Splits Calculator

What this page does

A splits calculator normally answers one question: if I want to finish in x, what time should I be at each mark? The answer is one pace repeated down a column, and it is the right answer to a simpler question than the one a chip-timed race actually asks. In a race with more than one starting wave, the time on the clock above your head at each kilometre marker is not the time you have been running.

So this page carries two clocks. You give it a chip time and a start-line gap; it prints every split twice — once on chip time, the number results are ranked on and the one you keep, and once on gun time, the number painted on the race clock at that mark. Pick your distance from 5K to 50K or type any distance from 0.1 to 100 km.

Everything runs in your browser. There is no server, no account and no upload, and the target you type never leaves the page.

Two clocks, and why the second one exists

When the pistol fires, two stopwatches start. One of them is attached to you, and it starts when you cross the timing mat at the start line. The other is attached to the race. By the time you reach the mat you may have spent twenty seconds walking forward in a crowd, or two minutes, or six — and what you spent getting to the line is already on the race's stopwatch and not on yours.

• Chip time (also called net time) — mat to mat. It starts when your tag crosses the start mat and stops when it crosses the finish mat. It is the time you keep, and in chip-timed races it is the time results are ranked on. Most organisers publish it alongside gun time in the results table.
• Gun time (also called gross time) — pistol to tape. It starts with the whole field and stops when you cross the finish line. Every kilometre marker's clock reads gun time. This is the clock you are actually looking at when you glance up at 10 km.

The two numbers differ by one quantity and that quantity never changes during the race: the time it took you to get from where you were standing to the start line. If that was g seconds, then at every moment chip time plus g equals gun time. At the finish, the gun clock reads your chip time plus g — which is why a 1:30:00 chip time can sit under a 1:31:47 gun time and both be right.

The arithmetic, in full

t_gun(k) = t_chip(k) + g    where t_chip(k) = E_k × T

• T — your target chip time, in seconds. 1:30:00 is 5400 s.
• g — the start-line gap, in seconds. Your entry.
• E_k — the fraction of the race covered at mark k. On even splits this is simply the mark's distance divided by the race distance.
• t_chip(k) — the chip clock at mark k. This is the column you keep.
• t_gun(k) — the gun clock at mark k, and therefore the number on the race clock above that marker.

The whole model is that one addition. It is worth stating plainly because it is the part that gets lost: the gap is constant. It does not shrink as you run well, and it does not grow if you fade. If you run a perfectly even race you will see the same offset at 5 km as at 40 km, and only at the finish line do the two clocks stop together and reveal the gap as a single number.

A worked example: half marathon, chip 1:30:00, gap 1:47

Set the distance to Half (21.0975 km), the target chip time to 1:30:00 and the start-line gap to 107 seconds — 1:47, which is a plausible gap for a mid-field start in a large half. The even chip pace is:

p = 5400 s ÷ 21.0975 km = 255.95 s/km ≈ 4:16 per km

Each kilometre then costs 255.95 s, and the tail segment is whatever is left:

• 21 whole kilometres — 21 × 255.95 = 5375.04 s, printed 1:29:35.
• Tail, 97.5 m — 0.0975 × 255.95 = 24.96 s. The tail's row is a guide rather than a stopwatch mark, so it rounds to the nearest five seconds; its exact cumulative value is the target, 1:30:00.

The gun column is that same list with 107 seconds added to every row. So the race clock above the 10 km marker reads 10 × 255.95 + 107 = 2666.6 s ≈ 44:27, while your watch — if your watch is showing chip time — reads 42:40. Both are correct. The gun clock is not 1:47 slow; it is 1:47 ahead, because it started 1:47 earlier.

And the finish: gun clock 5400 + 107 = 5507 s = 1:31:47. That is exactly the number the official clock shows as you cross, and exactly the number results will print beside your 1:30:00 chip time.

The same two clocks on the other distances

Nothing in the method is specific to the half, so the page carries the same worked figures for every chip it offers. Each row is the chip time you keep and the gun time the finish clock reads for a 1:47 start-line gap:

• 5 km, chip 25:00 — even pace 5:00 per kilometre, whole-kilometre rows 5:00 / 10:00 / 15:00 / 20:00, no tail. Gun finish 26:47.
• 10 km, chip 50:00 — 5:00 per kilometre, rows 5:00 through 50:00, no tail. Gun finish 51:47.
• 10 miles, chip 1:20:00 — 16.09344 km, so the whole-kilometre rows run to 16 and then a 93.44 m tail. Gun finish 1:21:47.
• Half, chip 1:30:00 — the example above: 21 rows and a 97.5 m tail. Gun finish 1:31:47.
• Marathon, chip 3:30:00 — 4:59 per kilometre, 42 rows and a 195 m tail. Gun finish 3:31:47.
• 50 km, chip 4:10:00 — 5:00 per kilometre, 50 rows, no tail. Gun finish 4:11:47.

Every pace in that list is the target divided by the distance and nothing else, so you can redo any row with one division and one addition: time at mark k is k × pace, and the gun clock is that plus your gap. The tail row on the 10-mile and marathon distances is the only place a race distance is not a whole number of kilometres — which is the whole reason it exists.

The rounding rule, and what it guarantees

Every whole-kilometre row is the exact cumulative chip time rounded to the nearest second, so no row can be more than half a second away from the arithmetic. On this example the largest gap on the whole-kilometre rows is 0.4995 s, at the 11 km row — and the page reports the figure for whatever you enter, rather than quoting a generic bound. The finish row is printed with your target exactly, with no rounding at all, because it is the one row that has to be exact.

The tail row is deliberately different. It covers a fraction of a kilometre, you cannot see the exact point where the race ends without a measuring wheel, and printing it to the second would fake a precision you cannot act on. It rounds to the nearest five seconds and says so on the row.

Turning the gap into a corral position

A start-line gap is a measurement. A corral is a place. They are related by how many runners are in front of you and how tightly they are packed, and the page converts one into the other using a published benchmark rather than a guess.

The benchmark is the wave structure the race published for the 2024 and 2025 New York City Marathon. Read across those tables, wave 1 running the first 1,000 metres of the field gives a mean delay of about 4.2 seconds per metre of depth into the field, and each later wave adds about 140 seconds, mostly because the waves behind start from further back and are held longer. The conversion the page runs is:

depth = g ÷ 4.2  metres    then    waves = floor(depth ÷ 1000), remainder in metres

• g — your measured start-line gap in seconds. This is the only number you supply, and it is the number that matters.
• 4.2 s/m — the mean wave-1 density read off the published NYC Marathon wave tables for 2024 and 2025. It is a benchmark from those public tables, not a measurement of your race.
• 140 s — the delay each later wave adds, same source.
• depth, waves — outputs, both rounded as described.

Take the 1:47 gap above. 107 ÷ 4.2 = 25.5 m — a start inside twenty-six metres of the line, which is wave 1 and the front of the field. Take a much later start, 480 seconds: 480 ÷ 4.2 = 114.3 m, still inside wave 1. Push it to 900 seconds and you get 214 m, still wave 1. The conversion only starts producing later waves once a gap gets large, because the density figure is what does the work: 1,000 metres of field at 4.2 s/m is 4200 seconds, which is over an hour of field depth.

That number — 4200 s, or seventy minutes — is the honest scale check on this whole section, and it is worth saying out loud. A one-wave field of a thousand runners, starting in a block, spans more than an hour of gun-to-line delay. Nobody is at the back of a race like that for a small gap. Which means: the corral estimate is a sanity check, not a map. It tells you whether your measured gap is the right order of magnitude for where you were standing. It does not know your race's corral layout, and the page does not pretend it does.

Supported range and limits

• Distances: 5 km, 10 km, 10 miles, half marathon and 50 km as one-click choices, plus a marathon chip and any custom distance from 0.1 to 100 km (about 0.06 to 62 mi). The custom field takes either unit.
• Time: 5 minutes to 47:59:59, entered as hours, minutes and seconds. Minutes and seconds must be 0–59.
• Start-line gap: 0 to 3600 seconds. Zero is a legitimate entry — it is what a chip-only race or a one-wave start looks like.
• Segmentation: whole kilometres with a tail for the remainder. The marathon therefore ends on a 195 m tail, not a fictional 43rd kilometre, and a 5 km race has no tail at all.
• Pace display: switchable between km and miles for the pace note. The table itself stays a kilometre table — see the next section for why.
• Cost and privacy: none and none. No account, no email, no paywall, no request to a server; every number is computed in your browser and nothing is stored.
• Works offline once the page has loaded, and on a phone.

What this page deliberately does not do

• No negative splits. The splits are even. A negative split is a shape, not a target, and printing one needs a model of how much slower your first half should be — which is a claim about you that this page cannot source.
• No hill adjustment. The course is assumed flat. The marathon split calculator already does per-kilometre elevation adjustment from a published energy-cost model, and repeating that machinery here under a different headline is how a site starts competing with itself.
• No mile-split table. The table is cut at whole kilometres on every distance, including the 10-mile one. A mile-split variant of the same table would be a near-duplicate of this page, so there is one kilometre table here and one job per page.
• No pace band. A band is a paper strip with one column on it; this page's whole point is two columns, so a band would throw away the thing it is for. The half marathon split calculator prints bands.
• No live race corral data. The corral estimate uses the published NYC Marathon wave tables described above as a density benchmark. It does not know any race's actual corrals, and the per-race wave tables for your race may differ.
• No weather, no age grading, no fatigue model. All three need published tables this page does not have, so it leaves them out rather than filling the space with a coefficient nobody can check.

Questions

What is a splits calculator?

A splits calculator turns one number — the finish time you are aiming for — into the time you should be at each mark of the course. This one also carries two clocks rather than one, because in a chip-timed race the clock above your head and the time you actually ran are not the same number.

What is the difference between gun time and chip time?

Gun time runs from the starting pistol to the moment you cross the finish line. Chip time runs from the moment your timing tag crosses the start line — which in a big race is seconds or minutes after the gun, because you were behind the line and had to walk or jog to it. Chip time is the time you keep and the one results are normally ranked on. Gun time is what the race clock reads when you cross the line, and it is the number you see on the clock at every kilometre marker.

Why does my gun time look so much slower than my chip time?

Because you did not start running when the gun went off. The gap equals the time it took you to reach the start line from where you were standing, and it carries through the whole race unchanged. Enter that gap on this page and the two columns differ by exactly that amount at every checkpoint, all the way to a finish where the gun clock stops at the gap plus your chip time.

How many corrals away from the front am I?

The page estimates it from a benchmark you can check: in the published 2024 and 2025 New York City Marathon wave tables, wave 1 over a 1,000-runner field gives a mean delay of about 4.2 seconds per metre of depth into the field, and each later wave adds about 140 seconds. Dividing your measured gap by 4.2 gives a depth in metres; if that lands beyond 1,000 metres the remainder is charged in whole waves at 140 seconds each. It is a sanity check on 'am I where I think I am', not a race's real corral map.

Which distances can I plan for?

5 km, 10 km, 10 miles, half marathon, marathon and 50 km as one-click choices, plus any distance from 0.1 to 100 km entered by hand. Each one is cut into whole kilometres with the remainder as a short tail segment, so a half marathon ends with a 97.5 metre tail rather than a made-up full kilometre.

Does the splits calculator do negative splits or hill adjustments?

No. The splits are even — one pace repeated — and the course is assumed flat. There are already two pages on this site for the other two jobs: one takes a per-kilometre elevation profile and adjusts every split for the climbing, and another turns a half-marathon target into a printable pace band. Adding a third variation here would just repeat them.