Marathon Pace KM
Menu

Calculator methods and limitations

Published by · Updated

Methods updated 23 October 2026

Pace, finish time and splits

We use a marathon distance of 42.195 km, a half marathon of 21.0975 km and 1 mile = 1.609344 km. Pace in seconds per kilometre is total seconds divided by distance in kilometres. Pace per mile is that value multiplied by 1.609344.

A checkpoint at distance d is total time × d / race distance. Calculations retain full precision; displayed elapsed times round to a second. Full split tables on the homepage and newer tools calculate segment times as differences between displayed cumulative checkpoints, so their segments sum to the displayed finish time. Some quick pace labels round to whole seconds and therefore cannot reconstruct the exact goal time.

For four hours, the average is about 5:41.27/km. Halfway is exactly 2:00:00. A sub-four target must be below four hours, with all pauses included. These calculations describe the measured course distance. For an example of how training and execution can differ from a calculated target, see my Sydney review.

Even, controlled-start and negative-split bands

The pace-band generator uses the same full-distance total for each strategy. Even pacing distributes time in proportion to distance.

For a controlled start, the selected time allowance accumulates gradually across the first 5 km. After 5 km, that allowance is recovered evenly over the remaining distance: 37.195 km for a marathon or 16.0975 km for a half marathon. Let D be the selected race distance, 42.195 or 21.0975 km. If the allowance is C seconds and the average pace is P, the opening cumulative time at d ≤ 5 is P × d + (2C/5) × (d − d²/10). Beyond 5 km it is P × d + C − C × (d − 5) / (D − 5).

For a negative split with a second half N seconds faster than the first, cumulative time is P × d + (2N/D) × (d − d²/D). Halfway receives half the selected difference; the finish remains the requested total. For a 2:00:00 half marathon with a two-minute difference between halves, the halfway checkpoint is 1:01:00 and the finish remains 2:00:00. For a 30-second controlled-start allowance, the first 5 km takes about 28:56 and the remaining 16.0975 km requires about 5:39.4/km on average. These are planning shapes, not scientifically validated optimal pacing strategies. The required later effort may not be sustainable.

Print wrist bands at actual size and check the 50 mm scale. The checkpoint-sheet option prints a separate reference. Share links preserve distance, exact goal seconds, units, strategy, reading direction and paper/layout choices. The standalone calculator supports 5 km and 10 km splits and quick printing, but disables unsupported pace-band exports. Saving as PDF depends on the browser's print options.

Race predictions are scenarios

The main predictor uses T2 = T1 × (D2 / D1)exponent with illustrative exponents 1.06, 1.08 and 1.10. Higher exponents allow more slowing with distance. The result is rounded to 30 seconds before checkpoints are built. Training-context answers are shown for interpretation; they do not create a calibrated fitness score or change these exponents.

Some article calculators expose different exponent choices or blend shorter-race estimates; follow the assumptions beside that calculator. None of these outputs is a personal confidence interval. My own half-to-marathon comparison shows why even a plausible calculation can be too optimistic. Pacing, endurance, course, conditions and fueling can all differ between races.

Fueling totals

The user selects a carbohydrate target, gel size, first-gel time, interval and any planned carbohydrate from drinks. A gel is scheduled at each selected time before the finish. Total scheduled gel grams divided by race duration gives the race-average gel intake. Planned drink grams per hour are added for the combined average. Spare gels are carried, not counted as consumed.

The calculator compares the scheduled total with the target; changing the target does not automatically change the interval. If planned drinks already cover the target, no gels are scheduled. Product timing and an intake estimate do not establish consumption or absorption. Burke and colleagues' nutrition review supports a personalised, practised plan; it does not validate every default in a scheduling tool.

Race-condition adjustments

The conditions tool is an illustrative model. It adds 0.25 percentage points per degree Celsius above 15°C and 0.35 per degree of dew point above 12°C, then applies the selected acclimatisation factor and a heat-component cap of 12%. If dew point is blank, it is estimated from temperature and relative humidity using a Magnus-style formula.

Wind above 10 km/h adds 0.15 percentage points per km/h, capped at 4%. Terrain adds 0%, 1.5% or 3% for the selected category. The selected planning preference scales the combined adjustment; the displayed lower and upper examples are 70% and 115% of that central value, capped at 25%.

These coefficients are site heuristics, not a validated prediction, probability range or safety threshold. Temperature, humidity, wind exposure, hills and individual response cannot be reduced reliably to one universal penalty. The linked guides explain why effort may need to change; the model's numbers must not overrule symptoms or event advice.

The monthly training planner

The baseline is the average of all four entered complete weeks, including zero weeks. The planner combines that baseline with selected days, recent longest runs, available time, training experience and the chosen phase. Its progression, session selection and distance caps are hand-set rules, not a validated prescription of a runner's optimal load.

A previous peak week is different from a sustained average. Race-derived training paces and generated sessions must be checked against current effort, recovery and suitability. The planner does not diagnose injury or provide rehabilitation. Shared links can include training and injury-status answers; see Privacy before sharing.

Check or correct an example

The site is maintained by Davin Pinto. Report the page, exact inputs and unexpected result through Contact and corrections. The editorial policy explains how calculations, research and experience are distinguished.

Sub-threshold pace estimates

The sub-threshold tool uses fixed seconds-per-kilometre offsets from recent race pace, followed by training-level and condition offsets. The linked table documents every value and a 50-minute 10K example. These are site heuristics, not a lactate test, a fitted research model or a validated physiological threshold estimate.

Treadmill speed and pace

The treadmill chart and converter use seconds/km = 3600 ÷ km/h and km/h = mph × 1.609344. Pace is displayed to the nearest second; speed to two decimal places. Time-at-speed distance excludes acceleration and pauses. Incline and running economy are not modelled.

Training start-date comparison

The 12-, 16- and 20-week calendar comparison subtracts 84, 112 and 140 days from the entered race date. UTC date arithmetic avoids daylight-saving shifts. Dates do not imply fitness, injury clearance or readiness for a programme.

Run-walk target arithmetic

The run-walk calculator subtracts stopped time from the finish target, counts full run-then-walk cycles and assigns the remaining moving time to a partial cycle starting with running. Stops pause the interval timer. Walking distance is walking seconds divided by walking pace in seconds/km. Running seconds divided by the remaining part of 42.195 km gives the required running pace. It assumes constant speeds and immediate transitions. The worked guide shows the calculation; it does not predict whether a ratio or pace is sustainable.

Taper worksheet

The taper worksheet multiplies one user-entered reference distance by each percentage remaining. Percentages apply to that same reference, not successively reduced weeks. Race distance is added separately: 42.195 km or its full-precision mile conversion. Displayed training totals are rounded to one decimal place. Selecting a unit reinterprets the reference-distance number in that unit. Default percentages are editable illustrations, not an individual prescription.

Pace-group timing and catch-up pace

The pace-group worksheet uses group pace in seconds/km = target seconds ÷ 42.195. Required catch-up pace = group pace − gap seconds ÷ catch-up distance in km. It assumes a time gap measured at the same marker and constant group speed. A result of zero or below is rejected. It cannot assess the effort or suitability of catching up. If runner and pacer finish together, runner chip time = pacer chip time + pacer start clock time − runner start clock time.

Half-marathon checkpoint plan

The half-marathon worksheet uses 21.0975 km. Average pace p = target seconds ÷ distance. With opening offset o seconds/km, first-5-km pace = p + o. Remaining pace = (target seconds − 5 × opening pace) ÷ 16.0975. A checkpoint d kilometres from the start uses opening pace × min(d, 5) + remaining pace × max(d − 5, 0). Offset zero gives even pacing. There is no course model or fitness prediction. Checkpoints are rounded independently to seconds, pace to tenths of a second; calculations retain full precision. The fixed opening remains 5 km even when mile paces are displayed.

800 m session arithmetic

The Yasso workout calculator counts n repetitions at r seconds each and n − 1 recoveries of q seconds: main set = n × r + (n − 1) × q. Warm-up and cooldown minutes are added separately. Fast distance = n × 0.8 km; repetition pace = r ÷ 0.8 seconds/km. Recovery distance and whole-session distance are unknown without easy-running speeds. No final recovery is counted separately from the cooldown. The familiar minutes:seconds to hours:minutes label is shown only for ten equal-time-recovery efforts and is explicitly not a forecast, error range or recommended marathon goal.

Daily carbohydrate target and food record

The carb-loading worksheet converts lb to kg using 0.45359237 kg/lb, then multiplies kg by the user-entered g/kg/day target. Each food contributes carbohydrate grams per defined serving × serving count. Planned and recorded totals are separate. A blank recorded count stays unknown; zero means none consumed. The recorded subtotal includes only entered records, with the completed-entry count displayed. The plan gap is max(0, target − planned); an excess is stated separately. Targets and food totals are shown to one decimal place without intermediate rounding. The tool models one day and does not assess diet completeness, absorption or glycogen stores. Entries are not persisted. CSV downloads preserve unknown records as blank and escape formula-like text cells for spreadsheet safety.

Long-run duration and weekly context

The long-run worksheet multiplies distance by matching-unit moving pace and adds entered stopped minutes for elapsed time. Weekly share = long-run distance ÷ weekly distance × 100; the weekly total must include the run. Recent change = (candidate distance ÷ longest distance in the preceding 30 days − 1) × 100. Three-hour moving distance = 10,800 seconds ÷ pace. Kilometres and miles use 1.609344 km/mile. Changing units reinterprets the input numbers rather than converting them. Terrain, fatigue and recovery are not modelled. Percentages describe the entered plan; no injury score, safe progression or ideal weekly share is inferred.

Half-marathon, track and pace-conversion references

The half-marathon chart uses 21.0975 km. A checkpoint at d km is finish seconds × d ÷ 21.0975. Track time is seconds per kilometre × metres ÷ 1,000. Pace per mile is seconds per kilometre × 1.609344. A 1,600 m repetition is distinct from a 1,609.344 m mile.

Reference tables use full-precision inputs before final rounding to seconds. Optional controls show paces to a tenth of a second. The pace-band link rounds its final finish target once to a whole second. Inputs accept minutes:seconds for pace and hours:minutes:seconds for finish time; seconds and minutes in finish-time fields stay below 60. Invalid entries hide the previous result and disable the custom band link.

The half-marathon worksheet multiplies a completed weekly baseline by a user-chosen percentage. Six non-negative day weights divide that pre-race budget. For display, it rounds the total to tenths and allocates remaining tenths by the largest fractional remainder, with earlier days breaking ties. Displayed daily amounts therefore add to the displayed training budget. Race-day warm-up/cool-down and the 21.0975 km event are added separately. Changing km/miles reinterprets the numbers entered.

These are arithmetic tools. They do not choose a safe training load, establish a physiological zone or predict fitness. Values remain in the page; no new storage or form-value analytics is introduced.