Straight answers about treadmill accuracy, pace, and indoor running. We measure belts with a tachometer for a living, so the numbers here are real.
What the research says, what we've measured ourselves, and four ways to check your machine.
What your watch is really doing indoors, what the treadmill display actually reports, and how to find the real number.
How to tell whether your treadmill needs calibrating, how to check its real speed yourself, and who should actually do the work.
What the 1% incline rule actually fixes, why the belt still feels harder some days, and the one gap no pace conversion can close.
What actually changes from one treadmill to the next, what a small difference costs you, and how to check a machine you have never used.
Where the sensor really sits, the stored number behind the miles per hour on the display, and why steady is not the same as right.
Why your watch cannot reach the belt, what Apple's and Garmin's own calibration instructions really do, and what happens when you calibrate against a display that reads long.
Every way to log an indoor run on an iPhone, sorted by where each one gets its distance. Only one of them looks at the belt you actually ran on.
Usually yes, because a pod reads your foot instead of your arm swing. What you calibrate it against decides whether your treadmill's error follows you.
Zwift shows the speed your treadmill sends it, and says so itself. Peloton never says where its number comes from. Neither one measures the belt.
It dips at every footfall and the motor wins it back inside the same stride. The dip you can feel is almost never where your missing pace went.
We put a handheld tachometer on four treadmills. Two gym displays read about 12 percent above the real belt speed, a third read about 67 percent above, and a hotel treadmill about 2 percent above.
Three things open that gap, and the machine cannot settle any of them. Its own pace field is the speed number inverted, so the two can never catch each other out.
What 1.00 and 0.60 actually mean, in both units. Then the reason a machine can show you three different distances at once, and only one of them is your run.
Rep against rep your session is honest, because the belt hands every rep the same error. As a number you take to a race it is not, and changing machines resets even the comparison.
Two percent is ten seconds a mile and four tenths of a mile on an 18 miler. The error lands the same way every session, so the miles add it up instead of averaging it away.
Every common setting turned into pace, for machines in miles per hour and in kilometres per hour. Then the part no other chart tells you: whether that number is true.
The calibration step copies the machine's distance into your watch. That makes them agree. It does not make either one right.
The number is arithmetic from your settings, not a measurement of you. What the published tests found, and the one number in your run you can actually check.
Your weight feeds the calorie estimate and cannot reach the speed or the distance. Skip the box and the machine quietly assumes a person of about 150 to 155 pounds. The number your training runs on has no box at all.
Everyone presses 6.5 and no two belts have to agree. Each machine turns the number into belt motion its own way, and no documented service schedule ever checks whether the speed on the display is true.
The 3.0 is a setting the machine was asked for, and the workout has no outside check that would catch a slow belt. The studies behind it took the display at face value too. Walking speed is the easy one to measure.
Every belt gives a little under a footstrike and you never feel it. A slip is the belt losing its grip under your foot, and the display cannot tell you. One slow-walk test tells a slip from the normal dip, and what to report to whoever fixes it.
The 1 percent incline settles the effort. The distance is a different kind of number: outdoors your watch measured the ground, on the belt the miles come from a setting nothing checks, and a winter block can drift on it without you knowing.
The goal-time-to-speed table for every common marathon time, the incline rule's real tested range, and why the pace you rehearse for months is a setting nothing on the machine checks.
A predictor multiplies whatever pace you feed it, error included. The pace from a treadmill time trial is a speed setting nothing on the machine checks, and a belt a few percent off hands you a race goal your legs never earned.
Your treadmill never measures the slope you run on. It moves the deck from its own zero point and shows the grade you pressed. Here is what two points of grade cost a hill session, and how to check the deck with your phone.
Your virtual 5K time is how long the display took to reach 3.1 miles. A road course is measured with a checked wheel before race day. Here is what a small gap does to the time you submit, and the 3.10 trap.
No. A tilted belt carries you the same mile as a flat one. Incline adds climb, about 525 feet a mile at 10 percent, and that climb is only as real as the deck's grade.
Strava never measures your treadmill run. It logs what your watch, GPS or synced display sent it, and none of those measured the belt. Here is where the number comes from and what to do before your next run.
Yes, and the effort debate is settled. The question nobody asks is whether the mile you log was a whole mile, because nothing on the machine checks the display against the belt.
Nobody can say which is closer. Lab studies put the watch's calories 10 percent or more off, the treadmill's readout has no side-by-side test, and neither device looks at the belt.
A treadmill-only block puts every run on one belt, so nothing inside the block can show that belt's error. Check it before week one and again before you pick a goal.
The fitness does. The paces may not, because every treadmill pace was a setting and race day was the first pace anything measured. Here is how to tell which gap you hit.
A new treadmill still shows the speed you set, and nothing on it looks at the belt itself. Your new belt and the gym's are two unmeasured belts, so measure yours in week one.
Yes, you can run a half marathon on a treadmill. The display's 13.1 is a count you cannot check, so know your belt's real speed before the run and let the clock tell you where 13.1 is.
The plan counts minutes, and the treadmill decided where 5K fell. Measure the belt you trained on and you will know your true 5K time before your first outdoor race.
The rule comes from one 1996 study of nine runners at paces from about 5:22 to 9:11 a mile. The 1% offsets the air you would push outside, and it trusts that your belt runs the speed it shows.
The room and the missing breeze explain part of it. A belt running faster than its display is the part you can check, because your heart answers to the real speed.
Your usual setting feels quicker today, so either the belt changed or you did. A service can change what a setting means, and the belt is the one you can measure.
The display keeps the same speed, incline and calories when you hold the rails, but your body does less of the work. One study measured how much.
The distance you type after an indoor run resets your watch's pace at every cadence from the next run on. Type a number from a treadmill that reads long, and every indoor pace inherits it.
The clock is fine. The mile is the machine's own count, and a belt running 3 percent slow turns a 7:00 display mile into about 7:13 of real running.
Yes, and most trackers count treadmill steps well. The miles are the weak part, which is why your watch and the treadmill rarely agree.
On an Apple Watch, treadmill runs never enter the estimate. Where your treadmill still costs you is the pace you pick for it.
Find the speed by heart rate, then treat it as a label for that one belt. The same setting on another treadmill can be a different real pace.
A tempo sits close to a narrow line. A belt that misses its setting by a few percent can push the run over it.
Brisk is an effort, not a speed. A 15-second step count tells you if you are there, whatever the display says.
Speed and grade numbers do not line up across sessions. Climb per hour does, if the belt runs the speed shown.
Watch the number on the display. If it moved, a program or a key changed the setting. If it held, the belt did not keep up.
Stay under the line by breath. The belt's error gets in through the number you carry in and the pace you log.
For a 12:00 test, 7.5 mph. Round up to the next tenth, and know that each 1 percent of belt error is 6 to 9 seconds at the pass line.
On a track you count the laps, and your lane decides the distance. On a treadmill the mile is the display’s word.
You can run it on a belt. The formula takes a fixed chunk off your distance first, so a belt 3 percent slow moves the score about 4 percent.
At 12% and 3.0, the climb makes about 60% of the calorie figure. One point of grade moves it about 5%, more than a slow belt does.
Easy paces tend to feel easier indoors and faster ones harder. The speed you compare is the setting you pressed, and a belt a little off changes the pace you ran.
Every surge starts slow and every float starts fast, so short pieces get rounded off. How much depends on how fast your machine changes speed.
A treadmill 10K time compares fairly with your own earlier time on the same machine. Each percent of belt error is worth about half a minute at 10K.
When your Fitbit counts steps, it works out distance as steps times a stored stride, and the treadmill reports its own figure. Only the belt settles which to trust.
A sprint from a walk is the longest climb a workout asks of the belt, so a short rep can spend most of its time getting up to speed.
Nike Run Club and the treadmill display can end the same run on two different distances. The gap shows they disagree. Only the belt can settle it.
Studies found no reliable step-rate gap between belt and ground. A belt that runs off moves your pace, and barely your steps a minute.
Lab tests of young adults put the walk-to-jog switch near 4.2 to 4.6 mph. Find your own with a slow ramp, then check the belt before you log your first pace.
In one study of trained runners, hard uphill intervals bought about 2 percent. That is the same size a belt can be off, so check the belt before you trust a retest.
Apple says watchOS 27 brings more accurate treadmill readings. What a sharper watch can and cannot tell you when it disagrees with the treadmill.
Your Polar learns its corrections outdoors, and a stride sensor can learn the display's number by mistake. Why neither one can settle a treadmill gap.
In the manuals we read, wear shows up as strain, not a slow fade. The error that stays quiet is a setting nobody ever checked.
There is no set speed. Judge each rep by heart rate at minute two, not minute one, and use incline as the second dial.
The display shows the speed you set, not the belt's. At walking pace a small belt error costs the most seconds per mile.
The speed for each stage in km/h and mph, why the belt skips the turns, and how a few percent off moves you most of a level.
Find your speed and grade, read off the flat pace with the same effort, and see why every row depends on the belt's real speed.
Runna logs treadmill runs from the plan's target pace, so the record can only be as right as the belt under it.
The Gerkin scores minutes until your heart rate hits its target, so the belt's real speed sits under every minute you walk.
Your Suunto learns running and walking separately from a known distance, and a treadmill total is not one. Check the belt first.
Your watch and the treadmill show different paces. Read the gap by where it follows you, then settle it with the belt.
Your WHOOP builds Strain from your body. Here is where a treadmill run's distance can come from, and which number to check.
Your watch and the treadmill gave two distances for one walk. See what the watch learned outdoors, what a treadmill walk changes, and how to tell which number is off.
Your ring counts a treadmill walk from your finger. Why the steps can come up short, and where the walk's miles come from.
A world ranking runs on gym treadmills. What the RUN X rules check, what they leave to your machine, and what 1 percent does to a 25-minute 5K.
Your strap learns its treadmill pace from outdoor GPS runs or from the treadmill display. That decides what the number on your watch is worth.
Your watch never sees the slope directly. Here is how the climb reaches its calories and distance, and the one number you can check.
Your Garmin counts your wrist. Hands on the rails or a keyboard can shrink the step count and the watch's miles together.
Strava builds best efforts from GPS data, so a treadmill 5K can miss the list. The number worth doubting is the belt's distance.