This week, the Boston Athletic Association officially announced the cutoff time for the 2025 Boston Marathon: 6:51.
As they have done in the past, they included with this announcement a collection of data and statistics that tell some of the story of the applicant pool and the field of qualifiers. Absent a full release of the applicants and their qualifying times, these clues are all that we have to try and understand what factors are at play – and what future cutoff times could be like.
This year, I analyzed a bunch of data. I initially predicted a cutoff time of 7:03, and I ultimately settled on a range of possible times: 6:30 to 7:30.
So, on the one hand I was right. But it was sort of like the math problem you get right … despite doing it wrong.
Although much of the modeling was sound and I was in the right ballpark, there were a couple factors that differed from the past and threw a wrench in the process of predicting a cutoff time.
What can we learn from the data the BAA released yesterday, and what does it tell us about future cutoff times? I haven’t had time for a deep dive, but here are a few initial thoughts.
The Number of Accepted Applicants Increased
This year, the BAA accepted 24,069 qualified applicants based on qualifying times. This is back to an (almost) all time high.
The only year with more accepted applicants was 2020 – with 24,127. There were also just over 24,000 in 2016.
Since 2015, the number of accepted applicants has been around 23,000 for five races. Last year was the lowest number of accepted applicants since before 2014, at 22,019.
The inclusion of additional applicants drove the cutoff time down by a modest amount. More on that later. But it also sets up a question for next year: Will the field size remain the same?
The overall field is again 30,000 runners, meaning the number of other bibs (charity, sponsors, international tours, influencers) is at its lowest point in more than a decade. If BAA faces push back from this side of the field, they could reduce the number of accepted applicants back to ~23,000 – a more typical amount.
In any case, it’s unlikely that that number goes higher. Increasing the number of accepted applicants was one relief valve available to BAA to help reduce the cutoff time, and that won’t be an option next year.
How Much Did the Increase in Accepted Applicants Matter?
That 2,000 runner increase likely reduced the cutoff time by about a minute.
The BAA released two relevant data points – the number of applicants with times between 6:51 and 10:00 (5,873) and the number of applicants with times between 0:00 and 6:50 (12,324).
If you divide the number of applicants by the size of those buckets, you get approximately 1,864 runners per minute for 6:51 to 10:00 and 1,799 runners per minute from 0:00 to 6:50. That number drops significantly once you move on to the 10:00 to 20:00 bucket. But it’s likely pretty consistent – just under 2,000 runners per minute – from 5:00 to 10:00.
Therefore, by adding 2,000 seats they likely reduced the cutoff time by about 1:00. It’s impossible to know the exact number of seconds without the exact distribution of BQ buffers, but it would have been within a few seconds of 7:55.
This is why my prediction was both right and wrong. I settled on the right timeframe – 6:30 to 7:30, with an actual cutoff closer to 7:00. But that was predicated on an accepted field size of 22,000. If I had assumed an increase of 2,000 runners, I’d have projected a cutoff time closer to 6:00 to 6:30.
So what changed?
Qualifiers With 10 Minute Buffers Were Much More Likely to Apply
The biggest difference – and the thing that screwed up my projection – was that qualifiers with just under a 10 minute buffer applied at a much higher rate than they did previously.
In the 2024 qualifying period, my sample included 14,122 qualifiers that beat their BQ by at least 20 minutes. There were 6,182 applicants with a 20+ minute buffer – for a conversion rate of 43.78%. Note: This is the ratio of my sample to actual applicants. Less than 40% of all qualifiers applied, because there are additional qualifiers not in my sample.
This was consistent with the 2025 qualifying period. BAA announced that there were 6,971 applicants that beat their BQ by at least 20 minutes. That’s an increase over last year, but the number of qualifiers also increased significantly to 16,197. The conversion rate was 43.04% – the same ballpark as last year.
In other words, the relationship between the number of qualifiers and the number of applicants was the same for applicants with a buffer of 20 minutes or more.
But if you look at runners with a 10 to 20 minute buffer, there’s a significant difference. In 2024, the number of applicants (8,858) was 62.59% of the number of qualifiers in the sample (14,152).
This year, the number of applicants in this bucket increased dramatically – to 11,199. The number of qualifiers (15,516) increased at a much smaller rate, and the conversion rate was 72.18%.
In other words, there were slightly more qualifiers in this bucket – but many more of them actually chose to apply. This shifted the distribution of applicants qualifying times to the left compared to last year.
Meanwhile, in the 0 to 10 minute range there were two divergent trends.
For runners that beat the cutoff, the conversion rate was 78.59% last year. This year, it was 86.83%.
Among runners who were slower than the cutoff, it went the other way. Last year, the conversion rate was 79.30% – so roughly the same as runners closer to 10 minutes. This year, it was 67.55% – significantly lower.
This visual above summarizes these findings. The blue bar shows the number of applicants in each bucket – released by BAA – and the purple bar shows the number of qualifiers in my sample. If you hover over the bar, you’ll see the conversion rates.
But you can visually see that the last two buckets are very different. For the third bucket (between the cutoff and ten minutes), the two bars move closer together. More of the qualifiers apply, and there’s less difference between the two numbers. For the fourth bucket (between the cutoff and regular BQ), the two bars get further apart.
These changing conversion rates make it difficult to predict the number of applicants in each bucket – and therefore the cutoff time.
What Explains This Shift?
Whose to say for sure … but let me offer two hypotheses.
First, runners can be divided into two groups – those who want to run Boston and those who don’t really care.
The data from 2024 suggests that as you move further below BQ times, the balance shifts towards fewer runners who care about Boston. Those with buffers of 0-10 minutes were the most likely to apply, then those with 10-20 minute buffers, followed by those with 20+ minute buffers.
A reasonable explanation for this is that for the runners who have Boston most front of mind – the first timers and the ones who have been striving for a BQ for a long time – their goal is just to BQ. Maybe by a few minutes if possible. But their target was likely somewhere around BQ to BQ-5.
But after last year, the general consensus was that you needed to run much faster to actually get into Boston. So for this group of people – who really wanted to run Boston – the target was now somewhere between BQ-5 to BQ-10.
This likely shifted some runners towards actually finishing faster.
This is born out by the data. When I look at the increase in the number of qualifiers in each bucket (0-6:50, 6:51-9:59, 10:00-19:59, 20:00 or more), the group that finished 6:51 to 9:59 ahead of their BQ saw the highest increase (15.82%). By contrast, the number of runners finishing 0:00 to 6:50 ahead of their BQ only increased by 4.85%.
Second, it’s likely that some number of runners with small buffers – say 1, 2, or 3 minutes – simply chose not to apply. They qualified, but they assumed their chances of getting in were minimal and they simply tuned out. That would also, partially, explain the drop in the conversion rate for runners qualifying but not making the cutoff. And it likely kept the overall number of applicants slightly lower than it otherwise would have been.
We know for a fact there were 12,324 runners who qualified, applied, and were rejected. But there are almost definitely more runners out there who qualified, wanted to apply, but assumed they would be rejected. They are within a couple minutes of the new qualifying times, and I’m sure some of them are making plans to notch a new, better qualifying time for Boston 2026.
What Were the Largest Qualifying Races This Year?
This year, the BAA released not only the name of the top five qualifying marathons – but they also released the number of entrants from each of those races.
The five races remained the same, but the order changed.
The Boston Marathon – which typically is the largest source of qualifiers for the next year’s race – dropped to #2. 1,822 of the runners at Boston 2024 qualified for and earned in a spot in 2025.
The Chicago Marathon moved from #2 to #1. It yielded more than 10% of this year’s field of accepted applicants – 2,584.
London and Berlin come next with 1,101 and 809 entrants respectively.
The field size of these three races is similar. London is the largest (54k), followed by Chicago (48k), followed by Berlin (43k). But they’re all in the same ballpark.
And they all yielded a significant number of qualifiers. I recalculated the number of qualifiers to check specifically for how many beat the cut-off – and there were 6,854 in Chicago, 6,743 in London, and 5,412 in Berlin.
But the rate of those qualifiers actually applying to run Boston is much higher for Chicago (37.7%) than for London (16.3%) or Berlin (14.9%). This reinforces the notion that qualifiers in certain races are more likely to apply – and therefore those races have a greater impact on what the potential applicant pool (and buffer) will be.
For its part, the California International Marathon came in fifth again. This year it yielded 749 entrants – from a field of just over 9,000 runners. And a whopping 2,181 of them beat the cutoff.
What Does This All Mean for Next Year?
Obviously, it’s too early to make any kind of specific prediction for what next year’s cutoff time will be.
But we’ve learned a few things that give us a general idea – and the results of the big marathons over the next few months will fill in the details.
Factors Suggesting a Cutoff Time for 2026
We know that there were 24,000 applicants with buffers of 6:51 or better. We can also assume that there were between 3,500 and 4,000 applicants with times buffers between 5:00 and 6:51. So despite a five minute shift in the qualifying times (for most runners), there are definitely enough runners out there to fill up the field – unless something drastically reduces the number of runners who qualify.
We also know that some runners are not going to see their qualifying times change. So there are some masters runners out there who are already capable of qualifying – but they were either eliminated by the cutoff or chose not to apply. So that will mildly dampen the effect of the new qualifying times.
We also know that there was a huge increase in the number of runners beating their old BQ by 7 to 10 minutes. These runners will still be eligible under the new qualifying times. It’s not the case that many of the newer qualifiers were in the 0-5 range, and they’ll be eliminated by the new qualifying times. To the contrary, there are likely many runners who were on the cusp – probably targeting a time of BQ-5 – who can improve on their training and score BQ-5 with the new qualifying times.
And of course there’s the latent potential of Boston. The weather in 2024 reduced the number of applicants coming out of Boston so much that Chicago became the number one qualifying race this year. It’s hard to see a scenario in which Boston gets worse. So if there’s mild or good weather, that would yield additional qualifiers.
Finally, we know that the number of accepted qualifiers is at an all time high (24,000). This could remain the same next year. But it’s just as likely that it reverts to the typical level of 23,000. And it’s incredibly unlikely that it increases. So the net impact of field size will either be neutral – or it will slightly increase the cutoff time.
Factors Mitigating the Size of a Cutoff Time in 2026
Nonetheless, I think it’s unlikely that there would be a huge cutoff next year. I’ve seen some people suggest that the cutoff time will still be 5 or 10 minutes. I just don’t see how that could happen.
To push things past (new) BQ-5 – roughly equivalent to BQ-10 today – would require another ~6,000 or so applicants. There were roughly 18,000 applicants with that range of times this year. That would be require a huge (33%) increase. Despite some large increases this year, no portion of the field came close to that.
There’s also the question of where these new applicants would come from. Some runners will get faster, yes. But this year’s large increase in the number of qualifiers was driven in large part by an overall increase in the field size across all marathons. That, in turn, is explained in large part by a rebound from the depressed field sizes of 2022-2023 in the years following the COVID shutdowns.
When we look back at 2024, I suspect that overall participation in marathons will have returned to its peak of around 2014-2015. To sustain huge increases in the number of applicants, the sport would have to continue to grow well beyond that previous peak. Possible? Sure. But after the surge in the last two years, we’ll likely see smaller increases moving forward.
Frankly, it is incredibly unlikely that the applicant pool will even come close to the record levels of this year or last. And without 33,000 to 36,000 applicants, there’s no way you get to a cutoff time of 5 minutes below BQ.
It’s obviously far too early to make a real prediction about what will happen, but I’d say it’s quite likely that the total applicant pool will be somewhere in the neighborhood of 27,000 to 30,000.
When January comes around, and some of the fall races are in the books, we’ll know a little bit more. And after April, once Boston and London take place, it will be much more clear.
But I’m going to go on record now saying that BQ-5 is almost definitely a safe target for the 2026 Boston Marathon. I’m running Chicago in a couple of weeks, and I’m going to put my money where my mouth is. After two years of coming up just short, I’m targeting 3:00 – which is BQ-5 for my age group (Men 40-44).
Wish me luck.