July 20, 2026

Marathon Pacing Strategy: Building a Split Table That Survives Mile 20+

I’m a Korean medicine doctor in Vancouver by day, a software engineer the rest of the time, and a runner in whatever hours are left over — Ryutility is a small side project I build under R-Lab, part of Ryu Clinic Inc. I’ve watched a lot of runners walk into my clinic the week after a marathon convinced something went wrong physically at mile 22 — and in most of those conversations, what actually went wrong was a pacing plan built for the first 20 miles with nothing built in for the last 6.2.

A marathon split table isn’t just arithmetic dividing your goal time by 42.195 km. The honest version of that table has to plan for the fact that most runners — even well-trained ones — physiologically slow in the final quarter of the race, and a plan that doesn’t budget for that isn’t a plan, it’s a best case.

Even splits vs. negative splits, precisely defined

An even split plan runs every segment at the same pace — the goal time divided evenly across the distance, with no adjustment.

A negative split plan runs the first half slightly slower than average and the second half slightly faster, so early kilometers “bank” a small time cushion and later kilometers spend it. The standard implementation grades this linearly by how far into the race a segment sits — the earliest segments run slowest, the pace gradually rising (getting faster) toward the finish, symmetric around the midpoint:

Segment pace factor = 1 + strategy% × (1 − 2 × segment_midpoint / total_distance)

At the halfway mark, the factor is exactly 1 (average pace). Before halfway, the factor is above 1 (slower than average); after halfway, it’s below 1 (faster than average) — and the whole thing is normalized so the total time still matches your goal exactly.

A worked example

Say your goal marathon time is 4:00:00 (14,400 seconds) and you’re running a 5% negative split, viewed in 5 km segments.

At the very first segment (midpoint ≈ 2.5 km of 42.195), the factor is roughly 1 + 0.05 × (1 − 2×2.5/42.195) ≈ 1 + 0.05 × 0.882 ≈ 1.044 — about 4.4% slower than dead-even pace.

At the last segment (midpoint ≈ 40 km of 42.195), the factor is roughly 1 + 0.05 × (1 − 2×40/42.195) ≈ 1 + 0.05 × (−0.896) ≈ 0.955 — about 4.5% faster than dead-even pace.

In pace terms, if your even-split pace is 5:41/km, a 5% negative split has you starting around 5:56/km and finishing around 5:26/km — a meaningful, deliberate gap between your opening and closing kilometers, not an accident of fatigue.

Why even splits still fail for most first-timers

Even splitting is the theoretically “optimal” strategy for a runner with perfectly stable physiology across the full distance — but that assumption rarely holds for anyone without extensive marathon-specific training. Glycogen depletion, dehydration, and muscular fatigue accumulate non-linearly, meaning the effort required to hold a constant pace rises through the second half even when the pace number stays flat. An even-split plan on paper often becomes a slow positive split in practice, because the runner can’t actually sustain constant effort at constant pace once fatigue compounds.

A small negative split is a buffer against exactly this. It isn’t primarily about strategy or racing tactics — for most non-elite marathoners, it’s a hedge against the predictable physiological reality that the last 10K is harder to hold pace in than the first 10K, regardless of training quality.

How much of a negative split is reasonable

Elite marathoners often run close to even splits or very mild negative splits (1–2%), because their training volume makes late-race fade smaller and more predictable. For most recreational and sub-elite runners, a 3–5% negative split gives a realistic cushion without sandbagging the first half so much that you’re leaving fitness on the table. Going much beyond 5–7% usually means you started too conservatively rather than executing a real strategy.

Using it in practice

Start with a realistic goal time from the Race Time Predictor — don’t build a split table around a time you haven’t validated against a recent result. Then use the Splits Calculator to generate the actual per-km or per-mile split table for that goal time, choosing a 3–5% negative split strategy unless you have specific evidence from prior marathons that even pacing works for you. Run the early splits conservatively on race day even if they feel too easy — that feeling is the plan working, not the plan being wrong.