A target time and a distance are enough to produce your full split chart. The calculator gives two: the even version, and the negative split version, where the second half is run two percent faster for an identical total.
The second column is the one to look at. It matches how good races actually get run, and it's almost always easier to hold than the first.
Pace /km—
Your splits
| Distance | Even | Negative split |
|---|
The negative split column starts 2 percent slower and finishes 2 percent faster, for the same total. That is how most good races get run.
Why the second half should be the faster one
This isn't an elegance reserved for advanced runners, it's the arithmetic of cost.
Starting too fast doesn't only cost the seconds you gained: it brings forward the moment glycogen runs low and acid production outpaces your ability to clear it. The first two kilometres taken ten seconds too quick are rarely paid back at ten seconds. They're paid at three or four times that, in the final quarter, when there's nothing left to do about it.
The empirical side points the same way. Over the marathon, nearly every world record has been set on an even or slightly negative split. Among amateurs, the second half is typically five to ten percent slower than the first, which is exactly the reverse. The difference isn't talent, it's the start.
The two percent, and why not more
The negative column starts two percent slower and finishes two percent faster. That figure isn't arbitrary.
Below one percent, the gap is smaller than the natural variability of terrain and crowd: the chart becomes indistinguishable from an even split. Beyond four percent, you're no longer running a negative split, you're recovering from a start that was too slow, and leaving time on the table.
Two percent, on a 3:30 marathon, means a first half in 1:46 and a second in 1:44. Six seconds per kilometre of difference, which feels like nothing at the start and like a comfortable margin at thirty kilometres.
What the chart doesn't tell you
It assumes flat ground, neutral weather and a race without surprises. None of those three exists.
Elevation changes everything and doesn't compensate symmetrically: a climb costs more than the equivalent descent returns. On a hilly course, the way to read the chart is to aim at even effort instead of even pace, and to accept slower splits uphill without trying to claw them back.
Wind gets paid for twice: into your face it slows you, and behind you it rarely returns the equivalent. The best move is running sheltered in a group when you can, and forgetting the chart on exposed sections.
Aid stations cost five to fifteen seconds each. Over a marathon, eight of them make two minutes, which is more than the gap between the two columns. The chart gives race splits, not stopwatch splits.
How to use it on race day
Three uses, in order of importance.
The first is disciplining the opening three kilometres, and that's where the chart earns its place. Everybody is fresh, the crowd pushes, the pace feels easy. It's the one moment in a race where you should trust the number against the sensations, because the sensations of kilometre 2 always lie in the same direction.
The second is knowing what to do with a lead. If you cross halfway two minutes ahead of the chart, that isn't good news, it's a warning. The answer is returning to the planned pace, not holding the rhythm.
The third is deciding when to go. At three quarters, if the pace is coming out easy and the breathing follows, the chart has done its job and you can leave it. That's the only place where improvising pays.
Pace references by target
| Target | Pace per km | Pace per mile | Halfway split | Finish |
|---|---|---|---|---|
| 3:00 marathon | 4:16 | 6:52 | 1:29 | 3:00 |
| 3:30 marathon | 4:59 | 8:01 | 1:44 | 3:30 |
| 4:00 marathon | 5:41 | 9:09 | 1:59 | 4:00 |
| 4:30 marathon | 6:24 | 10:18 | 2:14 | 4:30 |
The halfway splits are those of the negative column, so slightly beyond half the target time. That's deliberate: they describe a cautious start.
One useful check before choosing a target: the pace you aim for should be a pace you have already held on a long run, not one you hope to discover on the day.
The GPS trap
Your watch and the timing mat don't measure the same thing, and the gap always runs the same way.
A watch reads long, by one to two percent on a certified course. The causes: tangents nobody cuts perfectly, signal error between buildings, tunnels. Over a marathon that's four to eight hundred extra metres displayed, and therefore an average pace that looks better than the official one.
The practical consequence: take your splits from the course markers, not from the watch's automatic laps. And read the pace of the completed lap rather than instant pace, which jumps twenty seconds per kilometre at every change of street and is useless.
FAQ
Should I run an even split or a negative split?
A slight negative split, around two percent, for almost everyone. It costs a few seconds at the start and it protects the final quarter, which is where races get lost. A perfectly even split is the second best option.
What do I do if I'm ahead at halfway?
Return to the planned pace immediately, without trying to keep the lead. A lead at halfway in a marathon is almost always a debt, and it gets repaid at an unfavourable rate after thirty kilometres.
How do I adapt the chart to a hilly course?
By aiming at even effort instead of even pace. Uphill splits will be slower than the chart, descents slightly faster without compensating, and the total will come in slower than on the flat. Attempting otherwise is the most reliable way to finish at a standstill.
Why does my watch show a longer distance than the race?
Because GPS reads long by one to two percent: imperfect tangents, signal error, tunnels. It's normal, it happens to everybody, and it explains why your watch average pace looks better than your official one.
Should I write my splits on my arm?
It's still the most reliable method, because it depends on no battery and reads in a second. Four or five markers are enough: the quarters and halfway. A full chart at every kilometre is unreadable while racing.
How much do aid stations cost?
Five to fifteen seconds each depending on the crowd and how you drink. Over a marathon, eight of them make two minutes, better built into the target than discovered at the finish.
Is the wall at thirty kilometres unavoidable?
No, and its main cause is the start, not the distance. A marathon begun five seconds per kilometre too fast depletes reserves earlier, and that's what produces the sensation of a wall. Sensible pacing and consistent fuelling make it avoidable for most people.
Can I use the chart for a trail race?
Not as it stands. On terrain with elevation and technical sections, pace per kilometre stops meaning much and the useful reference points are time per section and perceived effort. The chart stays useful on the runnable parts.
Does the chart work in miles?
Yes, and that's what the Metric or Miles switch does. In miles it places a marker at every mile up to 10K and every five miles beyond, the same logic as the metric version, with the pace shown per mile. It's the format a US pace chart is normally read in.
A split chart is not a build-up
It says how to distribute a target time. It doesn't say whether that target is within reach, and that's the question that decides the day. The Vitality Archetype test measures your profile across 5 dimensions in 27 questions and 3 minutes, then generates a training week calibrated for your recovery and your available time, with a progression that follows your session feedback. The chart serves on race day. The week serves the three months before it.