Mike W Seattle Marathon RR (with Stryd power data)
Overview
Sharing this partly to show the value Stryd data can add. The Seattle Marathon was my first open marathon in over ten years. Based on my fitness, I thought I could go just under 3 hours on the hilly course. I ran by power as much as pace, holding back on the uphills and accelerating downhill relative to other runners. In the end I hit my target with a time of 2:59:21.
Training
Following IM Canada in July, I took five weeks off from training to recover and rest a foot injury, and do some mountain climbing. Then I ramped up my running over a couple of weeks, and followed the last nine weeks of the Marathon Balanced Advanced plan. I averaged 41 running miles per week on the marathon plan, with a high week of 53 and a longest run of 20 miles.
To figure out a realistic pace for the marathon, I looked at
5k time trial -- 17:51 (56.9 vdot), 4 weeks out from the race. Subtracting 2.5 vdot points projects a marathon time of 2:57:30
half marathon -- 1:23:12, run in the spring. Subtracting 1-1.5 vdot points gives 2:56:20-2:57:50
long runs -- pace ranged from 6:42 to 7:14/mile, depending on the terrain and workout goals. Averaging the long runs together would have projected to about a 3:05 marathon
Race Strategy
The Seattle Marathon is a hilly course, with a bit more climbing than New York or Boston. Figuring the hills would cost me a couple of minutes, I was aiming for a 2:59 finish. Many of the hills are in the last 7 miles, and even the fastest runners lose a few minutes there (from some Strava sleuthing). So my plan was to average 6:45 pace (2:57:00 marathon) for the first 19-20 miles, and give back 2-3 minutes in the hills at the end.
I had been using the Stryd foot pod for a couple of weeks before the race, and had the Stryd chest strap (which gives different power values) before that. From my runs with the pod, I figured my marathon pace would be about 285 watts on flat terrain, and my critical power (FTP) was around 315-320. I planned to stay around 300-310 for most of the hills, and 310-330 in the steepest sections.
I used the Stryd foot pod for power during the race, getting power on my Garmin 920XT through the Stryd Power Connect IQ data field (in the normal Garmin run mode).
Nutrition
I woke up earlier than planned on race morning, and was able to eat breakfast a bit early. I had a banana and granola bar with Gatorade 3 hours before the start, which was my last solid food. I took in gels 90, 45, and 15 minutes pre-race with sips of Gatorade. During the race, I had gels at about 0:40, 1:33, and 2:05. I hand-carried a 17 oz Gatorade bottle for the first 10 miles, then took in Gatorade Endurance at aid stations every couple of miles (switching to water when I had a gel). This approach seemed to work well and I didn't have any stomach issues.
Miles 1-8
The first 8 miles had a number of small hills (about 300 feet of climbing and descending) where I could see the difference from pacing with power. I caught myself up around 330W on one of the early hills, and backed off to 300W (and 7:40/mile) while runners around me streamed by. Then on the downhill I focused on good technique and spinning easily rather than fighting the hill, and I went by the same runners at 5:20/mile pace at 265W.
Miles 9-18
This was the flatter part of the course, with only a few rolling hills. I focused on keeping steady at goal pace, which was taking more and more of an effort. My heart rate is creeping up with fatigue building.
Miles 19-26
Overall 2:59:21 281W 166 bpm 6:51/mile
The hilly part of the race. It started on the wrong foot when I stepped in a pothole while drinking at an aid station. Gatorade in the eyes stings! Fortunately I didn't roll an ankle, and was able to get back up to pace quickly.
Just after mile 20 was the steepest climb on the course, including a block at about 14% grade. I got up to 365W (at only 11 min/mile!) in that section but quickly backed off to 300-320W for the rest of the climb. Then a steep downhill, followed by another climb and rolling hills. Mile 26 had a long downhill, but I could only manage 6:40/mile pace down the hill. In the last mile I could feel glycogen depletion kicking in, and I knew I would hit 2:59 as long as I didn't cramp up, so I backed off a little heading in to the finish (other than a 320W push to get over one last cruel kicker).
Stryd power data
Power (orange), speed (blue), and elevation (purple). Some decline in power the last few miles, particularly on the downhills as I couldn't push quite as hard as I could early on.
Power distribution (10W bins, centered around 280-290W). I spent 72% of the race between 270-300W, the main zone I was targeting. Another 13 minutes at 300-310W for the shallower climbs, and 7 minutes at 310-330W for the steeper ones. Only 38 seconds over 330W; when I found myself there I quickly reduced the effort. On the flip side, I had about 7 minutes below 240W; generally these were downhills toward the end of the race where I just couldn't get the legs to propel me faster.
Leg spring stiffness (red). A measure of running efficiency that shows how much of your energy is being recycled from one stride to the next. Higher is better. Mine was around 9.3 for the first 20 miles, but dropped to 9.1 for the last 4 miles. 
Form power (orange). Another measure of efficiency, form power is the energy from your running motion that does not move you forward. Generally, lower is better. My form power was pretty consistent at around 24% of total power. I avoided big increases in form power on the downhills, which can be a sign of braking and less efficient form.
Cadence (green) and ground contact time (blue). Cadence was pretty steady throughout, around 182-183 spm. Ground time increased from around 220-222 ms early in the race to 226 ms over the last 4 miles.
Vertical oscillation (green). Declining starting around mile 22 due to fatigue.
Summary stats
So, what value did Stryd add? During the race, pacing over the hills was a big help, letting me execute the way that I would on the bike. It's easy to get sucked into going hard with the pack early on over the hills, even at well over threshold effort. On the flip side, I could use power as a whip to push harder on the downhills, at least until the last mile when my legs were too fatigued. I've also found Stryd to be a really accurate footpod for instant pace (needing no calibration), so I was able to use both power and pace to stay on target.
After the race, I can use some of the other metrics to adjust my training. I do a lot of strides and this shows in my cadence (which was steady to the end) and my ability to push the pace downhill. But the decline in leg spring stiffness and vertical oscillation show opportunities to improve endurance and power, with greater mileage and drills like bounding uphill.
While my pace and power declined over the last miles, the decline wasn't as sharp as many others, so I think I went out at a realistic yet aggressive pace given my fitness.
0
Comments
Question: How did you calculate your run power zones? I saw that you estimated your marathon and FTP power. Is there any official calculation?
I love the details on Stryd. I've been toying with getting one. This convinces me that I must have it.
Have you used Stryd off-road? I've wondering how it does on trails where the surface can have a big impact of speed and effort.
I am behind the curve on utilizing Stryd but your walk through put it into perspective. It definitely gave me motivation to integrate it into my run sessions. I will test with it on Sunday to nail down FTP paces.
@Ashton - Stryd has 3 ways to calculate FTP or baseline metrics.
1. 3-6 lap test, run evenly to near exhaustion
2. 3'-9' test, run evenly to near exhaustion
3. 5k or 10k TT
4. Self Test - direction not provided on how to conduct this one.
Thanks! I agree on the value of Stryd in the hills. When I started using it, I was surprised how much power I was using on even small hills. On a long enough hill the extra effort would eventually show up in heart rate or RPE, but with Stryd I could adjust much more quickly.
Stryd has a test to determine your critical power (FTP), which involves running 1200 meters and 2400 meters all out. I've done it with the chest strap but not the foot pod (and the two give different power values), so my critical power with the pod is more of a guess. To get my marathon pace zone for the race, I looked at training runs where I had done a few miles in z2, and used those sections to estimate the power.
Jim Vance wrote a book called Run With Power that goes into running power zones.
Agreed on the marathon plans! I thought the plan struck a great balance between volume, injury risk, and time commitment. It was easy to ramp up to get into the plan, and I felt the most prepared that I've been going into a marathon.
I just looked at the EI. It was 0.843 meters/min/W for the first 8 miles, 0.842 for the next 10, and 0.823 for the last 8.2. EI doesn't take into account hills, though, and will tend to be lower in hilly terrain. In the last 8 miles the EI per mile was 0.83, 0.80, 0.75, 0.86, 0.82, 0.85, 0.86, and 0.87. So the main reason it was lower was the climbs in miles 20-21.
I haven't, but I know trail running is one of the big use cases for Stryd due to all the elevation gain (and a more challenging environment for GPS). You're right that the surface can be an issue, and Stryd doesn't currently have a way to differentiate there. But I think there would still be a lot of value for uphill pacing.
Another limitation of Stryd for trails is that it isn't designed to measure power while walking, so you might get zero readings for a while if you drop to a walking pace uphill.
I was an early adopter of the chest strap version , at the time I could not get it to display power while in run mode on the 920xt , so I used it as a HR monitor , it broke twice and they replaced it both times, great customer service, the last one they sent me is still in the box... Looks like its time to dust it off and try it out again.. Maybe even try the foot pod version?
How difficult was it to set up in run mode on the 920xt with garmin IQ ?
How does it affect the TSS in TP?
Do you use pace or power to calculate your TSS?
What is your opinion of wind in this equation of run power?
I have the Jim Vance book but have not read it yet. Have you?
Andy Coggan (who I absolutely can't stand) is running some threads on ST (he is receiving compensation from STRYD) states that wind is not really a factor... Having just run a downhill longrun into a vicious headwind last weekend would disagree :-) Did I mention I can't stand him? He maybe smart but an A$$!
http://forum.slowtwitch.com/forum/?post=6152797
The new Stryd footpod ("Summit") still reports walking power as '---', but a firmware update is planned to make it function the same way as the Pioneer.
Having used both, the foot pod has a few advantages over the chest strap for me. The chest strap power readings can vary based on how it is positioned (and Stryd even has a calibration procedure now to try to adjust for that), and I have had it slip during training. The foot pod gets some additional metrics (leg spring stiffness, form power) and improves on others (vertical oscillation for the center of mass rather than your chest). It's also very accurate for instant pace in my tests, much better than a typical foot pod and needing no calibration (though some people have had issues and Stryd is setting up a calibration process for them). One downside to the pod is that the clip fits very tightly on some laces, and moving it between shoes can be a pain.
Setup on the 920XT was pretty easy. I just used Garmin Connect to add the "Stryd Power" data field in the Connect IQ store, and then added the field on my watch to one of the data screens. The pod is paired as a foot pod, and power comes in automatically. So far you can only get instant (3s average) power. Another option is to use the Stryd Connect IQ running app, but then you would lose the ability to customize your running data screens.
On TP TSS, you need to set up a separate FTP for running in TP; otherwise your IF will be way too high. Even then, I find the IF is a bit higher than it would be when calculated from pace alone. Running has a higher power requirement for "just moving" and a smaller spread than biking going from z1 to z5, so any run with an IF below 1 will end up with a higher IF based on power than it would based on pace. I don't worry about it too much and figure it will get baked into the PMC over time.
TP is calculating my IF based on power; I don't know if there is a way to switch the priority to pace when both are present.
Wind can be an issue if you have have sustained winds of 20 mph+, but I don't run many races where that's the case (and if I did, I would be tucking in behind someone bigger). Below that, the impact is just a few seconds per mile IIRC. A bigger issue to me is the running surface, which Stryd also can't correct for. And if you are on a long run carrying fluids and don't update your weight, the power will be off by a few percent (since Stryd measures W/kg).
I've read Jim's book. I thought it was helpful in explaining the basics, but a lot more is needed on how to use the new metrics to improve training. Yes, you can just take a standard training plan and replace pace zones with power zones, but power (and the other running metrics Stryd measures) should offer more feedback and opportunity to customize training based on your strengths and weaknesses and race goals. I'm hopeful that we will see more developing there, in the same way that cycling power training has progressed over the last 30 years.
Mike, If and when you have some spare time, I'm in search of a "FOR DUMMIES" guide on how to make my 920XT and my Stryd Footpod have a proper "conversation"!!! .... I downloaded IQ ..... I connected Stryd as a foot pod .... and I'm not getting any power readings. Further, I tested STRYD on the treadmill and the cadence calibration is way way off (half a mile on a one hour run) .... Maybe you know of a link within STRYD forums (which I haven't yet found) that might enlighten me... HELP!! Tks
Juan, it's possible that you got a dud. Stryd has a new support site with chat/email links that might help: https://www.stryd.com/support
Some things that I would check:
On the treadmill run, I wonder whether Stryd was really connecting at all. If it wasn't, the distance is probably based on the accelerometer in the watch, which could easily be off by that much.
As i remember from my marathon career, seattle was by far the toughest marathon i have done. Course wise. I think that is where i bq 10 years ago.
Kind of hankering to run a marathon myself. Maybe 2018.
Maybe consider doing one of our Steilacoom running club races this winter/spring. 5 and 10 and 15 and 20. Mile or km series.
Thanks, Robin! Yeah, it's a tough course with the hills at the end. But the weather this time of year can be good for a fast marathon (if it isn't raining).
What's the Steilacoom course like? I've seen a lot of races there on the run calendars.
I'm looking at a couple of halfs in the next few months, probably Geoduck Gallop in Olympia February 4th, and the Mercer Island Half March 19th. But I might mix in some shorter races too.
I've had both versions. I stopped using the chest strap version as I didn't really like it, didn't think it was that accurate and it was messing up all my bike data. The new footpod version seems to be much better. I've mostly been collecting data, but reports like yours are really helpful learning to take the data and create actionable items. Training Peaks now accepts running power and WKO 4.0 also breaks out biking and running power so the reporting is getting better.
Any other race reports or insights you could offer on getting the most out of the Stryd would be greatly appreciated!
@Juan. The power meter does not record power when walking. For the treadmill pace, that is a known problem that they are working on. I talked to tech support on the issue. The foot pod thinks it's going backward on a treadmill and it's throwing off their algorithm for speed / pace. They are working on updating the math or finding a way to tell the foot pod it's on a treadmill. No word on timing. My Stryd pace is about 20-25% slower that what the treadmill shows. I have a 920XT. When I start up, I bring up the screen that shows power. It typically says "searching" when first connecting. I'll do a short burst of speed to get the watch and foot pod to sync, it will show a number in the power field, then I'm good to go. Also, use the phone app to check the battery charge level. It's not obvious how to charge it. I had it upside down (large flat side down) and had to get tech support to tell me to flip it over to charge it.
I am just getting started with Stryd and also have a goal of a 2:59:59 Marathon. (My current PR is 3:00:57)
By any chance, would you mind sharing your race weight for this marathon? I am curious as to what a reasonable W/kg ratio is for a 3:00 marathon.
Thanks!
I have a feeling a large part of my overtraining + speed problem is that my TSS was waaaaay to high for what I needed to do...
@Peter Noyes I think I was around 155 lb / 70 kg for the race, so racing at right at 4 W/kg and critical power around 4.5 W/kg. Weight is huge for running but efficiency also makes a big difference.
@Coach Patrick TrainingPeaks has me at 233 TSS for the race (0.88 IF, 1.01 VI). That's based on a guess at my CP with the footpod, which might have been a little high; I'd guess the race TSS was somewhere between 230-245. I know there has been some discussion about whether the cycling TSS formula should be changed for running power, and Andy Coggan is looking at altering it.
@Tom Box I pulled the numbers for each mile split out of TrainingPeaks and then did some of the calculations in Excel. I used WKO4 on free trial for a couple of weeks and even set up some custom charts, but there were some annoyances that so far have kept me from buying a license -- it can't handle the files with Connect IQ data straight from Garmin Connect (even though TrainingPeaks now can), some things I had fixed on TP didn't carry over (like elevation corrections from the time when my 920XT altimeter was out), and I just found WKO4 less intuitive to use overall. Unfortunately WKO4 is the only place I've seen to do some of the analysis I want, like looking at efficiency, leg spring stiffness, vertical oscillation within an interval.
@Tom Glynn If you found this report helpful, you might want to look at this one of a professional runner from the New York Marathon (which inspired a lot of my analysis): http://running.competitor.com/2016/11/photos/understanding-running-power-marathon-via-stryd-data_158787