LT2 development

Dear Team Athletica,

First of all, I’d like to share my great experience last season – your software got me through my first long-distance triathlon in sub-11 on a quite hilly course, and through a middle-distance race in the leading group up front. Thanks to your software, the plan and I grew together; even difficult times could be managed by moving, deleting, or adjusting sessions.

As experience grows and I’m still recovering from the long-distance race, I’d like to prepare training for next season. I couldn’t get past the (sub-)threshold training method – VT2 sessions right at the sweet spot of stimulus vs. exhaustion make a lot of sense to me, especially given that long-distance race intensity sits close to threshold rather than in the VO2max zone.

Looking at my current base-phase plan (15–20h/week, load 1000–1200), I see a polarized approach with VO2max sessions consistently 2x/week (bike + run), but no VT2/threshold work in these two disciplines over 8 consecutive weeks. Swimming is the only discipline with a weekly VT2 session.

I asked the AI Coach about this, and it explained that VO2max in base phase is typically kept to “max 1x/week, often every 10–14 days” with threshold work increasing in build phase. However, this doesn’t match what I’m actually seeing in my calendar – 2x/week VO2max at bike and run consistently.

I also looked ahead at my build-phase weeks (late Nov/Dec) and can see race-specific sessions like “Long Distance Over-Unders” (targeting Z3a, the intended IM bike pace zone) and “HIIT + MIT” runs with sustained IM-pace blocks. I understand the logic of introducing race-pace specificity in build rather than base – that makes sense to me.

My question is more specific: independent of race-pace simulation, shouldn’t a generic VT2/threshold stimulus (not tied to IM-pace specificity, just LT2 development itself) also have a small, regular place during base – similar to how it’s already programmed weekly for swimming? The physiological rationale I’ve read (e.g. the Norwegian endurance model) suggests threshold adaptation benefits from being trained somewhat year-round, not only introduced once race-specific work begins.

Given my long-distance focus, would it make sense to alternate or replace one of the two weekly VO2max sessions (bike or run) with a generic VT2/subthreshold session during base? I’d appreciate understanding the reasoning behind the current structure, or guidance on how to best adjust it.

Thanks for your great work – looking forward to your input!

The VT2 part caught my attention. When you’re doing these threshold sessions, do you use any practical cues like breathing rhythm or the talk test to judge whether you’re sitting around LT2, or do you rely mostly on pace, power, and heart rate? I find the breathing side especially interesting once the effort gets close to threshold.

Close to VT1 for long distance, VT2 you will hold at most for 70 mins, but more likely 30 mins as an amateur

Fair point, Paul – I conflated VT2 with race intensity, which was imprecise. IM pace sits much closer to VT1, as you say.
My actual interest in VT2/subthreshold work during base isn’t about simulating race pace, though – it’s about using it as a stimulus to raise VT1/AeT itself. The idea (e.g. from Norwegian-style training) is that repeated, controlled work just below LT2 improves mitochondrial density and lactate clearance broadly enough that the whole aerobic curve shifts up, including the pace/power you can sustain at VT1 – which is what actually matters for long-course racing.

Does that reasoning hold up in your view, or is there evidence that VO2max stimulus achieves the same AeT-shifting effect more efficiently, without the added fatigue cost of threshold work?

Concerning the first answer about valid methods:

The speech test has solid validation as a method to determine your threshold zones and usually overlaps with RPE and HR quite well, or lets say good enough for most of us age groupers. Lactate testing can be a reasonable addition but is usually recommended for fine tuning.