For the past nine months, I’ve tracked data about the number of finishers and qualifiers at marathons around the world to project the likely cutoff time for the 2026 Boston Marathon. With the qualifying period winding down, that projection has stabilized around 5:30.
A common – and fair – critique of the methodology behind this tracker is that it treats all results the same. This means that big swings in the results from large marathons – like Berlin – could influence the outcome. It also led me to exclude the data from the Tokyo Marathon, because historically it has had a much lower conversion rate than the other Majors.
After analyzing the conversion rate of Boston qualifiers to applicants, I’ve identified four main variables that correlate with a runner’s likelihood to apply: their qualifying race, their buffer, their age, and their gender. The dashboard below enables you to play with the weights of those variables and predict the number of applicants for the 2026 Boston Marathon.
Keep reading below for more details on how the dashboard works and some context. But when the dust settles, we will likely see more applicants this year than in 2024 and fewer applicants than in 2025.
How Does the Dashboard Work?
The dashboard is built on the same dataset as the Boston Marathon Cutoff Time Tracker. The only difference is that this includes the results from all races in the 2025 qualifying period, and it only includes results through the weekend of August 31, 2025 for the 2026 qualifying period.
The dataset includes races in North America with more than 200 finishers, along with London, Berlin, Tokyo, and a handful of small last chance races.
The dashboard includes four numbers per qualifying period: Finishers, Qualifiers, Weighted Qualifiers, and Applicants.
The Finishers statistic is a count of the number of runners who have finished a marathon in that qualifying period. To the extent that it is possible to do so, a runner with multiple results are grouped together and is only counted once.
The Qualifiers statistic counts the number of those Finishers where the runner’s best time met their qualifying standard for the Boston Marathon. The 2025 qualifying period applies the old qualifying times and the 2026 qualifying period applies the new qualifying times.
The Weighted statistic applies a variety of weighting factors (explained below) to modify the number of qualifiers.
The Weighted count from the 2025 qualifying period is used to calculate an overall conversion rate to achieve the actual number of applicants for the 2025 Boston Marathon (36,393). This conversion rate is then applied to the Weighted count from the 2026 qualifying period to predict the expected number of applicants for the 2026 Boston Marathon.
Note that the results set for the 2026 Boston Marathon are currently incomplete. Additional results will be added as the final races are completed. The final race of the qualifying period is 9/12/2025. Until that happens, this prediction should be seen as a floor – and the final prediction will be slightly higher.
What’s With the Chart?
The projection graph in the bottom right corner of the dashboard serves two purposes. It validates the assumptions made about the conversion of qualifiers to applicants and it offers a prediction of the likely number of applicants – and the cutoff time – for the 2026 Boston Marathon.
Based on the calculations in dashboard, the graph displays the number of expected applicants at any given buffer. The x-axis is marked in 60 second increments, from 0 (a runner’s exact BQ, at the right end) to -1500 (BQ-25 Minutes, at the left end).
The y-axis tracks the expected number of applicants, and the horizontal reference line marks 24,000 applicants – the approximate number of accepted applicants for the 2025 Boston Marathon. The expected cutoff time is the point at which this line intersects with the curve of anticipated applicants.
The three vertical lines mark statistics released by BAA when they announced the acceptances for the 2025 Boston Marathon. 6,971 applicants had a buffer of 20+ minutes and 18,170 had a buffer of 10+ minutes. The third line marks the actual cutoff time – 6:51 or 411 seconds – at which point there were 24,043 applicants.
These three points serve as a way to validate that the assumptions and calculations in the dashboard are plausible. You could enter many different values for the weighting factors and still end up with an accurate prediction at a single point. But if you enter values at random, you’ll likely have diverging results at or two of these points.
If you play around with the weighting factors, make sure that the three points of intersection – between the vertical reference lines and the blue curve – match up fairly well with the actual data. If that is the case, you can be fairly confident that the prediction – the intersection of the horizontal reference line and the orange curve – is also fairly accurate.
Again, keep in mind that the results from the 2026 qualifying period are not yet complete. Once the final weekend (9/6-9/7) is complete, a significant number of new results will be added – increasing the projected number of total applicants and moving the orange curve slightly higher.
Which Variables Correlate with a Runner’s Conversion Rate?
The four variables that I’ve identified which correlate with a runner’s conversion rate are: their qualifying race, their buffer, their age, and their gender.
In this analysis, I matched data from the 2025 qualifying period to the results of the 2025 Boston Marathon to estimate the conversion rate of runners from each race.
The majority of North American races had a conversion rate of approximately 35-40%, with a broader range of 30-60%. This included the three American majors – Boston, Chicago, and New York – along with most of the other large races.
A small subset of races had a conversion rate above 60%. This subset included downhill races (i.e. REVEL and tunnel races), last chance races (i.e. Erie, Beantown, and Last Chance BQ Grand Rapids), and some other small races.
The other Majors all had a lower conversion rate. Berlin and London each had a conversion rate of approximately 18-19%, about half that of the American Majors. Tokyo was well below that, with a conversion rate of approximately 12%. The 2023 Sydney Marathon (which wasn’t yet a Major) was even lower. Most other international races had a conversion rate below 10%.
In this analysis, I used the same dataset to explore some other variables.
A runner’s buffer has a big impact on their likelihood to apply. Generally speaking, the further a runner is below their BQ, the less likely they are to apply. Runners with a 5-10 minute buffer had the highest conversion rate and that declined as runner’s got faster.
This particular analysis didn’t produce any data on runners with a 0-5 minute buffer, but in exploring the data last year I’m fairly confident that this group had a slightly lower conversion rate than runners with a 5-10 minute buffer. This is because runners on the cusp, especially runners with a 1 or 2 minute buffer, should assume that they won’t get in – and therefore some of them will simply not apply.
A runner’s age also had a relationship to their conversion rate. Generally speaking, runners under 50 were less likely to apply than runners over 50.
Finally, gender showed a small correlation as well. Even after disaggregating by buffer and age, women were slightly more likely to apply than men.
How To Use This Dashboard?
I’ve assigned default values to each of the weights on the dashboard, based on my analysis of the conversion rate of qualifiers to applicants. But you can adjust these values to test your own theories.
For the weight by races, you can assign a weight to the High Conversion races (races with > 60% conversion rate) and to each of the Majors. The remaining races have a default weight of 1.0. The values you enter here are not the expected conversion rate of that particular race – it’s how you think that conversion rate relates to that of a standard North American race.
There are seven values to weight by buffer. You can assign an individual weight to each of those categories. I’ve assigned a default of 1.0 to the 5 to 10 minute group, and the other weights are relative to this group.
There are four values to weight by age, based on four broad age groups. I’ve applied the same values for Under 35 and 35-49 (default: 1.0) and for 50 to 59 and 60 and Over (default: 1.25). But you could nuance that if you’d like.
Finally, there’s one value for weighting by gender. Men are assigned a default of 1.0. If you don’t want to weight by gender, change the Women’s value to 1.0. Otherwise, assign a relative weight (default: 1.1) to account for the greater likelihood of women to apply.
Over the next couple of weeks, I will update this as additional results become available. Until then, remember that the predicted applicants for the 2026 Boston Marathon will be an underestimate – and the final prediction will be at least somewhat higher.
The current prediction is over 33,500 applicants. Once the final races are accounted for, this number will likely be between 34,000 and 35,000. That would be well above the number of applicants in 2024 (33,058) but fewer than in 2025 (36,393).