there, and probably could do a lot better already, but sounds like it’s not being asked to do the right things.
Garmin for instance, could take HRV as a life stress indicator and automatically adjust sessions and intensity. It just needs a weekly questionnaire to ask if there are any days in the next week where you can’t train, or what day for a long session etc. Then setting a week of sessions should be a piece of cake.
Tridot does allow you to move sessions to different days I think.
But people are so different, with different motivations and priorities - I can’t see how AI coaching will be able to match human coaching, for figuring out those aspects. Although can imagine it getting the physiology right & maybe see it being used as a tool by a human coach?
My Garmin does suggest bike\run training sessions per day, not sure why it thinks a 3 hour bike and 1 hour run is possible for someone working FT though as that would be a big commitment on a weekday.
Going on things I’ve read I wouldn’t be surprised if they end up in Connect+ with a bit more logic before long, it won’t impact me though as I ignore it but can’t remember where to disable it!
But today most people download a fixed training schedule from the internet. It just has to be a bit more intelligent than that - you choose the plan at the start, it monitors how you do against the plan (Training Peaks has been turning workouts Green, Orange or Red for years), and then suggest adjustments.
I see you missed your long ride, do you want to retry this week, or should we adjust for something more manageable?
You didn’t hit your target swim pace, try shorter efforts next week with longer recovery
If as @Hammerer says it is just doing a CTL / ATL calculation, then it should be able to work out if you are on track for your race day target or need to adjust.
I see you missed your long ride, do you want to retry this week, or should we adjust for something more manageable?
Well my knee has been a bit dodge and TBH it was raining and I CBA.
Good to know. I have a physio on speed dial, I can refer you. They only give me a 20% referral fee
Need some motivation? I can add a motivational podcast to your subscription for only $9.99/month
Rain you say, how about a warm weather training camp? My mate does one, I’m sure I could get you a place. Just let me know the long number and the security PIN and I’ll get right on it
You are probably really helping it out by identifying failure cases. Our youngest at uni has picked up a gig job helping to train a coding AI, he has to find ways of tripping it up- that’s what they are paying him to do. Really good money for a student too, we’re all 100% in the wrong job (except possibly @explorerJC by the sounds of it)