Code: Select all
Finish Time Average Split Avg split (% of total time) Num runners Num neg split % neg split
<3:00 0:05:02 2.92% 28 4 14.29%
3:00-3:20 0:06:46 3.53% 101 5 4.95%
3:20-3:40 0:09:39 4.25% 239 9 3.77%
3:40-4:00 0:11:53 5.14% 334 13 3.89%
4:00-4:30 0:16:57 6.66% 411 16 3.89%
4:30+ 0:24:24 9.94% 508 5 0.98%
So at the very pointy end of the field, there is the highest rate of negative splitting - but it's still only 14%. From 3:00 on, it's incredibly rare, 1 in 20 or 30 runners managed it.
Less than 1% of 4:30+ finishers negative split the race.
Now it of course must be noted that this marathon has a more challenging second half (downhill first 1/2, mostly flat second 1/2). But if you look at the percentages of the total time that the average person split, it's quite high - and goes up as the total time goes up.
So when planning a PB, maybe assuming a positive split isn't such a bad idea? Otherwise you're gambling that you're going to be in a very elite <5% of runners who manage the negative split...
jono