Six Star Finisher Follow Up: A Deeper Dive Into the Hall of Fame Stats

Feature image by Julian Mason on Flickr. CC BY 2.0.

A few weeks ago, I published an analysis of data that I collected from the Abbott World Marathon Majors Six Star Finisher Hall of Fame.

It provided some insight into who the Six Star Finishers were – and the path they took to collect all six of their Stars. It also generated a good discussion online, and from that discussion I garnered a few additional questions worth looking into.

Those questions include:

1. Which Star are Five Star Finishers most likely to be missing – and how many Six Star Finishers are we likely to see later this year?
2. Who holds the record for the longest Six Star journey from their first Star to their last?
3. Which races did runners complete the fastest? Which were the slowest?
4. Who are the fastest Six Star Finishers – among both elite runners and regular folks?
5. What can this data tell us about the difference between Boston, New York, and the flatter courses?

Which Star Is Missing?

One basic question worth asking is – of all the Five Star Finishers out there, which race are they still missing?

Among other things, this would give us a hint as to how many additional Six Star Finishers might be minted before the end of 2025.

This also gives us some insight into which races are more difficult to get into. In the previous analysis, Tokyo was by far the most common last race – which kind of implies it’s the hardest for people to get into. Although Boston and London came in a distant second, I was a little surprised Boston wasn’t higher up there, as many people struggle to get into Boston for years.

The Hall of Fame data also includes all Five Star finishers, and by identifying which races they’ve completed I can identify – through process of elimination – which race they still have to complete. The raw counts for each race are below.

When you look at the data this way, Boston is the most common race that runners still have to complete. Tokyo comes in second, but there are only about 4,000 runners who need to complete Tokyo compared to 6,000 who need to complete Boston.

Keep in mind that over 2,000 runners just completed Tokyo, so that likely reduced the Tokyo number. And it could have increased the Boston number, if some of those runners had four stars prior to Tokyo.

In the last three years, more than 2,000 runners finished their Six Star journey at Tokyo each year. So if you rewound the clock a couple years, the Tokyo bar would likely be a lot higher.

Boston, on the other hand, produces far fewer Six Star finishers per year. So in the long run, it may be that Boston is still the “hardest” race to get into – and there was a large number of runners who needed Tokyo and managed to complete it since 2023.

The flip side of this is that London – although a distant third – is still far more common than Berlin, Chicago, and New York. This further reinforces the general sentiment that Berlin, Chicago, and New York are the easiest stars to earn – and Boston, London, and Tokyo are more difficult for many people to get into.

How Many New Six Star Finishers Can We Expected This Year?

Based on previous years, the largest number of runners will complete their Six Star journey at Tokyo. But there will still be quite a few additional Six Star finishers throughout the year.

The Boston Marathon releases it’s entry list prior to the race, so that actually gives us some good insight into how many new finishers we can expect on Monday.

I took the list of Five Star Finishers who were missing the Boston Marathon, and I compared them to the entry list published by the Boston Athletic Association.

First, I ran an initial match of names of the Five Star Finishers against the names on the Boston entry list. This yielded 1,484 possible matches.

There were a handful of duplicated names, and eliminating them left 1,447 unique runners on the Boston entry list who appeared in the Five Star finisher list.

When I compared the country – according to the Five Star list – the country on the Boston entry list, there were 253 mismatches. However, when I compared against either the country of residence or country of citizenship on the Boston entry list, there were only 121 mismatches.

So of the 1,484 possible matches, you can remove some duplicates, some mismatches, and some other possible false positives … and you’re left with ~1,200 to 1,300 runners at Boston who are likely Five Star Finishers waiting to earn their final star.

After you take into account the fact that a small but significant portion of runners will no show for various reasons and a smaller number will not finish the race, I’d expect there to be 1,000 to 1,100 new Six Star finishers after Boston.

The list of runners who need London is much smaller (~1,800). I’d bet London yields another 3-500 finishers, and the remaining three races yield far fewer in the fall.

All told, the total number of Six Star Finishers will likely be between 22,000 and 23,000 by the end of the year.

Who Holds the Record for the Longest Journey?

After I published the last analysis, I mentioned that there were a few runners who took 40+ years to complete their journey. Somebody reached out to me to me to confirm some details, because he ran New York City back in 1980 – and he’s wondering if he’ll set a new record once he completes his journey in the next few years.

There are some quirks to the dataset – notably that if a runner completed the same race more than once, only their fastest time is displayed in the Hall of Fame. This introduces some uncertainty over the exact dates when a runner started and finished their journey (and the dates of each race in between).

At a large scale, that likely washes out. But on an individual level, it’s worth double checking things – and with a small number of finishers that’s possible to do manually.

The runner with the current record for the longest journey is Gregor Birk.

He ran Boston way back in 1978, and then he completed Chicago two years later. After a significant break, he ran New York City in 2007, Berlin in 2013, London in 2014, and Tokyo – finally – in 2024. That’s 46 years from start to finish.

I checked the results of earlier Boston Marathons, and he did not run the race prior to 1978. He would have been ~23 at the time. And I checked the 2025 Tokyo Marathon, and he didn’t run that either. Technically, it’s possible he ran the inaugural Chicago Marathon in 1977, but those results aren’t available for confirmation.

The next longest journey is Joan Samuelson – yes, that Joan Benoit Samuelson.

She first ran Boston in 1979 – when she won the race in 2:35.

She ran Boston two more times (1981 and 1983) before earning her second star at Chicago in 1985 (which she also won). That was followed by New York in 1988 (“only” a third place finish). Of course, at the time there was no Abbott World Marathon Majors.

Long after her professional running career was over, she ran Berlin in 2019 and London in 2022. She earned her final star at Tokyo in 2024 – which made headlines at the time.

From 1979 to 2024, her Six Star journey took 45 years.

After that, two runners are tired for third with 43 years.

Susan Hrabchak ran the 1982 New York City Marathon, although it appears she had a different name then. There’s a Susan Kellner of NJ in the results with the same time as her Hall of Fame record. She was 21 at the time. She completed her journey at this year’s Tokyo Marathon (2025), 43 years later.

Alfred Scaletta started out at Chicago in 1980. He would have been about 19 at the time. He ran Boston the next year (1981). It wasn’t until 2010 that he finally ran NYC, and he did the last three races post-COVID. Tokyo was his final stop – in 2023.

Which Races Are the Fastest? The Slowest?

I think the accepted wisdom is that Berlin and Chicago are the fastest of the Majors – followed by Tokyo and London. When you compare their courses, all four of them are pretty flat. NYC and Boston, by comparison, are hilly, and the they are considered much tougher.

But this dataset offers us some real world data of results from individual runners matched to six different races. And that makes for some interesting analysis about which races they completed the fastest and which were the slowest.

In the visual below, the group on the left shows the number of runners who ran their fastest time at each of the six races. And the group of bars on the left represents their slowest race.

Berlin and Chicago are, without a doubt, the most common fast races. That matches with the general consensus that those are the two fastest races.

Boston and New York City are the least likely to be the fastest races – which makes sense given the fact that they have the toughest courses. I was a little surprised to see Tokyo and London so close to them, though.

On the flip side, Tokyo is actually the most common race to be a runner’s slowest.

If it’s a fast course, how can that be?

The likeliest explanation is that many runners came back and completed Tokyo years after their prime. It’s by far the most common last race of a person’s journey. And if a runner started off in their 20’s and 30’s – but didn’t run Tokyo until their 50’s or 60’s – age will take its toll.

When I looked at the average time between the slowest race and a runner’s first race, Tokyo had, by far, the highest average. Which confirms that people are typically running it later in life.

After that, Boston and New York are the most common slow races, with Berlin and Chicago less likely.

Who Are the Fastest Six Star Finishers?

When I shared the last analysis, somebody asked to be able to see runner’s ranked by their finish times. That’s easy enough to do, but I thought it would be useful to break it out into the best of the best – the fastest elite runners – and a ranking of all runners.

When I started this analysis, I was surprised when I found some results with times less than two hours. It occurred to me at that point that wheelchair racers are also included in the Six Star Hall of Fame, and for the purposes of this comparison I filtered them out.

This table includes the ten fastest women – ranked by their best time of all six races – who have earned the Six Star Medal.

There’s an interesting span across time and generations here.

One of the OG’s, Joan Samuelson is #4 on this list thanks to her 1985 performance at Chicago. Her PR still holds up as a respectable time among top women, even if the world record has progressed since then.

Elizabeth Nuttall (McColgan) represents the 1990’s here, with her 1997 performance at London. Then there are a handful of runners from the early 2000’s, followed by a handful of more recent runners.

Other than Joan, there some other heavy hitters on this list – notably Deena Kastor, Edna Kiplagat, and Sara Hall.

On the men’s side, the biggest name to appear is at the top – Emmanuel Mutai. He has an impressive resume, including a win at the 2011 London Marathon and a lengthy list of silver medals at other Majors and the World Championships.

His best time as at the 2014 Berlin Marathon, when he finished second in 2:03:13. Dennis Kimetto beat him out by 16 seconds – setting a new world record in the process.

He’s the only top tier elite man on this list, but the remainder of the top athletes in their own countries. Yuki Kawauchi made the podium at the 2011 Tokyo Marathon, when he ran 2:08:37.

Unlike the women, there’s no one left on this list from the 80’s. The oldest result is Carey Nelson of Canada, who ran his best time at Chicago in 1994.

There are also fewer recent additions, with Lucas Baez being the only runner with a result in the last ten years.

How Does Everyone Else Rank Up?

It’s always fun to look at the elites, but what about the rest of us?

I took the list, and I calculated the median finish time for each runner. I then assigned a ranking based on that median finish time – split between men and women. I also included that runner’s best finish for comparison.

The list is long, and you can page through it. But if you’re just interested in finding yourself or a friend, there’s a search box where can enter a name.

Here’s the same table for women.

How Much Slower Are Boston and New York?

The final question that I really wanted to explore was just how much slower Boston and New York are than the other races. Phil Maffetone wrote a research study on this back in 2017, but it focused on elite finishers. I’m more interested in the masses.

On the one hand, these results offer a great way to make comparisons – because you can easily compare the results from the same runner.

On the other hand, this is fraught with some potential complications. For example, Tokyo isn’t the slowest race because the course is tough – it’s slower because runners complete it when they’re older. If too much time elapses between results, it’s less likely that the course made the difference – and more likely than age did.

To some extent, you can control for that by eliminating results where there’s a big gap in time. But another potential confounding factor is that a runner’s first race may not be their fastest. You get better as you go. Training in subsequent years can contribute to improvement.

That’s harder to control for without knowing something about each individual runner’s training history. But we’re just going to assume it kind of washes out in the large scale of the data.

I started with the full set of results, and I calculated the span of time from a runner’s first star to each result. If that result was more than 10 years after the first race, I excluded it from the dataset.

Then, I calculated a median finish time for each runner at Chicago, Berlin, and London. By using the median, we eliminate the impact of outliers – like a particularly bad race due to heat or injury.

That then became the baseline, and I could compare that to their finish at Boston and New York City.

I started by calculating the mean difference. For Boston, the mean difference in times was 5:23 – meaning that runners were, on average, 5:23 slower at Boston than on the flat courses. However, the standard deviation of this average was 23:51, which suggests there’s quite a bit of individual variation.

To get a better sense of that, I created a histogram to show the overall distribution. You can see that above. The number on the x-axis (i.e. -30) is the lower limit of each bucket. A negative number means they were faster at Boston, and a positive number means they were slower.

For example, -7 represents all runners who were 6:01 to 7:00 faster at Boston, while 10 represents the number of runners who were 10:00 to 10:59 minutes slower at Boston.

You can see that the highest bar (the mode) is 1 – so the plurality of runners were just slightly slower at Boston (1:00 to 1:59). But the median – the midpoint of the entire distribution – was 3:44.

You can see that the distribution spreads out pretty far, though. So while many runners are a few minutes slower at Boston, there are some runners who are a) much faster and b) much slower.

The other two reference lines here – -6:45 and 16:42 – mark the bounds of the lowest 25% and the highest 25%. That’s the interquartile range – or the range of the middle 50% of results.

To summarize, although there’s a broad distribution of individual differences and some runners are faster at Boston, it’s typical, on average, for runners to be a few minutes slower.

Here’s the same graph for New York City.

Starting with the raw numbers, the mean difference here was 5:26 – about the same as Boston. But the standard deviation was 23:37, a little bit lower. So the distribution might be a tiny bit tighter, but there’s not a whole lot of difference there.

When you look at the histogram, the mode is now 0 (0:00 to 0:59) and the median is 4:18. On average, the results are a tiny bit slower at New York City than Boston, but there is a larger group of people finish at similar times to their flat races.

The interquartile range is about the same – from -6 to +16.

I’ve been interested in this question of how much slower these two courses are, and I’ll probably come back to this and try to tighten things up a bit with a more narrow timeframe. But for now, I think this offers some pretty good evidence that Boston and New York City are both slower courses – compared to the rest of the majors.

That might seem like a “Duh,” conclusion. But when you start talking about downhill courses, there’s always somebody who throws out the fact that Boston is a downhill course.

It’s true. But it doesn’t make it easy. That net decline does make Boston a little bit easier than New York – but the total elevation gain in the back half still makes it much tougher than a flat course like Chicago or Berlin.

What Are You Thoughts?

What do you think about the data presented here. Anything surprise you?

If you haven’t, you may want to go back and check out the original analysis about Six Star Finishers here. It answers some other basic questions and provides a good picture of who the Six Star Finishers are.