Preparing for an ATHX isn't training hard for three months.
It's improving three qualities that compete with each other — maximal strength, endurance, work capacity — while knowing which to prioritise and when. An athlete who puts aerobic volume everywhere caps their strength gains. An athlete who stacks heavy blocks loses their engine.
The good news is that ATHX is the only hybrid format that makes this planning possible: all three events are fixed and public for the whole season. You know exactly what you're preparing for, six months out.
Here's the structure.
First things first: know where you stand
Twelve weeks before your date, three tests. On three different days, cold, under competition conditions.
| day | test | what you record |
|---|---|---|
| D1 | Strength — the three lifts in their windows | all three loads and the total |
| D2 | Endurance — 3 km then max SkiErg, 24 min | run time, transition time, ski distance |
| D3 | Metcon X — the sequence, 25 min | total time, or reps at the time cap |
That's your baseline. Without it you won't know whether your block worked — and you won't know which pillar to prioritise either.
⚠️ Don't do all three on the same day. A test degraded by the previous one measures nothing.
The structure, in three blocks
Weeks 1 to 4 — the base
The aim is to build, not to perform.
- Strength: 3 sessions a week. Linear progression on squat, deadlift and shoulder to overhead. This is the time for heavy volume, while you're fresh.
- Endurance: 2 sessions. One long moderate aerobic piece, one SkiErg volume session. If you've never skied, this is where it gets built.
- Technique: twice a week, 15 minutes. Shoulder to overhead light, movement standards, burpees over a bench. It costs almost nothing in fatigue and it buys seconds.
Weeks 5 to 8 — specificity
You start to look like the event.
- Strength: 2 sessions, heavier loads, less volume.
- The event pairings: 2 sessions. Run-to-ski in full. Sandbag squats straight into the carry. Ground-to-overhead straight into the bench press. These are the format's real transitions.
- One work-capacity session a week: the full Metcon X, or long portions of it.
Week 6: retest the Endurance only. It's the zone that improves fastest, and the one that drives your pacing most. You adjust your run pace for the six weeks that remain.
Weeks 9 to 11 — execution
You stop building and learn to race.
- Week 9 or 10: one full-chain simulation. Strength at 90 %, then Endurance, then Metcon, with the real rest durations. The goal is execution, not the score.
- The rest: maintenance. One heavy strength session, two event sessions, low aerobic volume.
That simulation is where you find out what no isolated test shows: eating inside ten minutes of Refuel without nausea, bringing your heart rate down across thirty minutes of Recovery, arriving at the Metcon in the real state.
Week 12 — nothing
Taper. Volume halved, intensity kept, no testing.
What you haven't built won't get built in a week. What you can still lose is freshness.
The measurement calendar
One thing that often goes wrong: how often you test.
Testing your maxes every week cooks your nervous system for nothing. Measuring only once before the competition leaves no room to correct anything.
| when | what |
|---|---|
| 12 weeks out | all three blocks, cold, three different days |
| 6 weeks out | Endurance only |
| 4 weeks out | full-chain simulation, for execution |
| 2 weeks out | nothing |
Between those markers, you train. You don't test.
Prioritising: which pillar to work on first
Take your baseline and compare it to your category's targets. Three cases.
Your Strength is below the floor. That's your priority, and it's the longest job: eight to twelve weeks of structured work. Start there, from week 1.
Your Endurance is the weak link. It's the quickest to fix — six weeks of volume is enough to move the curve. Start with SkiErg volume if you've never done any: that's often where the deficit hides.
Your Metcon collapses under fatigue. A gap of more than 20 % between a cold version and a full-chain version signals a work-capacity deficit, not a technical problem. Train long blocks, not maxes.
⇒ And if your three numbers are even, work all three: this is the format that rewards having no hole most.
What generic plans don't handle
A twelve-week plan written in advance assumes you improve as planned.
You don't improve as planned. Your strength climbs faster than your endurance, or the other way round. A heavy week at work breaks a block. A minor injury takes you out of a movement for ten days.
A plan that doesn't know that has you following a structure that no longer matches your state.
That's what we built the ATmetrYX cockpit for. The Garmin app captures your sessions pillar by pillar; the cockpit stacks them. What appears is no longer a number, it's a trajectory: three curves, one per pillar, climbing at different speeds.
From there, reading it becomes obvious. The pillar that's improving gets maintained. The one that's lagging becomes the next block's priority. You're no longer guessing your weak point on instinct — your own measured progress points at it.
And the cockpit installs on your phone like an app: an email, a code, and it's open.
Where to actually start
If you're starting from scratch and your competition is three months away, the order is simple.
- Generate your plan. You enter your profile, your current loads and your event date, and you get a structure built on the 2027 format.
- Do your three tests in the first week, and enter your numbers.
- Install the app on your watch, so every session is captured without you writing anything down.
- Look at your curves at week 6, and adjust.
The four-week plan is free. The version that keeps adjusting across your whole preparation is in Premium — but start with the free one, it's enough to tell whether the method suits you.
In short
Twelve weeks, three blocks: build, specialise, execute. Four measurement markers, and nothing in between.
What makes ATHX trainable is that its three events don't move all season. You're not preparing for an unknown: you're preparing for a test you can repeat as many times as you want.
Might as well use that.