Eastern States 100 - 2017 Analysis

Eastern States 100 - 2017 Analysis

In-Depth Look at Eastern States 100 - 2017

With Eastern States 100 being my biggest race of the year (and of my life, to date), I figured I'd give it a hard look at some of the stats and information I was able to gather about it.

Because the race opted to use UltraLive to track participants throughout the race, I was able to scrape data from their site and compile it here. It should be noted that the times are recorded by someone that was sitting at the start of the aid stations and asking for a bib number to record. That means if a runner forgot to check in, or didn't remember until they were leaving, then their time is reflected accordingly. I have no control over that, nor any way of knowing.


First thing to look at is the category of who started the race but then did not finish (DNF), and those that did finish (Finished).

Finish Statistics

Of the 204 runners that started, 121 finished and 83 DNF'd. This is a huge improvement from the previous years where the finishing rate was around 30%. It's worth noting that there were 27 DNS entries. The conditions were much more favorable this year than last year where the temps were up in the 80's during the day followed by torrential downpours and a tornado in the evening.

This may also mean that word of the dismal finishing rate became more widespread, meaning those applying for the race were more prepared than years. This is just speculation as it is a new-ish race, but there are also a lot of different factors that affects a person's decision to drop or continue. I don't have a good comparison other than looking around at other sites for their finish stats, and what I hear, but it seems a rate of 60% finishers is more common and typical of 100-mile distances.

Mapping and Overview

While I wish I had better map plotting skills... there are others that are much better. With that, below is an interactive map put together by the ES100 crew. It's a good overview, and gives an idea of how far apart the different aid stations are and elevation involved.

When I was plannign for the race, I wanted to know where the hard sections were, so I grabbed the total elevation gain in each section. You can see what that looks like in the elevation chart above. Below is a slightly different form, with the colors denoting the amount of climbing between each aid station (also shown in the height of the bar). Then with that knowledge, we can use a heatmap to visualize how paces differed from section to section.

Vertical Gain per Section

NERD NOTE (as if this post wasn't nerdy already) I averaged the paces going into each section based on a runner's finishing decile group. Decile group meaning if the runner had a top 10% time, they were in group 1. Any runner that DNF'd was still given a place (and consequently a decile group), which is why there's a few blank sections. Ranking was based on finishing place, then those that DNF'd were given a place based on the maximum distance first.

Heatmap showing Speed per Section

This is telling us that some runners covered some sections quickly, and sections were demonstrably slow. A valid assumption from this is the section from RAMSEY into LOWER PINE BOTTOM is pretty tough, and SLATE RUN to ALGERINES is extremely tough. Expect that to be the slowest section of the course! Another slow section for all groups is BLACKWELL to SKYTOP.

A quick note on the section at SLATE RUN. Unless you happen to be in group 1 or 2, more than likely, most runners reach that aid station by night fall. That means whatever difficulties that section brought on originally in the day, that should be magnified by the darkness. If I remember correctly, that section was not very pleasant anyway... not that I was there in the daytime, I just remember it being tough.


Just looking at finish times, we can see how long people were on the race course, and where they finished.

Time Elapsed by Place

A little subtle, but if you look at the 30th finisher, the time coincides with 30 hours. There's a slight but noticeable jump in finish times. This is likely from runners understanding they will have chance to get under 30 hours, and therefore push themeselves a little harder to beat a benchmark time. A quick summary of finishers by gender.

Gender Distribution (Finishers)
16 105

We always need a good histogram. Here's one to show when runners finished, and then how long those that DNF'd were out on the course.

Runners Histogram


But we can go deeper. I found a different chart style, called a beeswarm, so let's use that. This will require a bit of explanation (not always great for a chart, but stay with me). This is grouped by where the runners are ranked overall amongst everyone that started (not just finishers). It's a nice way of showing the distribution amongst runners within each grouping.

By the time you get to the 6th group (top 60%), there are not only runners that finished, but also start seeing those that DNF'd. Group 7, 8, 9, and 10 are all runners that did not finish, so the time elapsed is decreasing for those groups since runners were on the course for less and less time.

Beeswarm of all runners

Now that you've been stung, let's see where those that DNF'd stopped.

DNF Analysis - Where runners DNF'd and when

My heart goes out to the 3 three runners that were on the course until 9 AM the next day, and made it to Skytop before they hit a cutoff (or chose to call it quits). It looks like a lot of runners called it good at Halfway House just before midnight.


Remember that the times recorded here are by volunteers as every runner came into the different aid stations. That means some people were given wrong times or maybe were just missed.... that happens. Below is a runner's pace over the course of the race. Notice how everyone gets just a little bit slower. Wouldn't it be nice if we all got faster?? That doesn't happen.

I am the runner highlighted, FYI.

Pace over time for each runner

This last graph is my favorite. It shows how a runner's place changed over the race. Some runners started off really fast and stayed there, others started off a little too hot and couldn't hold it... then there are those that really started slow but had a great day and moved up a lot.

Below is just the top group, but I have the data for other groups. I may show all the groups if there's some interest, but for now, I'll keep it to this top grouping.

Top 10% place into each aid station

If you read my text, thanks for reading. If you looked at my graphs, I hope you understood and enjoyed it. Let me know if you have any questions or suggestions.

About Tomas Castillo

I'm an ultramarathon runner, and outdoorsman when I can, while keeping bees and playing with data on the side. I plan on sharing my explorations and adventures here.