Most runners steer their training by a number on their wrist. That works as long as the number is right — and that is exactly where the problem starts. GPS pace jumps in corners, drifts in forests and briefly displays values nobody actually ran after an interval. If you only follow the watch, you are reacting to a calculation. If you train your pace feel, you are reacting to your body — and you can still make decisions in a race when the technology gives up or your legs are tired.
Why your watch knows your pace worse than you do
A sports watch calculates pace from position points. Depending on the device and satellite geometry, there is one second between two points, sometimes more. Within that second a short average is formed, reshuffled with every new point. Under trees, between tall buildings or when cloud cover changes abruptly, a point drops out or lands 20 metres off — and the alleged pace spikes to 3:10 min/km for three seconds while you are running perfectly even.
Then there is the question of what the watch actually shows. Most devices display a smoothed pace of the last 5 to 15 seconds. That is sluggish: when you deliberately slow down on a climb, the old, faster number stays on screen for another ten seconds. You brake further although you already reacted — and then correct in the other direction. This lag creates exactly the uneven splits many runners mistake for a lack of discipline.
The third point is physiological: your pace is the result of your current effort, your fatigue and the terrain. Your feel knows all three variables directly. The watch only knows your position. Into a headwind, in heat or on a climb, the same feeling can mean a pace 20 seconds per kilometre slower — and still be the physiologically correct intensity. That is why pace is a tool, while effort is the standard.
Three sources of error: GPS, smoothing, interval mode
- GPS jumps: In wooded sections, on tight bends of a track and in dense city centres, errors of 10 to 30 seconds per kilometre are normal. On a 400 m track GPS pace is practically worthless — there the lap time counts, not the display.
- Smoothing: A heavily damped display feels calm but shows outdated information. An undamped one jumps. If you run by display, you always run behind the past — the stronger the smoothing, the bigger the lag.
- Auto-pause and interval mode: Every automatic pause shifts the reference point. After an interval, the watch often shows the average pace of the whole session instead of the current rep — and you read a number that has nothing to do with what you are running right now.
The practical consequence is not "throw the watch away", but: the watch is a control instrument with a known measurement uncertainty. If you know that uncertainty, you read it correctly — as a trend over kilometres, not as an instantaneous value per second.
What pace feel actually is
Pace feel is the ability to hit a given pace without technical help — experienced runners land within three to five seconds per kilometre, elite athletes even closer. It is not a gift but a learned comparison: over years your body stores how 4:30 min/km feels, breathes, sounds. This reference library in your head is remarkably precise because it draws on several senses in parallel.
Four channels supply the raw data: breathing (how deep, how fast, how much resistance), cadence, ground feel (how hard the foot lands, how long the contact lasts) and muscular tension in calves, glutes and trunk. From these four signals a single overall impression emerges — and that impression is often more reliable in a race than a number that is twelve seconds old.
The distinction matters: pace feel is not the same as perceived exertion (RPE). RPE describes how hard something feels — in heat, the same pace is more demanding. Pace feel describes how fast you are moving. Together they form the actual craft of pacing: you want to know whether your pace matches the planned intensity, and for that you need both values. How the RPE scale works is explained in RPE: Training by Perceived Exertion.
Calibration: four drills that actually work
Pace feel only improves with feedback. Simply running without a watch does not train your feeling, it trains your indifference. What matters is the cycle: estimate, run, check, correct. Four drills cover this, and all of them fit into existing sessions without changing the plan.
- Blind 400s on the track. Four laps at marathon pace, no watch — after each lap you hit the button yourself and only look afterwards. Beginners are typically 10 to 15 seconds off, after eight weeks usually under 5. The track is the only place where the reference is truly exact.
- The pace-shift test. During an easy run, three sections of two minutes each: one deliberately 15 seconds faster than your base pace, one identical, one 15 seconds slower. Check the splits afterwards. The drill trains fine resolution — most runners can distinguish two paces, but not three.
- Route anchors instead of numbers. Pick fixed points on your home loop (a tree, a junction, a lamppost) and check the time there. Over months you develop a feel for that route that works regardless of weather and daily form — you notice immediately when you are faster than usual today.
- The silent kilometre. In the long run, run one kilometre per week without glancing at the watch, aiming to hit exactly the average pace of the session. The drill is uncomfortable because it offers no distraction — which is precisely why it works.
The order matters: first estimate, then run, then check. Anyone who looks at the watch mid-run is not learning — they are only correcting. And correcting is the opposite of calibrating.
Breathing, cadence and ground contact as a speedometer
When no display is available you need substitute readings. Three have proven themselves because they change systematically with pace.
- Breathing rate: At easy pace almost everyone runs a 3:3 rhythm — three steps in, three out. As it gets quicker this flips to 3:2, at threshold pace to 2:1. The moment you fall into a different breathing rhythm is your most important sensory marker: it marks the transition from base to intensity more reliably than any heart rate display with its delay.
- Cadence: It does not rise linearly when you accelerate but in jumps of two to four steps per minute. If you know your typical frequency at base pace (168–178 spm for most), you notice a pace increase in your legs before the watch shows anything. The myth around 180 spm is covered in Running Cadence.
- Ground contact time: The faster you run, the shorter the contact. The feeling of feet springing off rather than sinking in is a very direct indicator — and it still works when your legs are tired and the watch has long become unreliable.
You do not have to measure these three markers. You have to notice them consciously once and link them to a known pace: after a controlled session, sit down and note how breathing, cadence and contact felt. After five such notes you own a reference no watch can give you.
Pace feel in the race: let feel lead, then check
In competition it becomes clear whether the training paid off. The rule is: feel leads, the watch checks — not the other way round. Concretely, that means running the first kilometres by effort and using the watch only for confirmation.
The reason lies in start dynamics: in the first three kilometres heart rate and breathing are elevated by nerves while the legs feel light. If you now look at the watch and see your target pace, you are trusting a number that knows nothing about your current capacity. If you run by feel, you notice "this is too fast for kilometre one" — and correct before the mistake costs 40 seconds.
- Kilometres 1–5: feel leads. The aim is for the effort to feel "too easy". If you think you could go substantially faster, you are right on target.
- Kilometres 6 to two-thirds: switch to block checks. One look at the split every 5 km, correct deviations of more than 10 seconds — but over the next few kilometres, not instantly.
- Final third: feel again. Now the watch is information about your state, not an instruction. Anyone who gets faster by feel over the last 10 km chose the right pacing strategy. See Marathon Pacing Strategy for the methodology.
A practical note for big city marathons: kilometre markers are the best control because they do not jump. One glance at the marker, the split roughly calculated in your head, then back to feel — that is more robust than looking 42 times at a display reading 4:12 while you are running 4:30.
When you genuinely need the watch
Pace feel does not replace the watch in every situation. There are areas where the number is objectively superior:
- Interval training with exact targets: With 6 × 800 m in 3:05 you want the spread between reps to stay small. Here the watch is the reference, because feel drifts systematically under fatigue — the last rep feels identical but is 8 seconds slower. On recovery design see Interval Recovery Periods.
- Time trials and tests: A 5 km test or a timed threshold run needs a hard measurement. Feel can accompany a test, never replace it.
- New routes: On unknown terrain without distance markers, the watch is the only anchor. There the rule flips: watch leads, feel checks.
The real gain is not "without a watch" but independence: if you have trained your pace feel, you can still run sensibly when the battery is dead, the satellite spins or the track blocks the signal. And you can make a decision in a race when the number and the body disagree.
The training plan for the feeling
Pace feel is not a project but a habit. An eight-week build that fits any plan:
- Weeks 1–2: run one silent kilometre in every second easy session. Just notice, do not correct.
- Weeks 3–4: blind 400s on the track, once a week. Four reps, note your estimate after each.
- Weeks 5–6: add the pace-shift test twice a week inside existing runs, and establish route anchors on your home loop.
- Weeks 7–8: one complete easy session without looking at the watch — check the splits afterwards. If you stay within 5 seconds per kilometre, you have a reliable feel.
The cost is about 15 minutes per week. The return is larger than it looks: better race pacing, less watch stress, and the ability to execute a plan when the technology does not cooperate.
How Peakora combines feel and numbers
In the Peakora engine every session gets a target pace, but control runs through the intensity zone rather than a rigid minute target. On a suppressed day the plan proposes an adaptation based on effort and form — not on holding a pace that does not match your recovery state today. Form is calculated via TSB from the Banister model, weighted at 65 %, plus a 35 % sleep factor. Pace stays the output of training control, and your feel remains the standard against which you check it.
Common mistakes
- Looking at the watch too early. Anyone checking the display after 200 metres has not built a feeling, only read a number. The first two minutes belong to the body.
- Confusing feel with fatigue. "It feels fast" does not mean "it is fast". In heat, wind and after hard days every pace feels fast. That is why the check exists.
- Training only without a watch. Without feedback nothing improves. The check step is the actual training stimulus.
- Mixing pace feel with daily form. A bad day says nothing about your pacing accuracy. Assess them separately — how to do that is in Daily Form: Adjust Training to the Day.
Frequently asked questions
What is pace feel in running?
Pace feel is the ability to hit a target pace without looking at the watch — usually within three to five seconds per kilometre. It comes from experience with your own breathing, cadence and effort level, not from talent.
How do I train my pace feel?
Run segments without the watch and check the result afterwards: estimate 400 m at marathon pace, then verify. Three to four such blocks per week, over six to eight weeks, measurably improve accuracy.
Why does my watch show a different pace than I feel?
Because GPS pace is a rolling average of the last few seconds and jumps in corners, under trees and between buildings. The displayed number is a calculation; your feel is a physiological state — on uneven ground and in wind it is usually the more honest value.
Can you run a marathon without a watch?
Yes, with some experience it works very well. Big city marathons have kilometre markers, so feel plus an occasional split check is enough. For target times under 3:15, checking every 5 km is advisable because small pacing errors get expensive.
Train with a plan and with feel
Peakora gives you target paces per zone and adapts them to your daily form — with a rationale instead of guesswork. Start free, no credit card required.
Start free