Ultramarathon Pacing Strategy: What 18 Years of Western States Data Shows

This past weekend was the Western States 100. If you are not familiar, this is the Super Bowl of ultramarathons in the United States. It is the race most ultra runners dream about. I spent the weekend refreshing the live tracker more than I probably should have.

So this felt like the right week to dig into a study that just analyzed 18 years of Western States data. Not a training intervention. Not a "do this and run faster" study. Just a detailed look at how thousands of finishers actually paced one of the hardest 100-mile races in the world.

The Research

Citation: A 2025 study published in Scientific Reports analyzed official results and checkpoint splits for every Western States finisher from 2006 to 2023.

Researchers pulled data on 3,837 runners across 18 race years, with elevation data and 17 checkpoint times for each one. They calculated how each runner's pace changed at every checkpoint relative to their overall race pace, then grouped runners by gender, age, and performance level to see how pacing patterns differed.

I want to be upfront about what this study is and is not. This is purely observational. There was no intervention, no controlled variable, no if-then conclusion to draw. The researchers collected data and described what happened. That makes this an educational piece, not a prescriptive one. We can learn from the patterns without treating them as proof of what works best.

What They Found

Every single performance group showed positive pacing. Nobody negative split this race. Everybody slowed down from where they started. The question was never whether runners would slow down. It was how much.

The front of the pack paced the most consistently. These are largely professional and highly experienced ultra runners who know the course and have raced 100 miles before. They got out fast and stayed fast longer than anyone else.

The back of the pack also paced consistently, just at a slower overall speed. They showed less variability between checkpoints than the middle of the field.

The middle of the pack showed the most erratic pacing and the largest slowdown of any group. They faded the hardest.

Men showed significantly more pacing variability than women across almost every age group. Women paced more consistently throughout the race.

What This Means For Your Training

Here is where I want to be careful, because the study does not answer the question everyone wants answered: is this optimal? Is it actually a mistake for middle-of-the-pack runners to go out fast and fade, or is that simply what happens at this distance regardless of strategy? The data shows what happened. It does not prove what should have happened. Some people will argue the middle-of-pack fade is a pacing error. Others will argue it might not matter much over 100 miles. That was not what this study was built to settle, and I am not going to pretend it was.

What I will say is this: a lot of the men in the middle of the pack going out fast looks like ego more than execution. Getting swept up in the start line energy, running with people who are objectively faster than your training supports, and paying for it in the back half. That late-race collapse is also what running economy durability research from Issue 005 tried to measure: how well you hold form and efficiency when fatigue sets in. Putting the ego aside and running your own race, especially in a hundred-mile event, is not a new idea. But seeing it show up this clearly in 18 years of data makes the point land differently.

A friend of mine, also an ultra runner, said something to me once that has stuck. In most races we talk about your fastest splits. What are your best miles. In an ultramarathon, what matters more is your slowest mile. Not because the fast miles do not count, but because the size of the gap between your fastest and slowest mile tells you more about how well you managed the day than any single split does.

My Own Numbers

I have real data on this, not just a rough memory. At my first Ice Age 50 miler, my fastest mile was 7:40. My slowest was 26:33. That is a delta of nearly 19 minutes between my best and worst mile of the day.

Fastest mile

Year 1