The Boston Marathon Is Fast. Training Data Shows Why. 

In our new monthly column, in partnership with Strava, we’ll take a deep dive into interesting data points that reveal the more human side of sport. 

The Boston Marathon is fast. How fast? A new look at more than 100,000 Strava uploads reveals that the field at the Boston Marathon is significantly faster than the New York City Marathon, and that runners train differently once they secure that Boston qualifier (BQ).

This year, 53 percent of the runners in the Boston field uploaded their data to Strava, for around 16,000 uploads. (Strava or it didn’t happen, right?)

With the help of analysts from Strava, we dug into the data from the Boston and New York City marathons to see how the two compared. To compete at the Boston Marathon, the majority of runners (with the exception of 5,000-7,000 runners who qualify via fundraising) must hit a qualifying standard, a benchmark that just over 10 percent of all marathoners reach. That means many of the athletes lining up in Hopkinton have run at least one marathon previously (with the exception of the charity bibs) to qualify, and have finessed their pacing over time.

Roughly 55 percent of runners at Boston finished in 3:30 or faster, compared to just 20 percent of the field at NYC. Almost three-quarters of Boston runners (73 percent) paced for a sub-four-hour finish, compared with just 42 percent of runners at NYC. Also, 47 percent of runners at Boston BQ’d, re-earning their spot on the start line in 2024, compared with just 7 percent of runners in NYC.

Boston Runners Are More Experienced

But why are runners at Boston so much faster? The course is notoriously hilly and not exactly PR friendly. When you look at the percentage of the field that achieves a negative split time, meaning they run the second half of the race faster than the first, the importance of experience emerges in the data.

At this year’s Boston Marathon, Strava data showed that 13 percent of finishers ended up negative splitting the course, up from less than 10 percent in 2022. Negative splitting Boston, with its punishing downhills on the front half and grindy gradients on the back half is no small feat, and a negative split is often the gold standard for a well-executed marathon.

The Berlin Marathon, a notoriously fast course (and home to Eliud Kipchoge’s 2022 world record run of 2:01:09) with a net downhill, saw just 16 percent of the field negative split in 2022. In addition, 11 percent of the fields at London and Chicago—still relatively fast, PR-friendly courses—negative split, and just 10 percent of Tokyo negative split as well. The heat and humidity at NYC tore the field apart, with only 3 percent of the NYC field negative splitting.

This tells us that Boston runners, relative to the difficulty of the course, are superior pacers, likely because of the experience required to qualify for the race.

Boston Runners Train More

For most runners, Boston is not their first rodeo. Not only do they tend to race smarter, but their training looks different, too. The median peak week for Boston runners was a 53-mile week and 7:31 of total training time. Compare that with a 37-mile median peak week and 5:49 for London Runners, 39 miles and 6:19 hours for NYC runners (note: we use medians to control for outliers and erroneous data).

This isn’t revelatory—due to stringent qualifying standards, Boston skews fast, and we generally see that faster runners tend to train more—being a more experienced runner tends to mean you have built capacity for higher training volume, and that higher volume and longer experience tends to make you faster. But, we can control for this in the data by comparing runners of different pace groups across marathon majors. When we do this, Boston runners still train more.

When you compare the relative finish times across all marathon majors, Boston runners trained for more hours per week (8:10) and ran more frequently for a median of five times a week for all finishers under 3:30, dropping to four runs a week for 3:30-4:00 finishers, and three times a week for 4-plus hour finishers. One standout data point is that runners who ran a sub-three hour marathon at any marathon major had a median five times a week run frequency. Regardless of pace, runners toeing the line at Boston have a higher run frequency across their entire training cycle than other marathon majors. Basically, the faster the athlete was, the higher their run frequency tended to be.

For another comparison, we looked at the training runners did for the Manchester Marathon (which takes place the day before Patriots’ Day and is the second-largest marathon in the UK after London). Looking at the 16-week build-up for Boston, compared with Manchester runners, Boston athletes logged more volume, completing an average of five and a half hours each week running, compared to four hours for Manchester competitors (this excludes an assumed two-week taper before race day). Over half (56 percent) of runners at Boston hit at least 50 miles in a week in their training build-up, compared to just 20 percent of Manchester runners. Almost 80 percent of Boston runners hit a 40-plus-mile week, compared to 42 percent of Manchester runners.

One reason Boston runners racked up more volume is due to logging longer long runs. About 74 percent of Boston participants had completed at least one 20-mile run in training, compared to 62 percent of Manchester runners. Also, 93 percent of Boston runners notched at least a 16-mile run in their marathon build, compared with 87 percent of Manchester runners.

Boston Runners Don’t Necessarily Run Longer, But They Hit the Hills

Lastly, the median long run didn’t vary much across marathon majors, even while weekly volume varied across races. Most runners logged a median longest run of between 17 to 20 miles. That makes sense with the standardized distance, but one major difference between the training approaches of Boston runners and other majors was this: hills.

To get ready for the notorious Newton Hills, Boston runners who ran between 3:00-3:30 logged a median of 1,047 feet of gain per week, compared with just 725  median feet per week logged by athletes training for flatter courses like Berlin, 732 for Tokyo, or 739 for Chicago. 2:30-sub 3:00 finishers notched a median of over 1,300 per week in training. Now those are some hill repeats!