The Tableau dashboard below collects data from marathons, tracks the number of finishers who meet their Boston qualifying time, and projects an estimated cutoff time for the 2026 Boston Marathon.
It will be updated regularly throughout the year, through the registration period in September 2025. For more details on the data, the assumptions, and other factors, scroll down below the dashboard. Follow me on Threads for the latest updates.
Want to Stay Up to Date with the Latest Data?
I will continue to update this dashboard throughout the qualifying period as additional races take place.
About once a month, I will write up an analysis of the recent races and what effect they’ve had on the projected cutoff time. Use the form to subscribe to my weekly newsletter, and I’ll let you know when those are published.
More frequently, I’ll just be adding data to the dashboard. I will update the text in the dashboard with the date of the latest update, and I’ll also post something on Threads. Follow me there if you want to be the first to know when new data is added.
What Data Is the Dashboard Built On?
For this dashboard, I’ve collected the results of marathons with 200 or more finishers in the United States and Canada. I’ve also included the results of London and Berlin, because those are part of the Abbott World Marathon Majors.
Based on its date, each race is classified as being in the 2025 and/or 2026 qualifying period. There will be a few races in September 2025 that fall into both qualifying periods.
Although I have the full dataset for the 2025 qualifying period, I’ve excluded the races that haven’t yet been run in the 2026 qualifying period. As the results become available for the 2026 qualifying period, I’ll update the dataset and include results for both periods.
There are some races that have zero finishers in one of the qualifying periods. If that race appears in the dataset, that means the race did not take place in that qualifying period – either because it was discontinued, it is new, or it was temporarily cancelled due to extenuating circumstances.
The dataset includes a finish time, age group, and gender for each runner. For the 2025 qualifying period, the old (2020) Boston qualifying times were applied. For the 2026 qualifying period, the new Boston qualifying times were applied. Each runner is identified as a qualifier or a non-qualifier and their individual buffer time is calculated.
The runners age when they ran their marathon is used to determine the appropriate qualifying time. It’s possible they age up before Boston, so the actual number of qualifiers is higher than the calculated number. But, we can assume that this difference is similar from year to year and that the difference washes out in the aggregate.
The data comes from a variety of sources, including Athlinks, Marathon Guide, and individual race websites. I have not yet done so, but I will be uploading the dataset to Kaggle later in the season.
What Assumptions Are the Dashboard Based On?
This dashboard is based on several assumptions. These assumptions were tested last year when I attempted to predict the cutoff time for the 2025 Boston Marathon, and they held up fairly well. They are not perfect – but they provide a good starting point for an analysis.
The Boston Marathon cutoff time is based on three things: a) the total number of applicants, b) the number of accepted applicants, and c) the distribution of applicants’ finish times.
The biggest variable is the total number of qualified applicants. There are a lot of factors that go into this, and it’s impossible to fully account for all of them. But the core assumption that I’m starting from is that the number of applicants is directly related to the number of runners who meet their qualifying times.
Not every runner who qualifies for Boston will apply. But the percentage of runners who do apply should be fairly stable. Publicly available results allow us to track how many runners qualify. We can then calculate how that number changes. Finally, we can apply that change to the number of applicants from 2025 to project the number of applicants for 2026.
More complex modeling would take into account the runner’s gender and age, the race at which they qualified, and their individual buffer. While I will attempt to explore some of these things later in the qualifying period, that’s better left for a thorough analysis. For the purposes of this dashboard, it’s more effective to rely on a single core assumption.
The next variable is the number of accepted applicants. We can’t know exactly how many runners the Boston Athletic Association will accept. However, this has been between 22,000 and 24,000 in recent years. There is no indication the field size will increase significantly, and therefore it’s unlikely the number of accepted applicants will grow beyond 24,000.
Therefore, I’m using that as the second assumption.
The third variable is the distribution of the actual qualifying times, which determines how many applicants need to be rejected to balance the field. Based on the data released by the BAA when it announced the 2025 qualifying field, I’m assuming this to be approximately 1,800 runners per minute.
An Example of the Math Underlying the Dashboard
Here’s a simple example of how these assumptions work in practice.
Let’s say that the number of runners who qualified at this point in the year last year was 35,000 and the number who have qualified so far this year is 31,500.
35,000 – 31,500 = 3,500.
That 3,500 decrease in qualifiers is equivalent to 10%.
There were 36,393 qualified applicants last year. If this number declines by 10%, there would be 32,754 applicants this year.
To get to a field of 24,000 accepted runners, we would need to exclude 8,754 runners.
If there are 1,800 runners in each minute beneath the qualifying time, we divide 8,754 by 1,800 and get 4.86 minutes – or 4:52.
Updates and Analyses
Here are links to updates and analyses that I’ve posted throughout the qualifying period.
Future Enhancements and Further Analysis
This dashboard is a starting point, but I plan to make some improvements over the next few months to improve its ability to project and predict the Boston Marathon cutoff time.
First, you can filter out individual races. However, I plan to add some additional international races – especially big European races – and include broader filters for types of races.
Second, I’m currently basing my calculations off the age of a runner on the date of their marathon. It’s impossible to know for sure who does or doesn’t age up before Boston. But, I can identify some people who will age up and apply their new qualifying times. Once I work that out, I’ll update the calculations in the underlying dataset.
Third, the data currently includes all finishes. It does not attempt to identify runners who have completed more than one marathon. As we get later in the spring, it’s more likely that runners will have completed two marathons in the same qualifying period. So I will likely include a method to deduplicate the results and isolate a runner’s best time – to the extent that this is possible.
Fourth, the percentage of runners who actually apply to Boston varies according to the depth of their cutoff time. Runners that are 5 to 10 minutes beneath their qualifying time are more likely to apply than runners who are 20+ minutes below. I may modify the assumptions to account for this.
This is really, really neat.
It’s possible I missed this, but does this analysis account for the adjusted cutoff times that were announced for the 2026 marathon? For instance, a 5:33 buffer for 2026 would be the equivalent of a 10:33 buffer for 2025. That it’d have been 10:33 is totally possible given the growing popularity and improved times, but I just wanted to be sure.
Yes, the calculations underlying the dashboard apply the older qualifying times to runners in the 2025 qualifying period and the newer qualifying times to runners in the 2026 qualifying period.
Note also that these new times are not exactly equivalent to an extra 5 minutes of cutoff. The qualifying times changed for runners under 60 – but runners 60+ have the same qualifying times as last year. And they make up a small but significant portion of the field.
To give you a sense of the impact of these new times, the new times put the total number of runners who qualified to date at about 6.5% lower than last year. If the old qualifying times were applied across the board, that number would instead be up by about 10%.
This fantastic and very helpful. Thank you!
I’m having trouble following. Are you predicting the new cutoff, for example, for 40-44 will be 5:32 below the new bq of 3:05, making it 2:59.54? Thats FAST.
Also, your cutoff prediction changes if I select certain results of races, etc but that makes no sense. The cutoff time likely won’t be based on where you ran (although I wouldn’t disapprove). Thanks! I think it could be a cool tool if I can figure out exactly what it’s saying.
Short answer: Yes. The X:XX projected cut off is how far under the new qualifying time you would need to be.
The filtering is a little funky, and I’m going to refine how that works. But the idea is that you may want to exclude specific races from the dataset to see how that could influence the eventual cutoff time. That’s because some races are more or less likely to actually produce applicants – and removing them from the dataset shows you what kind of influence they have.
I find it hard to believe Chicago Marathon increased finishers by a lot, but LOST over 2000 BQs. That makes no sense. I think you should scrape the data again. I tried but my scraper was having issues. I may do it manually.
I’ll double check the results, but it’s likely due to the Age Group World Championships.
That was at Chicago in 2023, adding about 2,000 finishers who were very likely to BQ. That moved to Sydney in 2024, and they were replaced at Chicago by regular runners with a more typical distribution of times.
It was also warm this year at Chicago. This would have less of an impact than the Age Group Champs, but could also push down qualification rates a little.
So, I thought about that. But that’s a nearly 30% swing for an 8% increase in finishers. Still don’t make sense. And it really wasn’t that warm this year. I ran a sub-4 at Chicago this year, first time ever. Last year was warmer.
If you verify the data is good, then I would check Excel or whatever database you are using to make sure the times didn’t become strings. Also, the data you scrape doesn’t have BQ times inherent within it. You have to do a calculation somewhere. You can do it in Excel to manipulate in Tableau. Or you have to use a Tableau data blend or do a join. I prefer the manipulation in Excel as I’ve found the blend or join results in errors. But something doesn’t feel right.
In the analysis, I was going to do it from the “supply” end. I’m grabbing the results from all the majors. Then, putting the results into these buckets: 0-4:59, 5:00-7:59, and 8-plus. Why? Because, there has never been a buffer greater than 8 in the history of BAA buffers. Therefore, anything greater than 8 would be a “shoo-in.” My assumption if that number was near 24k, then what’s left is subject to the buffer calculation. Right now, you are showing 16k at the 10+ level. I’m guessing if you circle the time down to 8, you will approach 20k possible shoo-in applicants. If true, that’s a big number, WITHOUT the spring majors. But that’s my way at looking at the data.
thanks for the analysis
I have a 6:32 cushion on my 2026 qualifying time and felt pretty comfortable about it until reading your analysis
My only comment is that it really is hard to qualify for Boston and I find it hard to believe that the cutoff could still exceed 5 minutes even with the new times, but I guess the numbers so far seem to show that.
What about the Houston Chevron Marathon that happened this last weekend (1/19)? I don’t see it in the results and it is a pretty large marathon
The tracker should be updated later today with the results from Houston and several other marathons from this weekend. There’s a manual component to collecting and incorporating the results – but I typically try to update things within a week, and within a couple days for big races.
This is really cool (albeit a bit discouraging with my 4:09 buffer).
One question you mention in the write-up that there 36,393 qualified applicants for 2025 and the dashboard lists 38,494 under “qualifiers by qualifying period”. So it safe to assume that the delta is how many qualified runners chose not to apply (so roughly an application rate of 94.5% of people meeting the standard)?
Yes, the total number of qualifiers will be greater than the actual number of applicants – because not every qualifier actually applies.
Note that the 36k applicants is for the entire year. The number of qualifiers in the dashboard is based on races to date (and the equivalent time period last year). By the end of the qualifying period, there will likely be closer to 50-60k total qualifiers.
If I had a 4:09 buffer, I’d be nervous, too. It’s definitely possible that things moderate over the next few months, and a ~4:00 buffer is possible. But it’s not a likely outcome.
Would you consider adding the Mobile (AL) Marathon to your dataset? Selfishly I ran it and there were quite a few BQs so I am curious to see it added!
Hi Tyler. I usually limit the dataset to races with >200 finishers – because it would be a lot of effort to track the many smaller races, and they make up such a tiny percentage of the overall pool that it doesn’t impact the final outcome much.
But, since you asked … I’ll be adding the Mobile Marathon when I upload the updated dataset tomorrow.
And congrats!
Thanks for adding it into the mix! Appreciate your analysis and keep up the good work.