Does Consistency Really Beat Intensity? We Went Looking for the Trial

The advice everyone repeats

Be regular first. Add volume second. Add intensity last. It sounds right, and it is probably roughly right for many people. But "probably roughly right" is not the same as "shown." So we went looking.

What the studies say about hard sessions

How many hard days? For trained athletes, about two high-intensity sessions a week is the usual target, according to reviews (Seiler 2010; Stöggl and Sperlich 2015).

Is more better? One small trial in elite cross-country skiers and biathletes (Tønnessen and colleagues, 2020) matched total training volume and compared four short interval sessions a week with two longer ones. Only the two-session group improved its time-trial performance (p = 0.04). It was a small study in elite athletes, so it hints and does not prove. But it does not support "more hard days is better."

What about recreational athletes? One six-week trial in 26 runners (Lenk 2025) compared one, two and three weekly interval sessions. Maximal aerobic capacity rose about 1.1%, 7.7% and 5.6%. The authors saw no clear extra benefit from a third session. The trial was small and the intervals were wide, so it does not show that a third session is harmful. It gives no firm reason to expect more from it either.

So the evidence supports "a couple of hard sessions is plenty" better than it supports "intensity is overrated." Neither says anything about how often to ride or run in total.

So where does consistency actually go wrong?

In our work with simulated athletes and published studies, we keep finding that consistency is lost to a short list of boring things.

1. A sudden big session

A cohort of 5,205 recreational runners followed for 18 months (Frandsen and colleagues, 2025) linked injury to how much longer one session was compared with the longest run in the past 30 days. Compared with sessions up to 10% beyond that longest run, going 10 to 30% beyond went with about 1.6 times the risk of an overuse injury. Beyond double the longest run, the risk was about 2.3 times.

Caveats: this is one observational study, injuries were self-reported, and it measures distance in runners. It shows an association, not that a rule prevents injury.

Meanwhile, the famous "10% a week" rule did not hold up in the best test of it. A trial of 532 novice runners (Buist 2008) found injury rates of 20.8% with the gradual programme and 20.3% with the standard one.

A practical reading: the size of one session relative to your own recent longest looks more useful than a weekly percentage.

2. A rule that cancels good days

Many apps turn a noisy morning number into a verdict. In our simulation of healthy, made-up athletes, the check-in rules we tested told people something other than "go ahead" on three days out of four, and cost 4 to 10% of weekly training load. A gentler design cut the loss to about 1.6%.

Your own heart rate variability moves 10 to 20% from morning to morning in good health (Buchheit 2014). A single low reading is weak evidence.

3. A plan that quietly sits below what you already carry

In one of our simulations, an engine delivered 99 to 103% of the plan it was given, and every simulated athlete still finished a 14-week season with lower fitness than they started. The plan itself sat at 70 to 93% of what it takes to maintain their fitness. Following it perfectly was a perfectly consistent decline.

Consistency matters, but only if what you are being consistent with is enough.

4. A gap in the menu

One simulated athlete stalled for weeks because the workout library had nothing between a 150-minute ride and a 3-hour one. Adding a single 165-minute ride moved their season-end fitness six points above the start. Sometimes inconsistency is a missing option, not a missing effort.

What you can do with this

These are our readings of the evidence, not proven rules.

• Anchor your long session to your own history. Stay close to your longest recent session, and think before going far beyond it.
• Use two or three mornings, not one. A single low reading is a prompt to check how you feel, not an order.
• Check that the plan is above maintenance if you want to improve. Easy weeks protect you only if some weeks stretch you.
• Keep hard days few and spaced. Two is a well-supported default. A third may help some people, and the evidence is weak.

What we are doing about it

We are improving the pace24 coaching engine on exactly these points, and testing each change on simulated athletes before it goes anywhere. The long-session gap above was found and fixed this way, and the engine now delivers its plan while we work on making sure the plan itself builds fitness. The newer Jev-plus-language-model reader, which handles athlete messages, is giving promising results in our tests.

What we cannot say

We found no trial that ranks frequency above volume above intensity. We found no study of the 10% session step in cycling or swimming. Our simulations describe rules, not outcomes. Real athletes, injuries and results need real studies.

The short version

Consistency is worth protecting. But the best protection looks less like willpower and more like avoiding a handful of avoidable breaks: the jump, the false alarm, the thin plan and the missing option.

Curious how pace24 plans around these? Get early access.