Hey all,
I use Calendar sync to send my Projected Training plan from Trainer road to Intervals.ICU and over the next 6 months My plan shows i will have relatively minimal fitness gain. I’m new to Trainer road so not sure if this is just a difference in philosophy between the two sites or maybe the AI adaption takes a while to catch up?
Since the TR plan changes over time based on how you are progressing, the future workouts are a bit of an unknown. The “default” plans are extremely conservative by design.
You are misunderstanding. Intervals.icu does not in any way try to predict your future performance.
The “fitness” on Intervals.icu is basically a copy of Chronic Training Load (CTL) developed by Coggan. It’s not telling you how much FTP or other performance improvement you will have. It’s only telling how how much average TSS per day you’ve maintained over the last 6 weeks (usually weighted heavier for the most recent weeks). It’s a gauge of the combination of intensity and volume of training you’ve done.
However, you can do a ton of training (high CTL) and have very little or a lot of fitness gains. There’s no way that a formula can predict that, especially with how little information is used to calculate TSS. All that future portion of the CTL graph is doing is calculating what CTL will be, assuming you follow your training plan exactly and also don’t do any training not in your training plan.
In addition, CTL (training load averaged over time), is only part of the story. You can get really fit but be too tired (fatigued) on race day to perform well. So “fatigue” is calculated in basically the same manner as Coggan’s Acute Training Load (ATL). This measures how tired you are, based on the last 7 days TSS.
So your ability to perform, relative only to yourself, is supposed to be optimized when your “fitness” is high AND your “fatigue” is low. So you’ve done lots of training but let your body rest up recently so you are fresh.
This all assumes that your FTP is accurate, you’ve recorded all power data, you don’t have significant stress beyond cycling (weight lifting or life stress), you fuel your workouts well and you recover well.
The giant caveat with all this is that it also assumes all TSS has pretty much equal impact on your fitness. But fitness isn’t just one thing. If you are focused specially on 40 km time trialing your event is going to consist of one long effort near threshold. So you might decide to train mostly at sweet spot or threshold. That will give you a nice high CTL and it will look like you’re getting really fit if you only focus on that.
But if you only do that, that training will become less and less effective for you. Your body needs different stimulus to get good gains. You eventually will need to do some VO2max training, which will be low TSS since the workout lengths are short, and result in a lower CTL, but in fact you will become fitter. TSS is not a measure of training effectiveness, only an approximation of training load.
And, because TSS is always calculated based on your FTP at the time of your workout, it doesn’t factor in gains in FTP or other parts of the power curve.
So long story short, that chart is just an estimator of training load, short and long term, not an estimator of training results. Obviously, training load is an important factor in fitness, but there’s a whole lot more to it than that.
I get what you’re generally trying to say, but I think saying “having flat CTL from January to July is no big deal” is a step too far.
It was essentially what I saw happening with trainerroad late 2021-2023, a CTL that hovered between 48-56 during an entire base-build-specialty cycle unlike with Join (in fact usually during speciality my ctl would drop as I couldn’t keep up with the ever increasing intensity, ignore that recent huge dip, that was a recovery week followed by the flu 
)
It doesn’t surprise me to see a small drop in Specialty, because that’s not really the focus, but certainly during Base and Build, I would want to see the base shoring up the foundation and the build building.
Best answer ever on this platform.
