Polarised, Pyramidal or Sweet Spot — Which One Should I Use?

At 12+ hours/week, polarised tends to win for trained riders. At 6-8 hours/week, a pyramidal or moderate sweet-spot model usually delivers more — there isn't enough total stimulus in pure polarised. Which label you use matters less than this: most amateurs gain the most from making easy easier and hard harder, regardless of model.

Polarised vs Pyramidal vs Sweet Spot — The Honest Comparison

These three terms describe how training intensity is distributed across a week. Polarised: 80% genuinely easy, 20% genuinely hard, almost nothing in the middle. Pyramidal: lots of easy, moderate amounts of threshold, smaller amounts of VO2max. Sweet spot: lots of moderate-hard work at 88-94% of FTP, less easy time, less true VO2max. The internet's polarised-vs-sweet-spot debate has run for a decade. The honest answer changes with rider hours and training history.

Stephen Seiler's published distribution work shows that elite endurance athletes spend roughly 80% of training time below first ventilatory threshold and 15-20% well above the second. That's the polarised model. But Seiler himself has said on the podcast that pros are typically slightly more pyramidal than the headlines suggest — they hold more threshold time once form is built. Dan Lorang corroborated this on the show: at the Pogačar/Vingegaard end, the distribution is closer to pyramidal than the textbook 80/20.

For amateurs, weekly hours change which model wins. At 12+ hours per week, polarised usually delivers — there's enough total stimulus in the easy 80% that the hard 20% gets to be genuinely hard. At 6-8 hours per week, pure polarised often under-delivers; the threshold time in a pyramidal or sweet-spot model adds the productive stimulus that the shorter polarised week is missing. The Stoggl-Sperlich head-to-head studies and the FasCat/CTS coaching observations both line up on this. There isn't a religious answer — there's a hours-banded one.

What sweet spot does well: efficient productive stimulus when time is tight. What it does badly: it lives in the grey zone, and the recovery cost is high enough that easy days have to be genuinely easy or the rider cooks. The 'black hole' criticism of sweet spot is mostly about amateurs who do sweet spot AND let their easy days creep into tempo — that combination is what produces the chronic-fatigue plateau. Used carefully, sweet spot is a perfectly defensible amateur tool, especially in build phases.

The Roadman position: pick the model that fits your weekly hours and stick to it for at least 8-12 weeks. Polarised at 12+ hours, pyramidal at 8-12, sweet-spot-leaning at 6-8. Block periodise across the year — don't try to hit one distribution every week of every season. And remember: the percentage of effort that's actually easy in your easy time matters more than which label is on the model.

WHERE THIS COMES FROM

Stephen Seiler — polarised distribution work

Seiler's published research is the most-cited distribution evidence. Trained endurance athletes adapt more reliably to roughly 80/20 distributions than to threshold-heavy.

Dan Lorang — pyramidal observation on the podcast

Lorang's on-the-record framing: World Tour pros typically run more threshold than the textbook polarised model implies. Closer to pyramidal once base is built.

Stoggl & Sperlich 2014 — head-to-head distribution study

Foundational study comparing polarised, pyramidal, threshold, and high-volume distributions over 9 weeks. Polarised and pyramidal both produced larger VO2max gains than the threshold-heavy and high-volume groups.

FasCat / Frank Overton — sweet spot literature

FasCat's published case data on sweet-spot programming is the cleanest amateur-focused defence of the model — particularly for time-crunched riders.

Roadman polarised guide

The Roadman pillar on polarised training, with Seiler quotes and the amateur protocol.

COMMON FOLLOW-UPS

Is polarised better than sweet spot for amateurs?

It depends on hours. At 12+ hours per week, polarised usually wins. At 6-8 hours per week, a sweet-spot or pyramidal model often delivers more total productive stimulus. Both can work for amateurs — the model isn't the thing that breaks; the discipline of easy days is.

Do pros really train polarised?

Roughly. Distribution research shows pros sit close to 80% below VT1, but the remainder is more pyramidal than pure-polarised — meaning meaningful time at threshold, not just at VO2max. Both Seiler and Lorang have framed it this way on the podcast.

Is sweet spot the 'black hole' that ruins training?

Only when paired with too-hard easy days. Sweet spot in isolation is a productive zone — the trap is the rider who does sweet spot Tuesday, tempo-creep Wednesday, sweet spot Thursday, and is cooked by Sunday. The black hole isn't the zone, it's the week.

Should I switch models across the year?

Yes. The conventional periodisation across the Roadman network is: polarised in base phase, pyramidal or sweet-spot-leaning in build, polarised again in pre-event sharpening. Block periodise, don't try to hold one distribution every week of every month.