Most training apps hand you a plan and ask you to trust it. You get a workout for tomorrow, but no explanation for why that workout, why that distance, why today instead of Thursday. The algorithm is a black box. You’re supposed to follow along and hope it knows what it’s doing.
Pacewright works differently. Every workout decision is based on published, peer-reviewed research — and we tell you exactly what that research says and how it applies to your specific situation. This article walks through the actual principles that drive your training plan, and the research behind them.
Three Foundations
Your plan is built on three principles with published research behind them. The principles themselves are open: not proprietary, not secret. What’s ours is the optimization layer that weighs and combines them into your best next week, and we explain that too.
1. Polarized Intensity Distribution
In 2010, Stephen Seiler reviewed decades of research on how the best endurance athletes distribute their training intensity. The finding was consistent across runners, cyclists, cross-country skiers, and rowers: roughly 80% of training sessions should be at low intensity, with only about 20% at moderate-to-hard effort.
This is the polarized training model, and it’s counterintuitive. Running faster in races doesn’t come from running fast most of the time; the research points the other way. Athletes who spend most of their time running easy — genuinely easy, not “kind of easy” — consistently outperform athletes who spend more time in the moderate zone. The easy running builds your aerobic engine without accumulating excessive fatigue, and the hard sessions provide the specific stimulus your body needs to improve speed and lactate clearance.
A 2014 study by Stöggl and Sperlich compared polarized training directly against threshold training, high-intensity training, and high-volume training over 9 weeks. The polarized group saw the greatest improvements in VO2max and time to exhaustion.
How Pacewright applies this: Your plan targets roughly an 80/20 split between easy and hard effort. When you log a run, the app tracks where it fell on the intensity spectrum and adjusts future workouts to keep the distribution in the right range. If you’ve been running too many moderate efforts — the “gray zone” that feels productive but isn’t — you’ll see your next run shift to either genuinely easy or genuinely hard.
2. Training Load Management (How Much Is Enough)
Injury research points at two things. One is how much you run in total. The other is how big a single run is compared to what you have actually been doing lately, and that second one turns out to be the sharper signal.
In 2025, Frandsen and colleagues published the largest study of its kind: 5,205 runners and 588,071 recorded sessions. They looked for what separated the runs that preceded an injury from the runs that didn’t, and the answer was the size of the jump. A session that went more than about 10% beyond the runner’s longest run of the previous 30 days predicted overuse injury, in every band of excess they measured. Avoiding exactly that was the authors’ own recommendation to runners.
Three things do the protecting in Pacewright, and they’re in that order for a reason.
The single-session spike guard. No run in your plan jumps far beyond your longest run of the last 30 days. It’s the best-supported injury lever in runners, so it’s the limit Pacewright leans on hardest.
Weekly volume limits. How far your total mileage can climb from one week to the next, scaled to the mileage you’re already at. That’s the next section.
Your own feedback. How the run felt, the effort you rated it, whether you’re sore. Subjective reports move before the objective numbers do, so they get to override the plan rather than sit in a log.
How Pacewright applies this: Those signals combine into a single reading, DIAL — Dose In Adaptive Limits. It answers one question: are you training the right amount? DIAL places you on a single axis between your floor, the least running that still moves you forward, and your ceiling, the most your current state can absorb safely.
- Dial it up (room to grow). You’re under your floor with safe headroom, so the engine builds you toward more.
- Dialed in (productive). Enough to keep adapting, sustainable, inside your limits.
- Dial it back (ease off). You’re at or near your spike, volume, or fatigue limits.
DIAL sits next to RFI, your Run Fitness Index. RFI is how fit you are. DIAL is whether you’re loaded right.
Worth being plain about: DIAL describes, it doesn’t gate. It reads your dose and nudges the plan. The hard limits are the spike guard and the volume caps, and those hold no matter where the needle sits. You can see your DIAL in the app at any time, along with what moved it. No hidden math.
A note on honesty: what we took out. Pacewright used to manage this with the acute:chronic workload ratio, or ACWR, which weighs your recent training load against your longer-term average. It’s a popular number, and it’s in a lot of apps. We removed it, because it does not hold up.
The ratio is mathematically compromised: your recent load is part of the baseline it gets divided by, which manufactures a correlation on its own. Its own proponents concede that. When Impellizzeri and colleagues tested it directly, neither the ratio nor recent load on its own gave a meaningful predictive advantage over a model with no predictor in it at all, and dividing recent load by a contrived, randomly generated baseline reproduced much the same effect. Their conclusion was plain: “We suggest ACWR be dismissed as a framework and model.”
In runners the problem is worse than weak. The relationship runs backwards. Nakaoka and colleagues followed 435 Dutch recreational runners in 2021 and found the highest injury probability at the low end of the ratio, around 9.6% below 0.70 against roughly 1.3% above 1.38. That study tested all three ways of calculating the ratio, and the variant Pacewright had been computing, the exponentially weighted one, was the only one that came back sparse and non-significant in those runners. That is one cohort in one country, so we won’t call it settled on its own. Frandsen’s 2025 study, the largest and newest, points the same direction: the ratio showed a negative dose-response relationship with first-occurring, self-reported overuse injury, and the week-to-week version showed no relationship at all, while the size of a single session’s jump was what stood out. So the widely quoted “safe zone” for that ratio is not a validated safe zone, and you won’t find it in Pacewright.
We’re not replacing one false certainty with another. Most of the ACWR evidence base comes from team sports rather than running. The inverse findings in runners may partly reflect reverse causation, since a runner who is already hurting runs less, which drags the ratio down on its own. And whether a deep training base is independently protective is plausible but under-tested. What we can say is that the ratio was never the validated guardrail it got presented as, and that the per-session jump is better evidenced in runners than the ratio ever was.
3. Progressive Overload With Mileage-Dependent Caps
Your body adapts to stress — but only if you increase that stress gradually and give yourself time to recover. This is progressive overload, and it’s the oldest principle in exercise science. Banister’s foundational 1991 model showed that fitness and fatigue respond to training load on different timescales: fitness builds slowly and persists, while fatigue accumulates quickly and dissipates quickly. The art is in managing the balance.
The common advice is “don’t increase your weekly mileage by more than 10%.” It’s a reasonable starting point, but it’s also a blunt instrument. 10% of 10 miles is 1 mile — a trivially small increase. 10% of 60 miles is 6 miles — a substantial jump that can push a high-mileage runner into risky territory.
Pacewright uses mileage-dependent volume caps that scale with where you are. The more you’re already running, the smaller the share you’re allowed to add on top — a beginner has real room to grow, while a high-mileage runner is held to a much narrower step. There’s a stability bonus too: hold the same volume for several weeks and you earn extra headroom on your cap. Consistency is rewarded because a stable training base handles increases better than a volatile one.
The spike guard from the last section applies on top of this. Even when your weekly cap allows more volume, Pacewright won’t pour that room into one run.
The cap also tightens when you’re returning to training after a long break. If you’ve been away for two weeks or more, the weekly increase is held to a much smaller fraction for the first two to three weeks back. The reason is mechanical: cardiovascular fitness recovers faster than connective tissue, so your lungs feel ready before your tendons do, and the standard caps are too aggressive for that mismatch. The tighter ramp closes that gap and prevents the comeback injuries that show up two to three weeks into a return.
What Happens When Your Plan Updates
When Pacewright generates your next workout, it’s not pulling from a library of pre-built plans. The optimizer makes a real-time decision based on your current state, picking the best-scoring option within hard safety limits. Here’s what it weighs:
Your training load history. What you’ve been running this week, this month, and how that compares to your established baseline. This is what your DIAL reads, and what the spike guard and the volume caps check against.
Your goal event. A marathon training plan looks fundamentally different from a 5K plan. The principle of specificity means your workouts need to target the energy systems and distances that match what you’re training for. A marathon plan emphasizes long aerobic runs and sustained tempo work. A 5K plan includes more interval work and shorter, faster sessions.
Your intensity distribution. Has your recent running been too heavily weighted toward moderate effort? Too many hard days in a row? The plan adjusts to keep the polarized distribution in the right range.
The weather forecast. Heat and humidity affect your body’s ability to perform; a humid day with a high dew point slows the pace you can hold at a given effort. Pacewright adjusts target paces based on current conditions so you don’t chase a pace that’s unrealistic for the weather.
Your schedule. Rest days, available time, the structure of your week — your plan fits your life, not the other way around.
What happened recently. If you missed a workout, Pacewright doesn’t just shove it into tomorrow. Some workouts get rescheduled, some get dropped, and some are never “made up” because doing so would spike your load. There are specific rules for this, and you can read them in our article on missed workout handling.
Your race, once it’s close. Heading into a goal race, the optimizer brings your volume down while holding on to your intensity and how often you run, so you arrive at the start line fresh rather than tired. That’s the taper, and it’s driven by the same optimizer as everything else — how early it starts and how far the volume comes down depend on your race and on how your training has been going. It isn’t a fixed script. Even a brand-new runner’s walk-run start works this way: it leads with easy intervals and lengthens them as your body adapts, not on a set schedule.
Every Decision Comes With a Reason
This transparency is a deliberate part of how Pacewright works, not an afterthought.
Every time Pacewright modifies your plan, you see three things:
- What changed. “Today’s run is now 4 miles instead of 5.”
- Why it changed. “Five miles would have jumped well past your longest run of the last 30 days.”
- What data drove the decision. “Based on your longest run in the past 30 days and the mileage you’ve logged this week.”
This isn’t a decorative feature. It’s the core of how Pacewright works. If you disagree with a decision, you have enough information to understand what the app is seeing and why it’s responding the way it is. You’re not guessing. You’re not just trusting a brand. You’re looking at the same numbers the algorithm is looking at.
The training science isn’t ours; it belongs to the researchers who published it. What we built is the optimization layer that applies it to your specific situation and then tells you exactly what it did and why. That part is ours, and every decision it makes is explained, with the underlying math visible.
Your plan is built on real research, and we show our work.