How Stridewise works

Training load is a single number for how much a session took out of you. A gentle 30-minute jog and a hard hour of intervals both count as training, but not equally. Load puts them on one scale so they can be added up across days and weeks.

Stridewise works out each session’s load from the best evidence it has, in this order:

1. Heart rate, when the heart rate trace passes the quality checks (see checking heart rate data). This gives TRIMP and hrTSS.
2. Your effort rating, when heart rate is missing or unreliable. This gives session RPE.
3. An estimate from duration and pace, when neither is available, so a run with no data still counts for something.

All three are expressed in the same points as hrTSS, where one hour at your threshold scores 100. Effort ratings and estimates are first turned into an intensity factor: how hard the session was compared with threshold, where 1.0 is threshold effort.

• From an effort rating, the intensity factor is 0.55 + 0.064 × RPE. An RPE of 7 comes out at about 1.0.
• For an estimate, it is your threshold pace divided by your grade-adjusted pace, kept between 0.5 and 1.15. With no pace data, a typical easy-run value of 0.75 is used.

Other activities, such as rides or gym sessions, count towards load only when they have heart rate or an effort rating. Each run shows which source was used and the numbers behind it.

Converting between heart rate load and effort ratings is an approximation. They measure related but different things: one is how hard your heart worked, the other is how hard it felt. A week that mixes sources is less precise than a week measured one way.

TRIMP (training impulse) combines how long you ran with how hard your heart was working. Time spent at a high heart rate counts for much more than the same time at a low one, because blood lactate rises steeply, not in a straight line, as effort climbs.

Here t is the duration in minutes and x is the fraction of heart rate reserve. The published constants are a = 0.64 and b = 1.92 for men, and a = 0.86 and b = 1.67 for women. They came from the shape of the blood lactate curve.

Stridewise calculates TRIMP second by second through the heart rate stream, so a session of hard reps and easy jogs is not treated like a steady run at the same average. If only an average heart rate is available, it uses that instead. Stridewise does not ask for your sex, so it uses one set of constants for everyone, roughly midway between the two published sets: a = 0.75 and b = 1.8. TRIMP is the basis of hrTSS, which is the number that goes into the fitness model.

• It depends on accurate maximum and resting heart rates. If those are off, every TRIMP is off in the same direction.
• The weighting constants were derived from small groups and do not fit every individual.
• Heart rate lags behind effort, so very short, very hard efforts such as strides and hill sprints are undercounted.
• Heat, dehydration, caffeine and illness all raise heart rate without the training being harder.

hrTSS (heart rate training stress score) scales a session’s load so that one hour at your lactate threshold scores 100. That gives numbers that are easy to read: an easy hour might score 50 to 60, a hard hour around 100.

hrTSS is the main daily load figure behind fitness, fatigue and form and the acute:chronic workload ratio. Anchoring it to your lactate threshold heart rate makes scores comparable between runners with different heart rates. If no reliable threshold heart rate is known yet, Stridewise assumes it sits at 88% of your heart rate reserve.

The training stress score idea comes from Andrew Coggan’s work with power meters in cycling, and TrainingPeaks introduced the heart rate version. TrainingPeaks’ exact implementation is not published, so Stridewise uses a common approximation based on TRIMP. Scores are close to, but will not exactly match, other apps. It also inherits every limitation of TRIMP and relies on a sensible threshold heart rate.

RPE is your rating of perceived exertion: how hard the whole session felt, on a scale of 1 (very easy) to 10 (maximal). Carl Foster showed that multiplying this rating by the session’s duration gives a load figure that tracks heart rate methods well, and it works for any activity, with or without a heart rate monitor.

After each synced run, Stridewise asks for your RPE. It is used in three ways:

• As the load figure when heart rate is missing or unreliable, and for strength sessions.
• Compared with the RPE the plan expected for that session. For a run with no planned session, the expected range comes from its heart rate: the zone you spent most time in sets it, and 15% or more of the time in zones 4 and 5 marks it as a hard session. Without reliable heart rate there is nothing to compare against. A run that felt notably harder than expected is an early sign of fatigue and feeds the adaptation rules and readiness.
• To calculate monotony and strain.

Foster’s method asks for the rating about 30 minutes after finishing, once the last few minutes no longer dominate how it felt.

It is subjective. Mood, sleep and how the last kilometre went all colour the rating, and people use the scale differently. It is most useful compared with your own earlier ratings rather than with anyone else’s.

Eric Banister’s impulse-response model treats each training session as having two effects. It builds fitness, which lasts a long time but grows slowly, and it causes fatigue, which appears quickly and fades quickly. How well you can perform on a given day is roughly your fitness minus your fatigue.

This explains why you can feel flat during a heavy block and then race well after a taper.

w(s) is the load on an earlier day s. The fitness time constant τ₁ is long (Banister’s runners fitted around 40 to 50 days) and the fatigue constant τ₂ short (around 10 to 15 days). k₁ and k₂ scale how much each matters.

Stridewise uses the widely adopted simplified form of this model, with fixed time constants and no fitted gains. That is what produces the fitness, fatigue and form numbers.

• In research the constants are fitted to each athlete from repeated performance tests. Stridewise does not have those tests, so it uses standard values that will suit some runners better than others.
• The model knows nothing about injury, sleep, stress or the type of training, only the amount.
• Fitted models predict past performance much better than future performance.

• Fitness (often called CTL, chronic training load) is a 42-day exponentially weighted average of your daily load. It rises slowly with consistent training.
• Fatigue (ATL, acute training load) is the same calculation over 7 days. It jumps after a big week and drops after a few easy days.
• Form (TSB, training stress balance) is fitness minus fatigue. Negative form is normal while building. Positive form means you are fresh, which is what a taper aims for.

ATL uses the same formula with 7 in place of 42. Form is worked out from the values at the start of each day, before that day’s training.

• The rate at which fitness is allowed to rise each week caps how quickly the plan increases (see progression).
• Form feeds readiness. Form below −25, or below −30% of fitness when that is lower, triggers the overreaching rule in the adaptation rules.
• The taper cuts volume so that fatigue falls faster than fitness and form rises towards race day.
• The trends screen charts all three.

These are bookkeeping numbers, not measurements of your physiology. “Fitness” here means recent training load, not aerobic capacity, and two runners with the same number can be in very different shape. The 42 and 7 day windows are conventions, not values fitted to you.

The acute:chronic workload ratio compares your last week of training with your last month or so. A ratio of 1.0 means this week matches your recent norm. A ratio well above 1.0 means a sudden jump.

Research in cricket, rugby league and other team sports reported that sudden spikes were associated with more injuries, and a “sweet spot” of about 0.8 to 1.3 became popular. It is also one of the most criticised measures in sports science, and that criticism shapes how Stridewise uses it.

The ratio is the acute average (N = 7 days) divided by the chronic average (N = 28 days). Williams and colleagues proposed this exponentially weighted version because the original rolling-average version treats a session 27 days ago the same as yesterday’s, and drops it abruptly on day 29.

Stridewise uses the ratio as one soft signal among several. It contributes to readiness, and a ratio above 1.3 is one of the triggers for the overreaching rule, which adds recovery and caps the week’s load. It is never used as an injury prediction, never shown as a risk percentage, and never the only reason for a change. The planner also avoids creating spikes in the first place by capping how fast load can rise. The ratio is not shown at all until there are 28 days of history, or while recent load is very low, because it is meaningless then.

• Mathematical coupling. In the original version the acute week is also part of the chronic window, so the two numbers are not independent. Lolli and colleagues showed this alone can create correlations that look meaningful but are artefacts of the arithmetic.
• A ratio hides the totals. 20 km after 10 km gives the same ratio as 100 km after 50 km. Ratios also behave badly when the chronic number is small, as it is for new or returning runners.
• Weak evidence of cause. Impellizzeri and colleagues argued that the ratio has conceptual problems, that the original studies used flawed analysis, and that there is no good evidence that managing the ratio prevents injury. The “sweet spot” numbers have not held up as universal thresholds.
• Most of the research was in team sports, not recreational distance running.

The underlying idea, that sudden large increases in training are risky, is still sensible. The ratio is just a crude way to spot them, which is why Stridewise does not lean on it.

Monotony measures how similar your days are. If every day carries about the same load, with no truly easy days, monotony is high. Strain combines monotony with the total for the week. Foster found that illness and poor performance tended to follow weeks of high strain, and that varying the load from day to day helped.

Stridewise calculates both over a rolling 7-day window and shows them on the trends screen. If every day is identical, the standard deviation is zero, so monotony is capped at 10. The planner keeps monotony down by design: hard days are separated by easy days or rest, and easy days are genuinely easy. A monotony above about 2, the level Foster highlighted, is a prompt to look at whether easy days are drifting too hard.

Rest days count as zero load, so a runner who trains every day can show high monotony even with a sensible plan. The thresholds come from small studies of well-trained athletes. Treat these as prompts for reflection rather than rules.

Heart rate zones split your working range into five bands, from very easy (zone 1) to flat out (zone 5). Stridewise builds them from your heart rate reserve: the gap between your resting and maximum heart rates. This is the Karvonen method. It fits individuals better than simple percentages of maximum, because two runners with the same maximum can have very different resting rates.

By default, zones are percentages of your heart rate reserve. When a reliable, plausible lactate threshold heart rate is known, Stridewise switches to percentages of that instead, because threshold is a more individual anchor than maximum.

Zones are used for time-in-zone charts on each run and the easy and hard split. Planned easy runs carry a heart rate cap at 70% of heart rate reserve, and recovery runs a little lower at 65%.

Zones are only as good as the maximum and resting values behind them. Heart rate also drifts upwards during long or hot runs, so the same pace can move you up a zone over time. That is why easy runs use a cap as a guide, not a hard rule.

Lactate threshold heart rate (LTHR) is roughly the highest heart rate you can hold for about an hour in a race. Above it, fatigue builds quickly. Joe Friel popularised a field test: run 30 minutes alone as hard as you can sustain, and take the average heart rate of the last 20 minutes.

You do not need to do a test. Stridewise looks through the last 365 days of runs with reliable heart rate for:

• Races lasting 20 to 130 minutes. It takes the average heart rate of the final 20 minutes and adjusts it for the race length: you race a 5K a little above threshold and a half marathon a little below.
• Hard training efforts: steady 20-minute stretches above 80% of heart rate reserve. Training is rarely all-out, so these only set a lower bound until you race.

Several races that agree give high confidence. The estimate is shown with its confidence label and how it was found. It anchors hrTSS and, when confidence allows, your zones.

An effort that was not truly all-out, a hilly route or a hot day will all skew the estimate. The adjustment from race heart rate to threshold is an average and will not fit everyone exactly. With no suitable efforts, Stridewise falls back to heart rate reserve alone and marks the confidence as low.

Maximum heart rate is the highest your heart will beat during an all-out effort. Resting heart rate is your lowest, best measured first thing in the morning. Both are personal and are not a measure of fitness on their own.

If you enter your own values, Stridewise uses them. Otherwise it estimates them from your runs. Maximum comes from the highest heart rate in races and hard sessions, averaged over 15 seconds, with faulty recordings excluded. Resting comes from your lowest steady heart rate while running, because Strava does not record resting heart rate. Each value is shown with where it came from (entered by you, measured or estimated) and a confidence label (high, medium or low).

With no usable heart rate data, maximum falls back to the age formula 208 − 0.7 × age (Tanaka and colleagues) when your age is known, or a placeholder of 185 beats per minute. Resting falls back to 60. These are marked low confidence: age formulas can be wrong by 10 beats per minute or more for an individual.

The highest heart rate in your runs is only your maximum if you have actually run flat out recently. If you have not, the estimate will be too low, and your zones will be too low with it. A single spike from a faulty sensor could do the opposite, which is why spikes are filtered first (see checking heart rate data).

Heart rate sensors, especially optical ones on the wrist, often record nonsense. Stridewise checks every heart rate trace before trusting it, and if the trace fails, it uses your effort rating for load instead and says so on the run.

It looks for these problems:

• Dropouts: more than 5% of the run missing, or long gaps.
• Spikes: readings more than 20 beats per minute away from the surrounding 30 seconds, happening repeatedly.
• Cadence lock: a wrist sensor picking up your step rate instead of your pulse. Stridewise flags it when heart rate stays within 3 beats of your step rate for 85% of a 5-minute stretch.
• Flatline: the same value for 2 minutes or more.
• Implausible values: readings more than 8 beats above your known maximum for a sustained period, or a first 10 minutes reading 15 or more beats higher than later in the run at the same or faster pace, a common optical sensor fault.

These are rules of thumb, not a medical-grade analysis. A real but unusual heart rate, for example from a heart rhythm condition, could be flagged as faulty, and some subtle errors will get through. If you have any concerns about your heart rate, speak to a doctor.

Grade-adjusted pace (GAP) is the pace you would have run on the flat for the same effort. Running uphill costs more energy, so a slow uphill kilometre can be as hard as a quick flat one. Gentle downhills save energy, but steep ones stop helping because you have to brake.

C(i) is the energy cost in joules per kilogram per metre and i is the gradient as a fraction (0.05 for a 5% climb). Minetti and colleagues measured it on a treadmill at slopes from −45% to +45%.

Stridewise calculates GAP second by second from the gradient stream for every run and shows it per kilometre split. Gradients steeper than 25% either way are treated as 25%, where the equation and the GPS data both become unreliable. GAP, rather than raw pace, is used for efficiency factor, decoupling and judging whether you hit pace targets on hilly routes.

• GPS altitude is noisy, so short-range gradients are smoothed and still imperfect.
• The equation describes steady running on a smooth treadmill. Trails, mud, steps, wind and tiredness are not included.
• It is an energy cost model. Steep downhills also cost muscle damage that the equation does not see.

Efficiency factor (EF) is how much speed you get for each heartbeat. If you run the same easy pace at a lower heart rate than a few months ago, your EF has gone up, which is a sign your aerobic fitness has improved.

Stridewise calculates EF for runs with reliable heart rate. On runs longer than 20 minutes, the first 10 minutes are left out, because heart rate is still settling. The trend is charted for steady easy runs. A rising trend at stable heart rate is one source of evidence for the secondary VDOT estimate and for the “progress sooner” adaptation rule.

EF is only comparable between similar runs. Intervals, heat, tiredness, illness and caffeine all move it. Look at the trend over weeks, not single values, and do not compare your EF with anyone else’s.

Aerobic decoupling (also called Pa:HR, pace to heart rate) compares the first half of a steady run with the second half. If your heart rate creeps up while your pace stays the same, or your pace drops at the same heart rate, the two have “decoupled”. A small amount is normal. A large amount suggests the run was longer or faster than your aerobic base currently supports.

Stridewise calculates decoupling only for runs of at least 45 minutes with reliable heart rate, using grade-adjusted pace. The first 10 minutes are left out as warm-up, and the rest is split into two halves by moving time. Joe Friel’s guideline is that under 5% on a long steady run suggests good aerobic endurance for that duration, and run detail marks that line.

When a long run shows decoupling above 5%, the hold long run rule keeps the next long run at that distance rather than extending it.

• Heat, dehydration and not fuelling raise heart rate in the second half regardless of fitness, and some cardiac drift happens to everyone on long runs.
• A route that is hillier in one half, or a deliberate negative split, distorts the comparison.
• The 5% figure is a coaching guideline from practice rather than a value established by controlled studies, so Stridewise treats it as a prompt to hold, not proof of anything.

VDOT is Jack Daniels’ single number for your current running fitness. It is based on a race result: it works out the oxygen cost of the speed you ran, and how much of your aerobic capacity you could sustain for that long. A higher VDOT means faster races and faster training paces.

It is not a lab VO₂max. Two runners with the same VO₂max can have different VDOTs because one runs more economically.

Two equations from Daniels and Gilbert’s tables, with speed v in metres per minute and race time t in minutes:

Training paces come from running the first equation backwards at set fractions of VDOT: easy at 65 to 74%, threshold at 88% and interval at 97.5%. Repetition pace is 15 seconds per kilometre quicker than interval pace, and marathon pace is the predicted marathon time spread evenly.

Worked example. A 20:00 5K gives a VDOT of 49.8. That sets easy pace at 4:54–5:27, marathon 4:32, threshold 4:16, interval 3:56 and repetition 3:41, all per kilometre.

VDOT sets every pace in your plan and the race predictions. It is updated conservatively:

• The main evidence is races and qualifying hard efforts from the last 42 days, such as a time trial or a parkrun run hard.
• A secondary estimate compares your pace at a fixed heart rate on easy runs over the last 28 days with the 28 days before (see efficiency factor). It can move VDOT by at most 3 points on its own reckoning.
• VDOT never rises from a single run. It needs at least 2 independent pieces of evidence pointing at least 0.5 higher, and then rises by no more than 1.0 a week.
• It drops when a race, or 2 hard efforts, come in at least 0.5 below your current level, by no more than 1.5 a week. Dropping faster than rising keeps paces achievable.
• After 5 days without running, VDOT fades by about 2% a week, up to 20% in total, until new runs show where you are.

• The equations were fitted to well-trained runners. They are less accurate for beginners and for very long events.
• A marathon prediction assumes you have done marathon training. A fast 5K does not give you the endurance for 42 km.
• A result from a hilly course, a hot day or a short course will give a misleading VDOT.

Pete Riegel noticed that, across a wide range of distances, race times grow a little faster than distance. Doubling the distance takes slightly more than double the time, because you cannot hold the same pace for longer.

Worked example. From a 20:00 5K, Riegel predicts 41:42 for the 10K, 1:32:00 for the half marathon and 3:11:49 for the marathon.

Stridewise uses Riegel as an independent cross-check on the VDOT predictions. Where the two disagree, the gap widens the predicted range rather than one being picked over the other.

The 1.06 exponent is an average across runners. Endurance-trained runners often slow down less than it predicts, and less experienced runners more, especially at the marathon. Riegel himself suggested it for events between about 3.5 and 230 minutes.

Stridewise predicts your current race time for 5K, 10K, half marathon and marathon. It shows each as a range, not a single time, because any prediction carries real uncertainty.

Each prediction combines the VDOT time with a Riegel time from your best race in the last 120 days. The range runs from the faster of the two to the slower, then is padded either side by 1% for the 5K, 1.5% for the 10K, 2% for the half marathon and 3% for the marathon, because longer races carry more uncertainty. When the two methods disagree, the range is wider. With no recent race, the range is built from the VDOT time alone.

For example, with the VDOT of 49.8 above, the VDOT tables give 41:28 for the 10K, 1:31:50 for the half marathon and 3:11:17 for the marathon.

Predictions are also used at onboarding: if your race date is too close for your goal and current fitness, Stridewise says so and suggests a realistic target or date.

A prediction assumes good conditions, a flat course, sensible pacing and training suited to the distance. It cannot know about the weather, your taper or how you feel on the day.

Readiness is a daily score that sums up how prepared you are for today’s training. It is not a measurement. It is a transparent combination of signals Stridewise already has, and every factor’s contribution is shown.

It starts from 75 and each factor adds or subtracts points. The total is kept between 0 and 100.

• Form: your training stress balance. Each point of form above or below −5 is worth 0.8 points, from −25 to +8.
• Load ratio: the acute:chronic workload ratio, as a soft signal only. Between 0.8 and 1.3 adds 5. Above 1.3 it subtracts, reaching −15 at 1.5 and at most −25.
• Effort: your last 4 rated runs in 10 days compared with the RPE expected. Each RPE point outside the expected range, on average, is worth 6 points, from −20 to +4.
• Missed sessions: −4 for each in the last 14 days, up to −12.
• Niggles: pain flags in the last 14 days cost 8 points per severity level in the first 3 days and 4 after that, plus 5 more for two or more flags, up to −40.
• Time off: −40 while you are on logged time off. After you return, 0.5 points per day off, up to −15, fading to zero over 14 days.

The score is labelled 80 and above: ready for a key session; 60 to 79: good to train; 40 to 59: take it easy; 0 to 39: rest recommended.

Readiness is information for you. The plan itself changes through the adaptation rules.

The weights are judgement calls informed by the research on each factor, not a validated model. Readiness cannot see sleep, stress or nutrition. If you feel unwell or unusually tired, trust that over the number.

Compliance shows how closely a run matched the planned session. It is a quick verdict, not a mark out of ten for effort. Running easy runs too fast is the most common way to drift from a plan, so it counts against compliance just as cutting a session short does.

Stridewise matches your run to the planned session, detecting laps and intervals so reps can be compared one by one. It checks four things, each with a status of on target, too fast, too slow, cut short or extended:

• Duration
• Distance
• Pace for each segment, using grade-adjusted pace on hills
• Intensity, using heart rate against the plan’s caps

Pace counts as on target within 3% of the target range on hard segments and 5% on easy ones. A hard session’s pace is on target when at least 70% of its reps are. Duration and distance are on target within 15% of the plan, and each segment within 10% for hard work and 20% for easy running. Heart rate may run 3 beats over a cap before it counts as too hard. Otherwise, an RPE more than 1 point outside the expected range counts as too hard or too easy.

The score weights pace at 35%, intensity at 25%, and duration and distance at 20% each, skipping any that do not apply. Running an easy session too fast costs more than running it too slowly, and a cut-short session scores in proportion to how much was done. A score of 85 or more, with no single check below 75%, earns “solid”. Patterns over several runs, such as missing pace at normal heart rate, feed the adaptation rules.

Interval detection works best with laps or a clear structure. A run interrupted by traffic lights, a dog or a chat with a friend will look less compliant than it really was. The verdict does not know why a session changed, so it never judges you, only the numbers.

Studies of successful endurance athletes across sports find a consistent pattern: most training, roughly 75 to 80% of sessions or time, is easy, below the first ventilatory threshold. A small share is hard, and relatively little sits in the “moderately hard” middle. This is often called polarised or 80/20 training.

The reasoning is that easy running builds aerobic fitness with little fatigue, which leaves you fresh enough to make the hard sessions count.

Stridewise plans each week so that no more than 20% of planned running time is at marathon pace or faster. That counts by time, not by number of sessions, and it includes strides as well as quality sessions. Warm-ups, recoveries and easy running count as easy. If a week comes out harder than that, the scheduler tries again with fewer or shorter quality sessions and keeps the first version that fits, or the closest one. Race week is exempt.

Easy runs carry a heart rate cap at 70% of heart rate reserve to stop them drifting into the middle zone. Run detail and the trends screen show your actual split, using heart rate zones.

• Most of the evidence comes from describing what elite athletes do, not from controlled trials. Intervention studies such as Stöggl and Sperlich’s are short and small.
• A meta-analysis by Rosenblat and colleagues found an advantage for polarised over threshold training only in some groups, and the advantage was modest.
• For runners on low weekly volume, a slightly larger share of quality work may be reasonable. The 80/20 figure is a guide, not a law.

Periodisation means organising training into phases with different aims, so that fitness peaks when it matters. A classic running plan moves from general aerobic work towards specific race preparation.

For a race, Stridewise works backwards from the race date through four phases:

• Base: mostly easy running, a growing long run, strides and strength. Builds the aerobic engine.
• Build: more threshold and interval work, matched to the distance.
• Peak: the most race-specific sessions, such as race pace work and long runs with a fast finish.
• Taper: less volume, kept intensity (see tapering).

The taper comes first, counted back from race day. The weeks before it are split roughly 40% base, 35% build and 25% peak, with base shortened or dropped when time is short. The emphasis changes with the goal: 5K and 10K plans have more VO2 intervals and repetitions, while half marathon and marathon plans have more threshold and race pace work. For general fitness, Stridewise uses rolling 4-week blocks instead, mixing easy volume, threshold and faster running, with the same deload rhythm as race plans.

At the start, Stridewise checks whether there is enough time. The minimum build is 4 weeks for the 5K, 6 weeks for the 10K, 8 weeks for the half marathon and 12 weeks for the marathon, and longer if your current volume is well below what the distance needs. If the date is too close, it says so and suggests a realistic target or date.

Research on periodisation models tends to compare whole programmes, so it is hard to say any one structure is best. What is well supported is the general principle: progress gradually, make training more specific as the race nears, and arrive rested.

Fitness improves when training increases gradually and you recover from it. Increase too slowly and you plateau. Increase too quickly and the risk of injury and illness goes up. A deload week is a planned lighter week that lets your body catch up.

• Each building week adds 10% or 1.5 km to the last one, whichever is more, up to a peak set by your goal and starting volume. Volume holds steady in the first week and through the peak phase.
• The week’s planned load is also limited so that fitness (CTL) rises by no more than 5 points a week. Whichever cap is lower wins.
• The long run grows by at most 2 km a week. It is capped at 16 km for the 5K, 20 km for the 10K, 24 km for the half marathon and 32 km for the marathon (35 km for marathon runners peaking at 80 km a week or more), and at 180 minutes for the marathon and 150 minutes otherwise. For general fitness the cap is 16 km.
• Runners already on at least 30 km a week whose recent weekly distance has been consistent (varying by no more than 25%) get a deload every fourth week. Everyone else gets one every third week. A deload week drops to 75% of the last building week’s volume, with the long run at 75% of its last length.
• When checking whether a race date is realistic, Stridewise assumes weekly volume grows by no more than 10% a week and the long run by no more than 1.5 km, with recovery weeks and a taper on top.
• The adaptation rules can hold progression or bring it forward based on how runs are actually going.

The popular “10% rule” is not well supported: Buist and colleagues found a graded programme did not reduce injuries in novice runners. Nielsen and colleagues found that large jumps (over 30% in two weeks) were associated with some types of injury, but individual tolerance varies widely. Stridewise uses 10% as a conservative planning default, not because it is a proven safe limit. None of these caps guarantee you will stay injury-free.

Every session in your plan comes from a library of standard running workouts. Each has a purpose statement, so you know why it is there, and a priority: key (protect this one), supporting or optional.

Recovery

Very short and very easy, to add movement without adding fatigue.

Easy

The backbone of the plan. Conversational pace, with a heart rate cap.

Easy with strides

An easy run finished with 6 relaxed accelerations of about 20 seconds, to keep your legs quick.

Long run

Builds endurance, fuel use and resilience. Easy pace, with a distance cap set by your goal and experience.

Long run with fast finish

About the last quarter at marathon pace, or half marathon pace in a half marathon plan, to practise running well when tired.

Progression

Starts easy and finishes steadily faster.

Tempo

A continuous block at threshold pace: comfortably hard, roughly an effort you could race for an hour.

Cruise intervals

Threshold pace broken into repeats with short jogs, for more time at threshold with less strain.

VO2 intervals

Repeats of about 3 to 5 minutes at interval pace, to raise your aerobic ceiling.

Repetitions

Short, fast repeats at repetition pace with full recoveries, for speed and running economy.

Race pace

Sections at goal race pace, so the pace feels familiar on the day.

Paces come from your current VDOT: easy (E), marathon (M), threshold (T), interval (I) and repetition (R). Hard sessions are never placed on consecutive days.

Following Daniels, the amount of faster running in a single session is limited to a share of your weekly distance: about 20% at marathon pace, 10% at threshold pace, 8% at interval pace and 5% at repetition pace, with sensible minimums and maximums.

Paces assume flat ground and mild weather. On hills, in heat or into a headwind, run by effort and let the pace slow.

Strength training helps distance runners. Systematic reviews find that adding it to running improves running economy, the oxygen cost of a given pace, and time-trial performance, without making you bulky. Stronger muscles and tendons may also cope better with the load of running.

If you choose one or two strength sessions a week, Stridewise schedules runner-specific sessions:

• 20 to 30 minutes, bodyweight or light equipment.
• Single-leg work, calf raises, hip and glute work and core. Higher-rep foundation work in the base phase, heavier work in the build and less volume at the peak.
• Placed on rest days or easy days where possible, spread through the week, and never the day before a key session or a race. Its minutes are reserved inside that day’s time limit, so a run on the same day is sized to fit around it.
• At most one light session a week in the taper, and none in race week.
• Counted towards your training load through session RPE.

If you flag a niggle, exercises loading that area are removed until you are pain-free.

Most of the performance studies used heavy resistance or plyometric training under supervision. Light home sessions are easier to stick with but probably give smaller gains. Evidence that strength training prevents running injuries specifically is less certain than the evidence for performance.

A taper is the reduction in training before a race, so that fatigue clears while fitness is kept. Done well, it typically improves performance by a few per cent.

Bosquet and colleagues’ meta-analysis found the best results from a taper of about two weeks in which volume was cut by 41 to 60%, while intensity and frequency were kept. Mujika and Padilla reached similar conclusions. Stridewise follows this:

• Volume steps down progressively over 7 days for the 5K, 10 days for the 10K, 14 days for the half marathon and 21 days for the marathon.
• Some race-pace and faster work stays in, in shorter doses, so you stay sharp.
• You keep running on most of your usual days.
• Strength work is reduced.

The meta-analysis pooled studies of trained athletes in several sports, mostly over shorter events. Individual responses vary, and some runners feel sluggish early in a taper. That is normal and usually passes.

A plan only works if it fits your life. Stridewise plans around the days and time you actually have, and around sessions you already do with other people.

• Available days: each weekday is either off or available with a time limit in minutes. Sessions are sized to fit the limit, and nothing is planned on days that are off.
• Long run day: the long run goes on your preferred day, and only on a day with room for it.
• Runs per week: the plan never schedules more than your maximum.
• Fixed commitments: a club track night or parkrun can be pinned to a weekday. Stridewise treats it as given and counts its load, makes it a key session where it fits, keeps other hard days away from it, and plans the rest of the week around it.
• Moving a session: tap Move on an upcoming session to see which days are valid. Days that would break a rule, such as two hard days in a row, are greyed out with the reason. The rest of the week is then re-planned and the change is logged.

With very few available days or short time limits, the plan has less room to progress, and it will say so rather than squeeze in more than fits.

Your plan is not fixed. Stridewise re-plans the next 14 days after every synced run, after every effort rating, when you log time off or move a session, and every night. Changes are made by a fixed set of rules, and every change is written to your changelog with what changed, which rule fired, the numbers behind it and a plain-English reason.

Missed session

A planned run with nothing synced for it is marked missed, and missed training is never crammed into the following days. A key session missed in the last 3 days can move to a day in the next 6, but only onto a rest day or in place of a lower-priority run, where it fits your time limit and your runs per week, is not next to another hard day or the day before a key session, and is at least 4 days from another long run. If no day qualifies, it is dropped. Fixed commitments and races are never moved. After 3 or more misses in 7 days nothing is moved at all, and the plan carries on from today. Sessions that fall in logged time off, or before your plan started, are marked skipped rather than missed.

Effort above expected

This fires when one run in the last 7 days felt at least 2 RPE points harder than the top of its expected range, or 2 runs felt at least 1.5 points harder. One session in the next 7 days is eased. The first choice is the next hard session, with about 25% fewer reps or a shorter block, 5 s/km slower. A fixed commitment stays, but is capped at threshold effort and the last 30% of its hard running becomes an easy jog. If there is no hard session to ease, the long run is cut by about 13% and kept fully easy, or the strides come off an easy run. Progression is held while this applies.

Pace missed at normal heart rate

If, within 10 days, 2 runs (or a single quality session) came in slower than the pace target while reliable heart rate stayed where the plan expected, the same easing applies. That pattern usually means tired legs rather than lost fitness.

Progress sooner

If, over the last 14 days, at least 3 rated runs felt easier than expected, none felt harder, and your efficiency factor across at least 4 runs with reliable heart rate has not dropped by more than 1%, the next quality session moves up a step: one more rep, or a block about 15% longer, as long as it stays within Daniels’ limits and your time limit. If there is no quality session to progress, the next long run grows by 1.5 km, up to its cap. This never fires while another rule is holding progression.

Hold long run

If your most recent long run in the last 14 days (a planned long run, or any run of 75 minutes or more) shows aerobic decoupling above 5%, the next long run stays at that run’s distance instead of growing.

Overreaching

This fires when the acute:chronic workload ratio is above 1.3, or form (TSB) falls below −25, or below −30% of your fitness (CTL) when that is lower. Your next day becomes an easy recovery run of up to 5 km, and the rest of the week is capped at 80% of its target load: easy runs are shortened first (to no less than 50% of their length), then the long run (to no less than 75%). A hard fixed commitment that week is capped at threshold effort, with its last 30% jogged easily. If the ratio is above 1.5, or form is below −35 (or 10 below the threshold, when that is lower), the next day is rest and the day after is a recovery run.

Niggle

If you flag pain, the next 7 days are adjusted. Hard sessions and strides become easy running, runs are about 30% shorter, and strength exercises that load the sore area are swapped or removed. A fixed commitment stays on its day as an easy run for the same time. At severity 3, hard sessions, including fixed ones, become rest days instead. Normal training resumes once you log a pain-free run, and a flag stops counting after 21 days.

Repeated niggle

If the same area is flagged 2 times within 14 days, or once at severity 3, the next 2 days become rest days, or 3 days after a severe flag or 3 or more flags. If it still hurts after the rest, Stridewise suggests seeing a physiotherapist before running hard again. Repeated pain is a signal to get a professional opinion, not to train through it.

Return to run

After logged illness, injury or time off, the plan restarts with a ramp based on how many days you missed. See returning after illness or time off below.

Rebalancing the week

When the other rules have changed a week’s planned distance by more than 5%, easy runs that no rule has touched are lengthened (by up to 20% each, within your time limit) or trimmed (to no less than 3 km) to bring the week back towards its target. Weeks already reduced for a niggle, overreaching or a return from time off are left alone.

The rules only know what your runs, ratings and flags tell them. They cannot see a stressful week at work or a bad night’s sleep. You can always move or skip a session, and the plan will adjust around it.

After illness or a break, your fitness has dropped a little and your tissues are less used to running. Coming back too fast is a common cause of injury and of illness dragging on.

A widely used rule of thumb is the “neck check”. If your symptoms are only above the neck, such as a runny nose or mild sore throat, gentle running may be fine. If they are below the neck, such as a chest infection, fever, aching muscles or an upset stomach, rest until they have gone.

When you log illness, injury or time off, sessions inside that period become rest and nothing is carried over. When you come back, the ramp depends on how many days you missed:

After illness, hard sessions return only once you have logged 2 symptom-free runs, however long the break. After more than 28 days away, Stridewise also recommends regenerating your plan from your current fitness. Your fitness numbers and VDOT will have dropped during the break, and the progression caps work from where you are now, not where you were.

The neck check is practical guidance, not a validated test. After a fever, a viral illness such as COVID-19, or any chest pain, palpitations, unusual breathlessness or fainting, speak to a doctor before running again. Stridewise cannot assess medical conditions.

Stridewise reads your activities from Strava to analyse your runs and plan your training. That data is yours, and it is handled under strict rules.

• Your Strava data is only ever shown to you. There are no public pages, feeds or leaderboards.
• It is never sent to any AI model, language model or other external service. Every calculation on this page runs inside Stridewise.
• It is never sold or shared.
• If you disconnect Strava or revoke access from Strava’s settings, your Strava-derived data is deleted. “Delete my account and data” removes everything.
• The full details are in the privacy policy.