There are races, and then there’s Boston. On April 21, 2025, I had the privilege of running the 129th Boston Marathon – my fifth Abbott World Marathon Major – and perhaps the most unforgettable one yet. This race took place on Patriots’ Day, a Massachusetts state holiday that commemorates the start of the American Revolution. In fact, 2025 marked the 250th anniversary of the battles of Lexington and Concord that ignited the Revolutionary War, giving Marathon Monday an extra layer of significance. As a runner and a history buff, I could feel the importance of the day: the entire city seemed to celebrate not just a race, but a rich heritage of perseverance and freedom. Under clear April skies, with colonial-era reenactments at dawn and crowds packing the streets by mid-morning, I felt Boston’s spirit all around me as I toed the starting line in Hopkinton.

Boston is the world’s oldest annual marathon, first run in 1897, and it has always been tied to Patriots’ Day. Originally held every April 19 (the actual Patriots’ Day date), the marathon now happens on the holiday’s observance – the third Monday in April – allowing the whole region to turn out in celebration. Marathon Monday is unlike any other race day I’ve experienced. The entire city has the day off, and even the Boston Red Sox play a morning game at Fenway Park as part of the tradition. This year, with the 250th Patriots’ Day anniversary, there was a tangible sense of pride along the course. I remember running through Ashland and Framingham in the early miles, seeing families with American flags and signs referencing “1775.” It felt like we runners were carrying forward a legacy. The symbolism of starting in rural towns (just as the Revolutionary battles began on village greens) and finishing in the heart of Boston was not lost on me. I was quite literally running through history on this day.

Having run other races, I can confidently say Boston is a marathon like no other. For one, you can’t just sign up for Boston – you generally earn your way in. The Boston Marathon famously requires runners to meet strict qualifying times (based on age and gender) or to run for charity. These qualifying standards have been a hallmark since 1970, meaning most participants have already run a fast marathon just to be here, creating a field filled with dedicated, passionate runners. Lining up in Hopkinton, I felt an electric camaraderie knowing everyone around me had put in serious work to qualify. This wasn’t just another race; it was a reward for years of effort.

Boston’s course and traditions also set it apart from the other Abbott World Marathon Majors. Unlike the big loops through major city centers (think London or Berlin), Boston’s route is a point-to-point trek from a small town to the city. We start on Main Street in Hopkinton with the famous words “It all starts here” painted on the road, and we end 26.2 miles later on Boylston Street in downtown Boston. The first few miles are downhill out of Hopkinton, and it’s easy to get carried away by the adrenaline and fresh legs. But Boston teaches patience – those early descents pound your quads and can come back to haunt you later. No wonder they say Boston is a course that finds your weaknesses.

Then there are the legendary landmarks and traditions of the course. Just past mile 12, I could hear a roar in the distance – the Wellesley College “Scream Tunnel.” For about a quarter mile, the women of Wellesley line the course with deafening cheers and witty signs (yes, even the famous “Kiss me!” posters were out). Despite being only halfway through the race, I was grinning ear to ear from that energy boost. Further on, the route weaves through the suburban hills of Newton. Boston’s layout is famously challenging, thanks to the series of uphill climbs in Newton that culminate in the notorious Heartbreak Hill. Many marathons have hills, but Boston’s Heartbreak is iconic – not because of its height, but because of when it hits. As one writer put it, by the time you reach Heartbreak around mile 20, even a small incline can feel like a mountain. This hill has literally broken runners’ hearts and ambitions over the years, which is how it earned its name back in 1936.

Of course, Boston’s uniqueness isn’t only about difficulty; it’s about tradition and community. This race started in 1897 with just 15 runners, inspired by the Olympic marathon, and it has been run every year since (with the sole exception of 2020’s pandemic disruption). It’s also the only major marathon that is a regional holiday; the race and the community are entwined. In Brookline, I passed a priest running in honor of Pope Francis, blessing spectators as he went. One runner wore 1775 colonial garb – complete with gold-buckled shoes and a tricorn hat. A teenager handed out tissues at Wellesley because “his mom always forgets them.” Everywhere I looked, someone was running for something or someone. Marathon Monday in Massachusetts is basically a city-wide block party meets elite sporting event, and being a part of it was something truly special.

This year’s Boston Marathon had its own share of exciting storylines. According to the official program and pre-race buzz, the 129th running was poised to be one for the books. For starters, the field was larger than ever: the Boston Athletic Association (B.A.A.) anticipated about 30,000 runners from 118 countries and all 50 U.S. states toeing the line. As I looked around Athlete’s Village in Hopkinton before the start, I could see the diversity and excitement – runners of all ages, speaking dozens of languages, wearing club jerseys from around the world. Even days earlier, during my layover at the airport in Istanbul, it struck me how people come from every corner of the globe to run Boston on Patriots’ Day. The marathon truly lived up to its nickname as the world’s most prestigious 26.2.

One of the most celebrated milestones of 2025 was the 50th anniversary of the wheelchair division. In 1975, Boston became the first major marathon to officially include a wheelchair race after Bob Hall famously completed the course in a wheelchair, proving it could be done and convincing organizers to create an official category. Fast forward to today, and the wheelchair division isn’t just included – it’s one of the marquee events. This year’s wheelchair field was touted as perhaps the strongest in history, featuring defending champions and course record holders. I heard that Marcel Hug of Switzerland, the “Silver Bullet” who won his seventh Boston title last year in a record-smashing 1:15:33, was back to defend yet again. On the women’s side, Eden Rainbow-Cooper of Great Britain was also returning after making history in 2024 as the first British woman to win Boston’s wheelchair race. Knowing these elite athletes were out on the course ahead of us – and that the B.A.A. had increased prize awards in honor of the 50th anniversary – gave me and many other runners a huge sense of pride. Before we even started running, we all applauded during the announcements for the wheelchair start, acknowledging how much this part of the event means to Boston.

The professional runners lining up in 2025 were no less impressive. Boston always attracts a stellar elite field, and this year we had several notable returning champions in the mix. The defending women’s champion, Hellen Obiri of Kenya, was back aiming to win Boston for the third year in a row. Only a handful of women in history have ever pulled off a “three-peat” at Boston, so all eyes were on Obiri. (Fun fact from the program: the only four women to win three straight Boston Marathons are Bobbi Gibb, Sara Mae Berman, Uta Pippig, and Fatuma Roba – so Obiri would join truly legendary company if she succeeded.) Also in the women’s field was her compatriot Sharon Lokedi, who nearly caught Obiri last year, as well as top Ethiopians like Amane Beriso and Yalemzerf Yehualaw. In the men’s race, we had Sisay Lemma of Ethiopia, the defending 2024 champion, toeing the line again. Lemma had run a blistering first half last year and held on to win, and with a 2:01 personal best on his résumé, he’s one of the fastest men ever. Challenging him was a stacked group including Evans Chebet of Kenya (who won Boston in 2022 and 2023) and John Korir of Kenya, the reigning Chicago Marathon champion who was hungry for a Boston win. As a fan of the sport, I was geeking out knowing all these champions were running on the same course just ahead of me. The buzz in the village before the start was that we might see something special with so many accomplished athletes in the field. It’s not often you have both the defending champs and multiple past champs all coming back, but Boston in this milestone year drew them in.

This year’s marathon didn’t just celebrate history – it made it. The elite performances lived up to the hype, with several record-breaking and memorable moments on the day:

• John Korir overcame an early fall (in which he even lost his bib) and powered through to win the men’s elite race in 2:04:45, the second-fastest time in Boston history.

• Sharon Lokedi delivered a masterclass in pacing and perseverance to defeat two-time champion Hellen Obiri and set a new women’s course record of 2:17:22.

• In the wheelchair division, Marcel Hug claimed his eighth Boston title, tying him for second-most all-time. And Susannah Scaroni obliterated her own course record by six minutes to win her second women’s wheelchair title.

• It was also a day of heartfelt farewells. Boston fan-favorite Des Linden, the 2018 champion, ran her final professional marathon – a fitting “love letter to Boston” that reminded us why this race touches so many lives.

The personal journey of those 26.2 miles was intense, emotional, and unforgettable. I had run four of the other Abbott World Marathon Majors before (New York, Chicago, Berlin, London), but lining up in Hopkinton for Boston gave me butterflies like never before. When the starting gun went off for my wave, I felt a surge of adrenaline – I was finally running the Boston Marathon! Like many runners in Boston this year, I started the day feeling amazing. The first 30 to 32 kilometers were almost dreamlike. Controlled, pain-free, with my guts being nice to me, The early miles out of Hopkinton, through Ashland and Framingham, felt nearly effortless. I settled into a steady, confident pace around 4:50–4:52 per kilometer, ticking off each marker without strain, buoyed by the downhill profile and the roaring energy from the crowds. However, starting in the last Wave 4 came with a challenge: I spent those early kilometers zigzagging past slower runners. The road was so packed that I often had to move onto the shoulder to overtake people. This made the first 5–10K slower than expected just by the relative effort and mentally tiring as I weaved through the throng, but physically I still felt strong and smooth. Everything clicked – my legs were fresh, my breathing calm, my heart rate under control. It was the kind of flow state you train for but rarely experience so purely on race day.

But Boston isn’t just about the start – it’s about how you handle the hills. As I entered Newton and faced the famous series of climbs (culminating in the iconic Heartbreak Hill), the race began to ask harder questions. I felt the accumulation of effort in my hips and calfs, the slight drift upward in my pace, and the mental grind setting in. Despite digging deep, my average pace began to slip – not catastrophically, but steadily – in those crucial late miles. Heartbreak Hill was still controlled within my limits and the spectators were unbelievable there, practically carrying us up with their energy. Eventually, I crested Heartbreak, worried about what’s ahead but determined to maintain the original pace which would have given me my personal best. From there, it was literally (and figuratively) downhill toward Boston.

In the final stretch, I turned right on Hereford Street and then left onto Boylston – the famous finish approach – and the finish line came into view. I crossed the line in 3:33.14, about six minutes shy of my personal best. After a winter of focused preparation, and despite battling an infection just days before the race, I had dared to hope that a PR was within reach – and for two-thirds of the marathon, it truly was. That’s what stings a little: it was so close. But moments after finishing, as I stood on Boylston Street with the medal around my neck, pride outweighed disappointment. Boston demands everything, and I gave everything I had on the day. That’s the marathon – and that’s why I’m already looking ahead. Ten months from now: Tokyo. A fresh dream. A final star.

After catching my breath, I was eager to understand exactly why my race unfolded the way it did. The data from my Garmin watch covers each segment of the 2025 Boston Marathon, offering insight into how my pace varied with heart rate, cadence, stride, form, and terrain. I’ve crunched the lap-by-lap stats (every kilometer split) to uncover what helped or hurt my speed over the 26.2 miles. Let’s dive into the numbers to see where things went well and where I hit the proverbial wall.

Looking at the scatter plot of my average pace versus heart rate for each kilometer, you might expect a higher heart rate to coincide with faster running – but the red trend line is nearly flat. Surprisingly, my data shows almost no linear correlation between heart rate and pace. Early in the race, my heart rate climbed from around 123 bpm up to the 140s as I settled into race pace, yet my pace actually slowed slightly in those opening kilometers, perhaps due to the need to maneuver between slower runners. Later on, even as I slowed down significantly to 5:30–5:50 min/km in the final stages, my heartrate stayed elevated around 145–150 bpm. In other words, by mile 20+ my legs were tiring and forcing a slower pace even though my heart was still working hard. This decoupling of heart rate and speed is common in marathons – as fatigue sets in, heart rate remains higher while pace drops, so the typical HR-to-pace link breaks down.

The chart above, which plots my pace (blue line) and heart rate (red line) against distance, visualizes this disconnect. For example, in the first 5 km my heart rate rose sharply (red line trending upward) while my pace slightly eased off (blue line rising from ~4:45 to ~4:55 per km). And in the last 5 km, you can see my pace dramatically slowing (blue line peaking above 5:40) even though my heart rate stayed in the mid-140s, perhaps bit on the lower side given the end in sight, indicating there might have been a reserve to be used for a proper finish. My heart was working well, but my legs simply couldn’t keep up the speed at that point.

Insight: Don’t be alarmed if late-race heart rate doesn’t drop along with pace – it’s a sign of fatigue and cardiovascular strain outpacing muscular ability. In my case, heart rate was not a reliable indicator of pace changes over the course of the marathon (correlation essentially zero). Other factors like hills and overall fatigue played a much bigger role in my pace than heart rate did.

One big question I had was whether my cadence (steps per minute) or my stride length had a greater impact on my pace. Essentially, you can run faster either by taking more steps per minute or by covering more ground with each step. My Garmin data tracked both, and the results are clear.

The relationship between cadence and pace turned out to be a strong one. I maintained a fairly consistent cadence in the high 160s to low 170s spm (steps per minute) for most of the race. There’s a visible trend that faster paces came with a slightly higher cadence, and slower kilometers saw cadence dip into the mid-160s. The scatter plot above shows this inverse relationship: when my cadence dropped, my pace (time per km) got slower. That said, the changes in cadence were not huge – my step rate only varied by about 10 spm between the fastest and slowest segments. This gave me a clue that something else was contributing more to the pace fluctuations.

It turns out stride length was the bigger factor. The plot of pace vs. stride length shows a tight correlation: almost all my slower splits correspond to a shorter average stride, and my fastest splits had the longest stride length. Early on, I was springy and fresh, averaging about 1.22–1.25 m per stride, which helped me cruise at roughly 4:45–5:00 min/km. As the miles piled on, my stride length gradually shrank to around 1.10 m or less in the toughest miles – and those were exactly when my pace slowed to the 5:30–5:50 range. Basically, my steps got shorter as I got on to the hills, which had a larger impact on slowing my pace than any small change in cadence.

Insight: Maintaining stride length was key for me. When fatigue hit, I unconsciously took shorter steps; even though I tried to keep my leg turnover (cadence) up, the reduced stride length meant less distance covered per step – a recipe for a slower pace. In my data, stride length had a stronger correlation with pace than cadence did, suggesting that loss of stride length contributed more to my late-race slowdown than a drop in step rate.

Running efficiency metrics from my Garmin – namely Vertical Ratio and Ground Contact Time (GCT) – also tell the tale of a fading form as I tired. Vertical ratio is the percentage of vertical oscillation relative to stride length (how much you bounce upand-down versus how far you travel with each step). Ground contact time is how long your foot stays on the ground with each step.

In the early kilometers, my form was relatively efficient: I had a vertical ratio around ~7.5–7.8% and a ground contact time of ~250 milliseconds per step. Those are decent values indicating an economical stride (minimal bounce and quick foot contact). As I got tired (especially after ~30 km), both metrics worsened – vertical ratio crept up to ~8.5% and GCT extended to 270+ ms in the slowest splits. This means I was bouncing more relative to my stride length and lingering on the ground longer with each footfall when I was fatigued.

These form changes aligned with my drop in pace. In fact, the correlation matrix later showed that my vertical ratio and GCT were very strongly correlated with pace – on the order of +0.84 in each case. In plain terms, whenever I slowed down, I was also bounding less efficiently and spending more time on the ground. It’s a classic sign of fatigue: as the legs lose springiness and form deteriorates, pace suffers.

Insight: My running form broke down as I got tired on the hills – more up-and-down motion and longer ground contact coincided with slower running. Staying light on my feet (low GCT) and minimizing bounce (low vertical ratio) were linked to faster splits. This is a clear reminder that maintaining good form is crucial for keeping up the pace, and it’s something I’ll need to work on (through strength and form drills) to stay efficient in those later miles.

Boston is famous (or infamous) for its hills, especially the series of Newton hills culminating in Heartbreak Hill around mile 20. I expected the uphills to show up clearly in the data as slower splits – and they did. Each split’s total ascent (in meters) gives a sense of how hilly that kilometer was.

The scatter plot above of average pace vs. elevation gain per kilometer shows (unsurprisingly) that laps with more uphill tended to be run at a slower pace. Points further to the right (more ascent) generally sit higher up (indicating a higher min/km pace). There’s an observable trend that segments with significant uphill gain were run more slowly. For example, the kilometer that included Heartbreak Hill (around the 32–33 km mark) had one of the steepest climbs of the day (~17–23 m of ascent), and my pace on that split dropped to ~5:17–5:34 min/km with my heart rate peaking around 150 bpm. This was a noticeable slowdown compared to the ~4:50–5:00 pace I could hold on flat sections earlier. In fact, every time the course went uphill (even smaller climbs of 5–10 m gain), my pace for that split was a bit slower. It’s clear the hills cost me time, as expected. The correlation isn’t perfect – some of my slower splits had little ascent – but all the laps with significant elevation (say >15 m of gain) clustered on the slower end of my pace range.

What about the downhills? Boston’s course drops overall and has plenty of downhill stretches (especially early on, and right after Heartbreak). Interestingly, downhill segments did not correspond to dramatically faster paces for me. The data showed only a weak negative correlation between descent and pace – meaning that if anything, more downhill helped me run a bit faster, but not by a huge margin. In the first 10K, there were substantial downhill portions where I did clock some of my fastest splits (~4:40–4:50 min/km). However, by the later stages of the race, even when the course went downhill or flat, I couldn’t fully take advantage because my legs were already trashed. For instance, the final 5K of Boston is mostly flat or downhill, yet my pace was stuck around 5:40 min/km due to fatigue.

Insight: The hills had a noticeable impact on my race – uphills slowed me significantly (I was roughly 20–30 seconds per km slower on the major climbs), and downhills provided only a modest boost, especially toward the end. This reinforces how important it is to train for terrain: while I managed the climbs at a steady effort (keeping my heart rate high), the toll they took showed up in later miles. I effectively lost more time on the uphills than I could gain back on the descents, likely because the late-race downhills came when I was already exhausted.

Finally, the distance into the race itself turned out to be a strong predictor of pace – basically, the fatigue factor. My splits got slower as the race progressed, which is evident both anecdotally (I felt it!) and in the data. There’s a moderate correlation (about +0.59) between cumulative distance and pace, meaning my pace tended to increase (get slower) as the miles accumulated. I went out relatively quick and hit my fastest 5K early on, then gradually slowed down, with a pronounced drop-off after ~30 km (around mile 18–20).

In Boston, the notorious “wall” often hits runners in the final 10K, and I wasn’t immune. My slowest splits were kilometers 38–41 (roughly miles 23–25), where I was slogging along at ~5:45–5:50 per km. That’s about a minute per mile slower than my early pace. Importantly, those late splits had no significant hills, confirming that it was mostly fatigue dragging me down at that point. My heart rate was still high (mid-140s), but I simply couldn’t turn my legs over any faster. You can see this in the data: the splits with the largest cumulative distances (toward 40+ km) are all at the slower end of my pace range. The final 2 kilometers especially, I was running on fumes – shorter strides, slower cadence, you name it.

Every marathoner knows that feeling when you’re counting down each mile because the wheels are coming off. My data pinpoints exactly where it happened to me: after the big hills, in the last 7K of the race. Understanding this pattern will help me plan future races – perhaps I’ll go out a bit more conservatively or fuel a bit better to delay that late slump.Luckily, I hear that Tokio is flat!

Here’s a correlation heatmap summarizing how various factors correlated with my pace (red tones = positive correlation, blue = negative). Notably, my stride length, vertical ratio, and GCT show some of the strongest correlations with pace, whereas heart rate and descent (downhills) have weaker relationships.

To sum up the analysis, here are the main insights from correlating my Boston Marathon performance metrics with my average moving pace:

• Heart Rate: Surprisingly weak correlation with pace. Early on, my HR climbed while pace stayed steady, and later on, my pace slowed without any drop in HR. In a marathon, a high heart rate in the final miles doesn’t guarantee speed – it likely just means I’m maxed out physiologically. In Boston, fatigue essentially broke the usual HR–pace link for me.

• Cadence vs Stride: Both mattered, but stride length ultimately had a greater impact on my pace. I kept a fairly stable cadence (only a slight drop when completely exhausted), but my stride shortened significantly as I tired, and that loss of stride length strongly correlated with slowing down. My faster splits came from maintaining a longer stride, whereas a reduced stride length in the later stages was a big contributor to my slower times.

• Vertical Ratio & Ground Contact: These form metrics deteriorated as I slowed. I bounced more and spent more time on each foot strike once fatigue set in. The strong positive correlation with pace means poor running economy (higher vertical ratio, longer GCT) went hand-in-hand with a slower pace in my race. It’s a reminder that holding good form is crucial for maintaining speed when fatigue mounts.

• Elevation (Hills): Hills undeniably affected my pace. Uphill sections resulted in slower splits (shown by a positive correlation between ascent and pace). I lost time on Newton’s hills and Heartbreak Hill – evident from the drop in pace during those laps. Downhills gave me some time back, but by the later miles fatigue had muted their benefit. Net effect: the hills contributed to my overall slowdown by taxing my legs early and leading to compounded fatigue later.

• Power (Effort): Interestingly, my running power (measured in watts) showed little direct correlation with my pace. On uphill segments I often had to output very high power just to keep moving (yet I still slowed due to the incline), and in the later kilometers I was pushing hard simply to maintain a slower pace. In other words, raw power output didn’t always translate to speed on this course. Factors like efficiency and endurance played a bigger role – high effort could only do so much when my legs were fatigued.

• Distance/Fatigue: There was a clear trend of slowing down as the race went on. The correlation stats (Distance vs Pace) and my split times confirm that the final 10K was significantly slower than the early miles. This was due to accumulated fatigue compounded by all the earlier factors (hills and form breakdown). Essentially, I hit “the wall” in the last quarter of the race, which showed up in every metric – from heart rate decoupling, to shorter strides, to overall slower speeds.

In a narrative sense, my Boston Marathon unfolded as many do: I started strong and efficient, weathered the mid-race hills at a cost, and then fought through mounting fatigue to the finish. The data enriches this story by quantifying those effects – showing exactly how much my pace dropped, how my body responded (heart rate), and how my form changed when the going got tough. For fellow runners, the takeaway is that maintaining form and pace in a marathon is a complex balancing act of energy management, strength, and technique. When one element falters (like stride length, in my case), the whole system (pace) slows down.

Overall, analyzing all this Garmin data after the race gave me a deeper appreciation for what my body went through over 42.2 km. It paints a clear picture of how heart rate, cadence, stride, hills, and fatigue all intertwined to shape my Boston Marathon performance. And of course, it provides targets for improvement – perhaps more hill training, form drills, and better pacing or fueling – so I can aim for a stronger finish and more even splits in my next marathon. Running a marathon is as much a data story as it is a personal journey, and this data-driven look will definitely inform my training going forward

Running the 2025 Boston Marathon was an experience I’ll treasure forever. It combined the thrill of an elite sporting event, the company of a global running community, and the profound context of history and tradition. I ran on a day that honored the rebels and heroes of 1775, alongside some of the world’s best athletes of today, cheered by entire towns as if we were all champions. I cheered with families, danced past live bands, and got high-fives from kids waving “Run Dad Run!” signs. I passed by colonial reenactors, priests, bananas, elves, and even people dressed as Larry Bird. I took in every second and tried to leave it all on the course. Boston tested my limits with its hills and humbled me with its heritage.

As I sit here writing this, medal in hand and my legs still sore, I feel a deep sense of gratitude. Grateful to the spectators who carried me through Heartbreak Hill. Grateful to the volunteers who handed me water and Maurten energy gels with encouraging words. Grateful to the city of Boston for embracing us runners and making Marathon Monday a celebration of the human spirit. This was not just another marathon for me – it was a homecoming to the roots of the sport I love.

Boston is tough. It’s earned. And for me, it was perfect.

Five down, one to go in my Abbott Majors quest. Boston is now a part of my story, and I am a small part of Boston’s marathon story. I can’t wait to hopefully cap off my Six Star journey in Tokyo and bring it full circle. But no matter where I run, a piece of my heart will always be in Boston, right there on Boylston Street where I realized a long-held dream. For now, I’ll be holding onto the memory of Boston — the race that made me feel like part of something bigger. Something historic.

Boston Strong. Five stars down. One to go.