How did Adam Fogg go from 3:55 to 3:49 in the Mile in Two Weeks?
This essay will be a deep dive into Adam Fogg’s training and racing leading to his breakthrough performance, a 3:49, in the Millrose Games Wanamaker Mile. Fogg has put in a relatively long consistent training block starting in late September and running straight through to mid-February, and he’s still at it. His next race is the UK Indoor 1,500m Championship, in which he will try to qualify for the World Indoor Championships in March in Glasgow, Scotland.
Adam Fogg is a British-Australian (since 2021 eligible to represent GB, according to World Athletics) professional distance runner representing Under Armour and based in Baltimore, Maryland. He ran collegiately in the United States for Drake University in Iowa from 2019-2022. His collegiate career was solid but not indicative of what he is currently doing, finishing up with a 3:56.60 PR and sneaking an NCAA Division 1 All-American honor in the indoor mile. And it took him a first big breakthrough (running the race of his life to that point at indoor NCAAs) to even get there. Before that, he was a competitive, but not AA caliber, D1 athlete sitting on a 3:57 PR, which does not even come close to qualifying for nationals anymore.
So what has he been doing since then? All of his training is publicly available on his Strava. Let’s go back to the beginning of this training block to figure out how he went from 3:55, which he ran in 2023, to 3:49 in two weeks.
It all started at the end of his vacation in Portugal on September 25 with a 10k run at just over 7:00/mile pace. The next day, he was back in England for another 10k run. That first week, he ran 62 miles (100 kilometers) with an introductory session of 6x5 minutes+4x10 second hills. The 6x5 minutes appears to have been run at his threshold at that time (3:15–>3:08/km pace or 5:15→5:04/mile).
Over the next four weeks, Fogg would build his mileage up to 86 (140km). His long runs were built from 12 to 18 miles and done at varying fairly fast paces. His easy runs continued to be between 7:00 and 7:45 pace, certainly quite easy, even at that time of year, for a professional runner. He also seems to have done some speed development work once per week, which in modern training programs are usually things like short hill sprints, track 40s or 60s, and/or plyometrics. He did not post the work on Strava nor specify what he was doing, with the exception in late November he did 6x10-second hill sprints. But this cannot be underestimated as a big part of his success as a miler. He also seems to have lifted weights once or twice per week, but he did not go into detail about what he did in the gym. This essay is going to focus on his workouts.
His session the next week was 3x2 miles from 5:30→5:20 pace but with only 90 seconds rest, then 6x45 second hills. We should note how slow 5:30 pace is for a future 3:49 miler. Fogg was probably running these at tempo (Zone 3). But the key here is that Fogg was not in 3:49 shape when he ran this workout. He was training at his current fitness level, not his goal fitness level. This was his only session this week.
The next week he did 3x2 miles again at a pace I cannot tell (he ran laps around a turf field which notoriously throws off the GPS measurement), but probably on the easier side of threshold. He came back that evening for 10x800 2:26→2:20 on 1-minute recovery. That’s a much heavier session than the AM one; likely cranking well over Z4 by the end. But I see this as similar to a Norwegian PM 20-25x400 session. At the end of the week, he paced the Baltimore Marathon for 16.5 miles at 5:29 pace, which was a very aggressive long run for him at that early point in the build-up. If his heart rate monitor is to be believed, he averaged 158bpm for 90 minutes, which sounds low, but his Strava zones indicate that he was in Zone 4 for much of the effort, too hard for a continuous effort of that duration done more than once every few weeks. But he was pretty disciplined in taking some long runs easier than others throughout the build-up. Perhaps his HRM was off or his zones are inaccurate (more on that later). Nevertheless, this would have provided a huge stimulus, if he could properly recover and absorb it.
His first session of the next week was 12x350ish meter hill repeats with jog-down recovery, and about 4 minutes after 6 reps. He did these between 4:00 and 4:20/mile effort based on Strava’s Grade-Adjusted Pace feature. The hill was between 5 and 6 percent gradient, which I think is perfect for this sort of session. This is a nice mini-VO2 max stimulus while also working neuromuscular power and running economy. At the end of this week, he did 5xmile AM (5:14→4:58) on 60-second rest. In the PM, he did 10x800 on 1 minute (between 2:24 and 2:20). There was no hard long run this week; instead he strangely did 12xhill again on Sunday. These hills progressed to be a bit quicker than earlier in the week, going sub-4 effort on the last few. This was a big week. Also, it is clear from these hill sessions that Fogg straps up for most workouts with an accurate HRM.
His first workout of the next week was 17 miles in Manhattan (mostly in Central Park) at 5:37 pace. He did a 5-ish mile progressive warm-up, 11 miles hard around 5:20 pace, then a mile cool down, all continuously. His heart rate was much more impressive than his Baltimore Marathon pace job at 150 bpm and only sneaking into Z4 for 10 percent of the effort. That’s big work for a miler doing a 17-mile effort.
To me, this makes a few things clear. Elites get out of shape just like the rest of us, it’s just that their out of shape is still pretty fucking fit. But the more important takeaway is that elites get in shape extremely quickly. Fogg was improving rapidly months away from his target races.
His second workout day of the week was 3x2 miles from 10:37 down to 10:14 on 75 seconds; super easy. His PM session that day was 12x300/100 jog. All of the 300s were between 47 and 49 seconds, with the 100s well over 8:00 pace, giving him a 33-37 second moving recovery. He covered 4,800 meters in 17:10. This is a nice little VO2 max stimulus with 300s around date 3k pace and short active recovery to keep the HR high, with the obvious added benefit of specific running economy work at a pace not far off of his mile rhythm.
On Halloween, his session was 12xhill again. These hills were quicker still than last time (getting well under 4:00 effort for most) but at about the same heart rate. A few days later, he went out for 17.6 miles at 5:24 pace, again at 150 bpm. He progressed for a while, split a 1:08 half marathon in there, including the last four miles at about 5:10 pace, before running a quick cool down. He did not stop once. I’m most impressed by the same heart rate at a much faster pace than the previous week. FoggDog was in for a big season if he was in such flying form already in early November.
The following week he began an altitude camp in Flagstaff, Arizona. After some easy days, his first session at elevation was 6xmile 5:10→5:00 on 90 seconds recovery. You can see that he took more recovery between reps on his workout at altitude. His PM session that day was 2x800 in 2:32, 2:30 on 75 seconds, then 12x300 in 51→45 on 30-40 seconds rest, but 2 min in between sets of 4. I like this session. His long run for this week was 17.4 miles at a meager (I say that in jest) 6:13 pace, which is certainly wise considering the work he had been putting in and especially the altitude. His average HR was 144 bpm, certainly much higher than it would have been at sea level at that pace.
After four weeks in which he averaged 86 mpw, he took a down week at 75 for his second week in Flag. On November 15, he did 3x(3x600, 350) hills. This looks like a long, grindy, VO2 max session. He ran the 600s at around 4:35 average GAP and the 350s under 4:00 GAP. Looking at workouts like this on Strava reminds me why I retired from running. The suffering of a long hill workout at 7,000 feet elevation is unbearable and I would have perceived it as eternal. His double session that week was AM 3x2k/90 seconds at 6:37 down to 6:26, PM Mile/600/4x400/200/600 (5:00, 1:33, 68, 31, 67, 30, 66, 29, 65, 29, 1:27).
The following week he ran 4x800, 4x300. 2:21→2:15, 45, 43, 45, 43. He did this session down in Scottsdale so he could run faster paces at sea level. It seems he flew to England that evening. This was a pretty short altitude camp at just 2.5 weeks, certainly shorter than is generally recommended by exercise physiologists. Also, this was just a 56-mile week, which makes sense as he was about to race. He finished up the quality for the week with a 1-mile cross-country race in which he ran 4:26 for second place. Time is utterly irrelevant in the British cross-country scene as the courses are gnarly. After the race, he did 10x2 minutes/1 minute with the efforts around threshold and the recoveries not much slower than 6:00 pace.
The following week he went back up to 85 miles. Later during this week, he finally did what might be considered a bit of a banger workout to the layperson. The session was 5x1k at 2:41 average (2:36 finish) on 3-minute recoveries. This (1k reps around 5k pace on equal rest) is a very traditional VO2 max session. He would have gained a lot from this one with the huge base he had just spent over 10 weeks building. He tossed in 6x30-second reps on grass for some turnover after the main set. His long run for the week was 14 progressive miles averaging 5:48 pace. That’s pretty lite werk for FoggDog at sea level. He did not do another real session this week, illustrating the importance of recovering from (and coming fresh into) the big effort the 1ks surely were.
He also announced this week that he had been selected for the British team for the Mixed Relay at the European XC Championships, which was probably a contributing factor for him to blast that hard session which was far different from what he had been doing up to that point. I’ve always felt that you only need one or two real specific sessions like 1ks at 5k pace to get near (or to) your peak form. With European Cross just a week away, I have a feeling Fogg and his coach drew this one up on the fly to find some racing sharpness for the big target of December.
His race-week workout was a mile at threshold (4:53) followed by 8x400 in 65 down to 60, then 6x30 second grass reps. At the Euro XC Mixed Relay event, he ran his mile leg in 4:32 to keep his team in the lead. The course must have been ridiculous for a guy about to run 3:49 to grind out a 4:32. This being another race week, he only ran 56 miles.
We are well into December now, and he started the next week with an easy long run of 15.5 miles (25k) at 6:28 pace (4:01/k)*. His session for this week was 8x(600, 400) hills. That’s 5 miles of running uphill, all around 4:10-4:20 GAP. The hill was not steep, around a 3-percent gradient, so his actual pace uphill was still well under 5:00/mile. The downhill jogs were taken at an honest pace, mostly under 7:00/mile. He took about a 4-minute jog between sets, but his pace for the whole 10-mile run was still 5:47. This is an excellent strength-building session. He ran 75 miles this week.
*Apologies to my non-American readers, I am simply not going to convert every single pace or split to kilometers. If you are a runner, you ought to have it all down by now anyway.
Still in England, the following week he ran 89 miles. His opening workout was 3x2 miles at sub-T (aka Tempo or Marathon effort or Z3 or LT1 or aerobic threshold to cover everyone’s different way of referencing the same intensity) on 75-second jogs. 10:35, 10:23, 9:59 were the splits. In all, he covered 6.5 miles continuously at 5:16 pace. But that’s still wicked easy for a runner like Fogg, especially at this point in his training. His heart rate crept into Z4 just 2 minutes of the workout. He came back in the evening with 16x600 at 1:42 average on “100 walk/jog in 60 seconds”. That’s 68-pace (4:33/mile) for a lot of work (9,600m or 6 miles) on short rest. I’m seeing major parallels to the Norwegian method’s PM 25x400 workout. The idea is just to push over Z4 slightly in short intervals and on short recoveries so you get in a huge stimulus of both running economy at close to longer-distance race pace (maybe 10k-pace) and aerobic work because the heart rate and lactate do not drop too much in between reps. This of course piled on top of six miles in the AM in Z3. So FoggDog ran 12 miles of quality on this day. It of course makes sense that he takes his easy days quite easy to recover.
That being said, he did his fastest long run of the block to date that Thursday; 18 miles at 5:22 pace. His heart rate was surprisingly 157 bpm (much higher than it was at a similar pace about a month prior, despite clearly improved fitness), but he said it was quite windy (and his Strava post says it was a howling 24 mph!) That can shoot the HR through the roof.
On Sunday, he did the 8x(600, 400) hill session again, with pretty much the same overall effort as prior but the hills were just a bit quicker and the jogs a bit slower. All told, his average pace was 5:45 for 10 miles. The only comment he made on Strava was “sufficiently bleak.” I certainly resonate with that.
On Christmas Day, which was the start of the final week of the year, he did a half marathon!... at 6:59/mile… Just an easy medium-long run. His first session of the week was 6x1 mile at 5:00 on the treadmill with 75 seconds rest. In the afternoon that day, he did 10x1k on the treadmill in 3:00 with 60-second recovery. These are both pretty standard threshold sessions, if not on the easier side, for a 4:00-miler, let alone a 3:49 man. The next day he did a strange 10ish mile run that took over 90 minutes due to 1,500 feet of elevation gain and “horrendous” footing.
After an off-day, he went for a real ripper: 30 kilometers (18.6 miles) at 5:11/mile pace, crushing the last 10k right at 5:00 pace. His heart rate was 156 bpm, spending 27 percent of the run in Z4, with pretty much the rest in Z3. But at the pace he ran, that’s quite incredible work. This was his last quality effort of the year and helped cap an 84-mile week.
His first session of 2024 was 5x(1k, 600) essentially broken miles. He averaged about 2:46 and 1:37 for the reps, with 90 seconds jogging between the reps and 2:30 between the sets. The “miles” would have thus been about 4:24 on average, which is a nice lower Z5 effort. He flew back to the United States after this session.
The first quality effort he did back in Baltimore was 2x 3k, 1k with 75-second jogging recovery. He ran these easily, averaging about 5:04/mile for the ons. That’s still a bit quicker than his AM components of the double-threshold days had been up to that point. That afternoon, he ran 3xmile on 60 seconds, then 12x300 on 100 jogs at 4:56 and 47 avg. He continued to do almost every workout well under control and significantly slower than many runners who race way slower than he does. This was a 70-mile week.
The following week he ran his highest mileage of the entire block, at 92.4. His long run was 18 progressive miles averaging 5:43 at 150 bpm, which is not a great heart being so much slower than his big effort just 10 days prior. But this shows that heart rate can fluctuate a lot from day to day and linear improvement is not always possible. Torturing oneself over its ups and downs is wasted emotional energy (I had to learn that for myself the hard way.) Anyway, it’s wise to back off from hard long runs often, so cruising 5:40s (still not super “easy” for a runner like Fogg) was prudent. Most of the hay, in terms of his aerobic engine and general endurance, was in the barn at this point.
That being said, it was time to start tuning up for the indoor season. FoggDog hit the track that Thursday for his first real speed endurance or sharpening session of the block. The date was January 11. That means he had been training for about sixteen weeks before he even started ripping hard 400s, which is specific to running a fast mile (and he only did ONE during this session, which was mostly 5k-focused!). If this does not help to illustrate how important an aerobic base is and how little sharpening one needs to run a mile PR, I do not know what does. Let’s discuss this hard workout in detail: He started with 5x1k at 2:44 avg. on 2 min rec, then did 2x(300, 200), 400, 200 at 44, 28, 42, 26, 53, 26 on 45 seconds in between reps and 2-3:30 between sets. That’s a big dose of lactic acid, especially the last two sets with short rest between 42 and 26 and then 53 and 26. But racing a 3:49 mile induces a massive lactic acid figure, so one needs a bit of preparation to handle that. That being said, even this session was endurance-first with 5k of work at 5k rhythm before he even touched the real traditional speedwork.
Commencing the succeeding week was a 13-mile run with 2x5 minutes and 5x2 minutes at 5:05ish pace and then 4:40ish pace, on 90 seconds between all. This would be classified, in my running lexicon, as a mini-workout.
But it makes sense, as he opened up his indoor season on Saturday, January 20 with a 1:48.38 indoor 800, good for a 3-second PR. I don’t know how he chopped 3 seconds off his 800 with this training. He basically did a single workout with a bit of 800-pace work, just over a week prior.
I make two observations: talent is talent, and aerobic training is truly king. First, only a true elite miler could run 1:48 after barely touching that pace besides in short sprints for many months. And next: strength = speed, or so that old saying goes. A slow-twitch athlete can run faster than we may have previously thought in the mid-distance events if properly developed aerobically.
But with his engine and having run 1:48 early in the year, big things were obviously coming. I also observed that even with this breakthrough, he is essentially a Jakob Ingebrigsten-lite runner with a 1,500m focus, subpar 800m credentials, likely lackluster sprint times, but a huge engine that allows him to hold a ludicrously high percentage of his maximum speed for 3.75 or 4 laps. The reason for this is not because he was born running 13:00 5ks (not that he can do that right now, but I bet he is in better 5k shape than most observers would estimate), but because of his prudent aerobic training approach, part of which is based on the imperial (my adjective) Norwegian model.
He would finish the race day by pacing the 3k with 1.3 miles at 4:09 pace, a solid workout. He then finished the week at 69 miles.
The final week of January was another race week, starting with a workout of 4x800, 2(2x300). He was probably hitting about his threshold at this point on the 800s averaging 2:21 on 75-second recovery, then going 44, 41 for both sets of 200s on 45 seconds between reps and 3 minutes between the sets. Again, he was doing a race-week tune-up workout, but you can see how he’s just not doing much volume at mile pace or faster. Little doses coupled with a massive aerobic base are all one needs.
He PRed in the mile that Saturday in the Armory in New York City in winning the Dr. Sander Invitational in 3:53, which at the time was a 2-second personal best. This also qualified him for the Millrose Games Wanamaker Mile two weeks later. I noticed this result at the time, but 3:53 milers are practically a dime a dozen at this point in international terms (even plenty of NCAA athletes do that nowadays) so I wasn’t exactly bullish about his prospects squaring off against Yared Nuguse and company in a world record attempt at Millrose. I should have known better. He did another mini-session post-race and an easy long run the next day.
The next week he started with a double-threshold day of 2/2/1 miles in the AM at his usual Z3 effort except for the 1 mile rep, which he ran in 4:51. In the afternoon, he came back with 10x1k at 2:58 down to 2:54 on 75 second jogging recovery. 2:54 is rolling for a threshold rep, but with what Fogg was about to do, it’s right in line with what he should be running. His heart rate was placid.* That was a banger of a double-T day.
*All heart rate analysis is based on his Strava heart rate zones, which may or may not be accurate. A lot of people overestimate their max HR which inflates their ego when they run sub-5:00 pace for a threshold rep and their supposed HR was 88 percent of maximum, when in reality it was 96 percent of maximum and they raced their workout, to their own future detriment. I’m not saying Fogg does this; quite the opposite. He’s a 3:49 miler now, and 3:49 milers can do some unbelieve things in training. But Fogg doesn’t do that either, he’s quite smart and consistent. Rant over for now.
I figured he would do his last (and kind of his only) big session of the block about 10 days before Millrose (this is a traditional American coaching theory, based on the idea that one absorbs the training effect of a hard workout about 10 days after it is completed. That may or may not be based on science). He did his big effort 9 days prior to the race. The workout was 800 at 1:57 (but cutting down from 60.x to 56.x for his 400m splits), 6x500 at 1:20→1:16, 3x400 at 57, 57, 55. That’s a good, but still not totally bananas, workout for a 3:50 miler. He took long set rests but short rep rests; which is standard practice for a miler in a sharpening session (there are physiological reasons for this, which I would boil down simply to shooting the lactic acid to the moon within the sets, but then recovering enough between the sets to get more volume done at a quick pace). Still, despite the fast paces run here, it does not indicate 3:49 fitness. But that plays into the point I am trying to make with this long essay: one does not need to indicate anything in training. Consistent, disciplined work will get the job done. This was a 62-mile week.
Fogg started his Millrose race week with a 13-mile long run including 2x10 minutes steady on a 3-minute float. He averaged 6:12 for the whole run and 5:32 for the steadies. That’s another mini-workout. He didn’t do a double-threshold this week; his only real session was 4x800, 4x300 (2:21→2:15, 43 avg.) That’s a light tune-up for the lad.
Now we come to the reason I decided to spend hours writing this piece: FoggDog’s 3:49.62 mile at the Millrose Games. The race did not even play out ideally for him; too fast up front with Fogg forced to lead the chase pack and a large gap up to a flying Nuguse, Hobbs Kessler, and George Mills (Nuguse was attempting a sub-3:47 world indoor record, and the other guys were right on him). The pacers took the race out way too fast through 400m and 800m, and after they stepped off, Nuguse front-ran a way-too-slow 59-second 400m split. The record was basically out of reach at that point, but Nuguse kept grinding in the front. Fogg made a heroic attempt to bridge up to the leaders with a few laps remaining and almost did it because of Nuguse’s major error. But the 3:43 outdoor man wound it up from the front and eventually held on to win in 3:47 to just miss the World Record. Kessler ran a huge 3:48 PR with Mills right behind. Fogg ran his 3:49 with brilliant tactics, almost completely even-splitting the entire effort despite the fartlek run in the front of the race. He of course backed up his smart racing strategy with the legs to still muster a 57.22 final quarter-mile after the brutal opening pace.
Not many people in world history have run an indoor mile under 3:50, yet Fogg was just 4th in this race. There’s no reason that indoors is slower than outdoors anymore; in some cases, it’s faster. There’s a lot of debate about that hot topic, but I am not going to give Adam Fogg (as good as he is), of all people, 3 seconds for free and a 3:46 outdoor mile. That’s what some old conversions used to suggest. He’s a 3:49 guy for now; world-class, but not good enough to compete for an international podium yet.
So next up, Fogg will attempt to qualify for the British team for World Indoors in the 1,500m at the UK Indoor Championships starting tomorrow. Whether he manages that or not, it has to be expected that he will be shooting for the Olympic team with Paris 2024 this summer. Great Britain’s 1,500m line-up is stacked with the last two world champions in the ranks, but if Fogg can make one more jump, he’s got a decent shot. I think he can do it if he responds well to a bit more sharpening work.
So that is how Adam Fogg went from 3:55 to 3:49 in 15 days and is poised to achieve even more in the future. Consistent training at the right effort will get the runner further than they might have thought, especially when coupled with elite talent like the athlete in question here.
Fogg is an interesting story. I’m hoping he can mix it up even more this summer. I think he has a lot more room to improve given his relatively young age and the potential for him to increase training volume in the future.