A *positive split* (red) is when the second half is slower than the first (e.g., 2:00 first half vs 2:30 second half). A *negative split* (green) is the opposite. We defined *even splits* as halves within ±2% of each other (yellow).
Each point shows a runner’s first‑half vs second‑half time from the 2025 Richmond Marathon. The fastest finishers (Boston‑qualifying times, blue) cluster tightly near the even‑split diagonal, while the overall field shows a much wider distribution with many large positive splits
Zigxy on
Oof
Someone ran 1:29 first half and 3:06 second half.
I’ve had (non marathon) runs like that. Must have felt like poop.
pocketdare on
Actually even for many of the slower times overall, the relative drop off in the second half (as a percentage) doesn’t look that much worse than those at the faster end. But I could be wrong (just eye-balling it). Like if you drew an actual best fit line would the deviations (r squared) be that much worse for slower runners than faster runners?
rollem on
I was in that race! I’ve been trying to BQ for 2 years now. Unfortunately I didn’t this time and I think I can see my dot….
![[OC] 2025 Richmond Marathon Split Times: A Tight Band of Even Pacing Among the Fastest Finishers](https://www.byteseu.com/wp-content/uploads/2025/12/f0vl97w7s76g1-1536x864.png "[OC] 2025 Richmond Marathon Split Times: A Tight Band of Even Pacing Among the Fastest Finishers")
4 Comments
* Data source: direct from listing site [https://www.richmondmarathon.org/results/year-by-year-results/](https://www.richmondmarathon.org/results/year-by-year-results/)
* Tools: R and ggplot2
Full analysis: [https://rivercitydatascience.com/analysis/rva-marathon/](https://rivercitydatascience.com/analysis/rva-marathon/)
A *positive split* (red) is when the second half is slower than the first (e.g., 2:00 first half vs 2:30 second half). A *negative split* (green) is the opposite. We defined *even splits* as halves within ±2% of each other (yellow).
Each point shows a runner’s first‑half vs second‑half time from the 2025 Richmond Marathon. The fastest finishers (Boston‑qualifying times, blue) cluster tightly near the even‑split diagonal, while the overall field shows a much wider distribution with many large positive splits
Oof
Someone ran 1:29 first half and 3:06 second half.
I’ve had (non marathon) runs like that. Must have felt like poop.
Actually even for many of the slower times overall, the relative drop off in the second half (as a percentage) doesn’t look that much worse than those at the faster end. But I could be wrong (just eye-balling it). Like if you drew an actual best fit line would the deviations (r squared) be that much worse for slower runners than faster runners?
I was in that race! I’ve been trying to BQ for 2 years now. Unfortunately I didn’t this time and I think I can see my dot….