Most runners who get injured aren’t doing anything obviously reckless. They’re not sprinting into a wall or doubling their mileage overnight. They’re adding a few extra miles here, an extra hard session there, maybe jumping back in after a week off with the same workouts they were doing before the break.
The injuries come from a pattern, not a single mistake. The question is which pattern. A lot of training apps answer that question with a number that doesn’t hold up in runners, so this page lays out what Pacewright actually watches, what it does when something looks off, and where the science is solid versus where it isn’t.
The Single Big Jump
The best runner-specific evidence points at one thing: an individual run that lands well beyond what you’ve been doing lately.
The largest and newest study on this followed 5,205 runners across 588,071 recorded sessions, and it asked the question directly in its title: how much running is too much? The answer wasn’t weekly mileage. It wasn’t a ratio of this week to last month. It was the size of a single session compared to the runner’s longest run in the previous 30 days. Sessions that went more than 10% past that longest recent run carried a clearly elevated injury rate, and every band above that threshold carried it. Hazard rate ratios came in at 1.64 for small spikes of 10-30%, 1.52 for moderate spikes of 30-100%, and 2.28 for runs that more than doubled the recent longest.1 The authors’ own advice to runners was to avoid exactly that jump.
This is the guardrail Pacewright leads with. No run it prescribes will jump far beyond your longest run in the recent past.
Say your long run has been sitting at 8 miles for the past month. You have a good week, you feel strong, and 12 miles sounds reasonable. It isn’t, and the plan won’t offer it. It’ll step you up from 8, then step up again from there, and the ceiling moves with you as your longest recent run grows. The jump is what hurts you, not the distance itself. A 12-mile run is a fine thing to do once 12 miles is close to what you’ve already done.
This is also why the guardrail tightens on its own when you’ve been away. Your longest recent run is a rolling window, not a personal record you keep forever. Take a few weeks off and that window empties out, so the cap comes down with it, automatically.
Weekly Volume Has Limits Too
The single-session cap governs any one run. Your weekly total gets its own limit.
Pacewright caps how much your weekly mileage can climb from one week to the next, and the cap scales to how much you’re already running. A beginner adding a couple of miles to a small week is doing something ordinary. A high-mileage runner adding the same percentage is absorbing an entire extra session’s worth of stress on an already-loaded body. A flat percentage treats those as the same event. They aren’t. Volume caps covers how the tiers work, along with the long-run limit and the planned recovery weeks that stop you from building indefinitely.
One honest note on this. The weekly cap is a conservative progression heuristic, not a validated injury line. The famous flat “10% per week” rule didn’t survive being tested, and nobody has established a precise safe rate of increase to replace it. Pacewright’s caps are a deliberately cautious default, chosen because the cost of ramping slightly too slowly is small and the cost of ramping too fast is an injury. They aren’t a line where safety switches off at 1% over.
Coming Back From a Break
The most common way a reasonable runner gets hurt is picking up where they left off.
You ran 30 miles a week for two months. Then you traveled, or got sick, or life simply happened, and for two weeks you barely ran at all. Your head still thinks of you as a 30-mile runner. Your tendons, bones, and connective tissue have already started to give that back.
Pacewright doesn’t drop you into your old plan. It rebuilds you, and the mechanism is the one already described above rather than a special case: your longest recent run has shrunk, so the single-session cap has shrunk with it, and your recent weekly volume is low, so the weekly cap starts from where you actually are instead of where you were. You’ll see a reduced plan with a clear path back to your previous volume.
Your Own Feedback
The third guardrail is the one you supply.
After a run you can tell Pacewright how hard it felt, whether anything hurt, and how you’re recovering. That’s the only input the engine can’t derive from a GPS file, and it’s the one that overrides the rest. A run that looks fine on paper (right distance, right pace, inside every cap) can still be the wrong run for you on that day, and the plan should respond to that. Reported effort, soreness, and how you say you felt all feed back into what comes next.
This isn’t a courtesy feature. Subjective load is the adaptive part of an adaptive plan. If you tell it the easy runs are wrecking you, that means something the mileage doesn’t say.
How Training Load Is Calculated
All of this needs a way to measure “how much” you trained, not just “how far” you ran. Five easy miles and five hard miles aren’t the same stress on your body.
Pacewright calculates training load as:
Duration (minutes) × RPE (1-10)
RPE is your Rate of Perceived Exertion, meaning how hard the session felt on a scale of 1 to 10. A 45-minute easy run at RPE 3 produces a training load of 135. A 45-minute tempo run at RPE 7 produces a load of 315. Same duration, very different stress.
This approach is called session RPE, and it was validated by Foster and colleagues.5 It’s a deliberate choice. Duration × RPE works without a GPS watch, without a heart rate monitor, without any gear at all. It accounts for things that pace and heart rate miss: heat, sleep quality, accumulated fatigue, stress at work.
It’s also why your feedback carries real weight. The load number has your honest answer built into it.
What About ACWR?
If you’ve read much about running injuries, you’ve met the acute:chronic workload ratio. Your training over the past 7 days divided by your rolling average over the past 28. The idea is that the ratio tells you whether you’re ramping up faster than your body is prepared for, and most training apps treat some version of it as the safety number.
Pacewright doesn’t. Here’s why, because you should be able to check this yourself.
There’s no validated sweet spot. The often-quoted “0.8 to 1.3 is safe” band traces back to a systematic review that found, in its own words, a trend toward that range showing the lowest injury risk. Look at the actual estimate and the uncertainty around it is wide enough to be consistent with that range helping, doing nothing, or hurting.9 The authors of the most recent review, pooling 22 cohorts in 2025, land on “use with caution.”11 That is a long way from a safety threshold you’d let an app enforce on you.
The math is compromised, and that part isn’t controversial. Your acute load is part of the chronic average you’re dividing it by. The numerator sits inside its own denominator, which manufactures correlation on its own. This isn’t a fringe objection. The ratio’s own proponents have acknowledged the coupling problem in print.10
It doesn’t add predictive value. Impellizzeri and colleagues showed that the ratio is essentially a rescaling of your recent load. Swap in a fake chronic denominator, even a random one, and you reproduce nearly the same injury odds. Neither ACWR nor recent load alone beats a null model.4 In a 2021 follow-up the same group concluded, in their words, “We suggest ACWR be dismissed as a framework and model.”3 The 2016 and 2014 papers that made the ratio famous were team-sport work, on rugby players and cricket bowlers, carrying collision loads rather than endurance loads.6, 7
In runners it doesn’t just weaken. It inverts. A prospective cohort of 435 Dutch recreational runners found the relationship ran backwards and L-shaped: the highest predicted injury probability, 9.6%, sat at an ACWR below 0.70, while runners above 1.38 came in around 1.3%.2 The 5,205-runner study reported the same direction: a negative dose-response for the ratio, meaning more of it went with fewer injuries, not more. The week-to-week ratio showed no relationship at all.1 In the runner data, the “danger zone” wasn’t dangerous and the “safe zone” wasn’t safest.
The “improved” version isn’t an upgrade. Weighting recent days more heavily, using an exponentially weighted moving average, was proposed as a better way to calculate the ratio.8 It’s the version training apps tend to reach for, and it sounds more rigorous. In the only head-to-head comparison in runners, it produced the weakest and sparsest results of the lot.2 It is not a more sophisticated calculation, and we won’t describe it as one.
So ACWR is demoted. It is not a guardrail here, it is not a safe zone, and you won’t see 0.8 to 1.3 presented as a validated band anywhere in this app. A hard rule that eased your training to pull the ratio down could push you toward exactly the low-ratio state that carried the highest injury probability in the runner data.
Now the honest half, because replacing a false certainty with a new one is the same mistake wearing different clothes. Most ACWR research is team-sport research, and that cuts both ways. The inverse finding in runners may partly reflect reverse causation, where runners who are already breaking down back off and show a low ratio as a result rather than a cause. Whether a well-developed base is independently protective in runners is biologically plausible and genuinely under-tested. None of that rescues the 0.8 to 1.3 guardrail. It just means we hold the replacement to the same standard, and the single-session finding is where the runner evidence is strongest right now.
DIAL: Are You Training the Right Amount?
Everything above answers “is this too much?” That leaves the other half of the question, and it’s the half most runners actually lose sleep over. Am I doing enough?
DIAL answers that one. It stands for Dose In Adaptive Limits, and it’s one number that places your current training on a single axis between two points:
- Your floor, the least you can do and still improve. Below it you stop adapting and settle into a rut.
- Your ceiling, the safe maximum, set by the guardrails above.
DIAL is the number the gauge shows. It pairs with RFI, your Run Fitness Index. RFI tells you how fit you are. DIAL tells you whether you’re loaded right.
DIAL is a read, not a rule. This distinction matters and we don’t want it blurred. DIAL describes your dose and nudges your plan. It never gates a workout. The single-session cap, the volume caps, and your feedback are what actually hold the line, and they do it whatever DIAL happens to say. A number that tells you how much you’re doing and a mechanism that stops you doing too much are two different jobs, and Pacewright keeps them separate on purpose.
Two more honest notes. The floor is the softer side of the science. The evidence for how much is enough to keep improving is thinner than the evidence for what’s too much, so DIAL puts its firmer conclusions on the ceiling side and treats “you could do more” as a gentle suggestion, never a push past safety. And DIAL is a lagging signal that reads across weeks, not days. Early on, before it has enough of your history, it says so instead of guessing.
The Three Zones
Dial it up
You have room to grow. Your current load sits below your effective floor with safe headroom above it, your fitness has gone flat, and the runs are coming back easy. That combination usually means you’ve plateaued, not that you’re being careful. Pacewright will nudge your plan upward toward your goal, inside whatever you’ve told it about your week. If you’ve set your own running days, this stays a read on the dial and nothing moves without you.
Read this one as a multi-week observation rather than a verdict on today. Holding steady isn’t a failing.
Dialed in
You’re in the productive band. You’re doing enough to keep adapting, and you’re doing it inside your limits. This is where you want to spend most of your training. The plan proceeds.
Dial it back
You’re at or near the edge. That might be proximity to the single-session cap, weekly volume climbing near its limit, or the fatigue side of the picture: effort trending up, soreness, or you telling the app the runs are too hard. Pacewright pauses the build and eases off.
Easing off is temporary and it recovers on its own. It doesn’t lower the target you’re building toward.
Before the zones: establishing your base
If you’re new or coming back, there isn’t enough history to place you on the axis yet. Rather than show you a number built on nothing, Pacewright says it’s still establishing your base. You’ll get the number when it means something.
What You Can See in the App
Pacewright doesn’t hide the math. Your DIAL and its zone read out on the gauge. The Today screen carries a training status card. And every workout comes with the reasoning behind it, including whether you’re doing enough and whether it sits inside your limits.
When a guardrail catches something, the explanation is specific rather than a shrug:
“Saturday’s long run has been adjusted from 12 miles to 10. Your longest recent run is 9.5 miles, and 12 is a bigger single-session jump than your plan allows. Your long run will keep stepping up from here.”
The numbers behind every change are yours to examine.
The Short Version
Your safety comes from three things. No single run that jumps far beyond your longest recent run, which is the strongest injury signal in the runner research. Weekly volume limits, held deliberately conservative because nobody has established a precise safe rate. And what you tell the app after you run, which outranks the rest.
DIAL sits alongside those and answers a different question: whether the amount you’re doing is too little, about right, or too much. It’s the number on the gauge, and it’s a description of your training rather than a rule over it.
ACWR isn’t the safety number. It was never validated as one, and in runners it points the wrong way often enough that we won’t build your safety on it. We’d rather show you the mechanism, name the studies, and let you check our work.