After 7 years hiatus from triathlon, I started racing again this year. Finished eagleman in 5:13. I want to improve this next year.
I enjoyed the racing part but I really dont want to plan my work outs anymore.
I want to improve in all three next year but I dont have the salary of a heart surgeon so i am financially limited.
Ive looked at a couple of ways of doing this thats within my budget.
Which one is cost effective that can help me in a long way?
Trainingpeaks Plan - ranging from $60 - $200 usd per plan
Tridot - ranging from $10 to $199 usd/month
Purple Patch Squad - $129usd/month
Endure IQ - $100usd /month
Human Coach from Australia thats $144/month usd (https://www.jarasport.com.au/) I just found him through google search and the reviews looks great. I may be financially stretch but I can probably do a little OT.
Im racing 1 IM and Eagleman again next year and would like to be in great shape.
I love the idea of an AI coach… if it’s not great today it should be better than most over the next year or two.
Imagine if your coach brought a cardiologist-level knowledge to your workout analysis, and evaluated all your workout files closely, then adjusted workouts to match.
They could far more intelligently evaluate your rest, blood O2, weight…
Imagine if your coach brought a cardiologist-level knowledge to your workout analysis, and evaluated all your workout files closely, then adjusted workouts to match.
…
You’re giving cardiologists way too much credit for breadth of knowledge but I do agree that AI was huge potential for coaching given enough inputs,
I’ve used Athletica AI the last 18 months for Olympic, 70.3 and IM training/racing and I’ve primarily had it set for mid-volume that put me in at 10-15 hours/week. For the Olympic and 70.3 it works well. For IM, it maxed me out at 4.5 hours on the bike and that was only a couple of sessions. I don’ think 4.5 is long enough for IM training. That left me with the choice of adding time to those long rides or bumping up to the high-volume plan that would get me those longer sessions and taking out some workouts to keep me in the range of hours I could handle with work. I’m fine with either of those options, but I’ve been doing this for close to 30 years and know enough to make those tweaks and appreciate not having to build entire plans. At $200/year, I feel it’s worth it.
I tried Athletica.ai and ended up moving away from it.
You could probably do the free trial and get a basic idea of the program. A weakness of the program is that there’s very little variability to it, so once you see a couple of weeks of it, that’s the basics of the program for the length of the training interval.
I ended up moving to a human coach and have seen immediate improvement.
If you’re in AUS I’d suggest checking out https://www.jtmultisport.com.au/. I don’t know Simon from Jarasport personally. I do know James and Grace from JTM and both are top notch coaches. They would make my short list of coaches if I was looking for a coach. I’m a direct phone call away from a lot of top coaches across the world and no more than a phone call to get a # from the rest.
FWIW I currently coach, and have coached, a lot of AUS athletes and spend a fair bit of time there for both work and family. If you’re not in AUS you can still check out James and Grace. I know they have a few US based athletes.
ai vs human: It depends if you value human input, having someone know about you, your situation etc. AI is most often a if this than that algorithm vs tailored to your situation. It could be great, it may not. Some ai plans are better than others. Think of it as putting the meat of the bell curve to your particular situation. If you go AI I’d check out athletica.ai
Training peaks plans are hit and miss. IMO most are tailored to the weekend warrior athlete who wants to finish with a smile vs actually kick ass when racing. There are coaches who will tailor a plan to you specifically. (contact me if interested)
Some, many, of the 80/20 plans are low volume lots of intensity.
The other option you didn’t mention is consult with a coach about how to structure your plan, what you need to do when etc. Then just go at it alone. Have a coach help you put together an outline of what to do when and then you fill in around the edges based on your schedule.
+/- 1 hrs per sport depending on week and life stress. That’s 13-16 hrs per week.
Sleep 56 hrs per week, Walk 15,000 steps per day inclusive of your run volume.
Repeat for 50 weeks, Taper for two weeks. Guaranteed you will be in shape.
If you want get a real human coach to keep you on track
You’ll be in shape. No need to over complicate things. There is no rocket science to endurance sport. You will be in better shape on the 15 hrs average plan versus the 10 hrs vs the 5 hrs version of yourself. No real shortcuts available in endurance sport.
I love the idea of an AI coach… if it’s not great today it should be better than most over the next year or two.
Imagine if your coach brought a cardiologist-level knowledge to your workout analysis, and evaluated all your workout files closely, then adjusted workouts to match.
They could far more intelligently evaluate your rest, blood O2, weight…
A huge probleme for AI is garbage in = garbage out, no matter how good your algos are.
So much data in tri is garbage. Even something as reliable as running pace, is affected by heat, humidity, elevation (even with elevation correction on), shoes, etc.
Don’t even get me started about recovery metrics.
Until the inputs are rock solid, the outputs will be similar.
Yeah. I plugged in a race several months away just to see how it would react, and you end up just repeating the same 3 weeks over and over.
The positives I took away from it were:
If you don’t know how to structure a week, it will help
You’ll get very good at z2 riding
If you like 30/30s you’re in luck.
It’s cheap to use and track with.
The negative:
Very redundant.
Most weeks, the only time you’ll do sustained efforts above z2 is when you do low cadence. You rarely/never hold race power at race cadence at any point.
Unless something has changed, it 100% defaults all VO2 workouts to 30/30s.
What I found, and why I ended up switching, is I could hold z2 and I could do 30/30s…but ask me to hold above z2 (thinking 70.3 pace or FTP), and I had no real ability to hold it because I just never trained at that power. I ended up having to swap to VO2 rides on Zwift to replace the 30/30s.
I switched to a human coach, who maybe just won an IM recently, and have seen immediate improvements.
I can agree with the lack of variety sentiment. I know that the best in the world do very basic workouts week in and week out, but to I feel like there has to be at least SOME variety eventually. It was very “polarized” in the sense that you were either at zone 2 or doing 30/30’s.
If you struggle to be objective and disciplined about your own training like me then go with a human coach. And by this I mean doing too much for your current life style. The older you get the easier this is to do I’ve learned the hard way.
If you don’t struggle with the above and you understand your hydration, nutrition and recovery thoroughly then I think you can still improve with a plan as long as you can be think outside the box if the situation presents itself also. Sickness, injury, family obligations, travel, etc.
No experience with AI except for watching all the Terminator movies.
I love the idea of an AI coach… if it’s not great today it should be better than most over the next year or two.
Imagine if your coach brought a cardiologist-level knowledge to your workout analysis, and evaluated all your workout files closely, then adjusted workouts to match.
They could far more intelligently evaluate your rest, blood O2, weight…
A huge probleme for AI is garbage in = garbage out, no matter how good your algos are.
So much data in tri is garbage. Even something as reliable as running pace, is affected by heat, humidity, elevation (even with elevation correction on), shoes, etc.
Don’t even get me started about recovery metrics.
Until the inputs are rock solid, the outputs will be similar.
Not to mention you need really quality training data for the models, even if you can build them, and almost no one has this. Strava, TP, Garmin … I guess Garmin is trying this, but my watches training program seems pretty trash to me.
Also in that approach (roughly 2 hrs per day average) you have to prioritize 56 hrs per week sleep (or minimally over 50) then it is sustainable assuming you already trained your body to get on the 700 hrs training per year plus 2600 hrs per year of sleep program. If you sacrifice too much on the sleep in training has to dial down. No free lunch without the recovery time