70.3 Coeur D'Alene race report - John Katsoudas (run pacing with power experiment)
70.3 CdA was my “A” race this year and my first long-course triathlon since I raced Vineman 14 years ago. I started training for triathlon again last July from a completely untrained and overweight state. At the time, I weighed 188 pounds and couldn’t run one mile without the pain in my left foot flaring up that had plagued me chronically for over a year. The first order of business was addressing body comp, and over the last year, I’ve worked my way down to 155 pounds, which is the lowest I’ve ever trained and raced at. The last year of training has not been without its share of injury, illness, and competing priorities (holidays, travel, etc.). The TrainingPeaks PMC doesn’t lie when it comes to illustrating gaps in consistency, but despite what the numbers and charts show I feel like it was a very successful and solid first year of training. My primary goal for the season was to show up happy, healthy, and ready to race on race day. All other intermediate season goals and race time goals were secondary to this, so in this regard, the season has been an unmitigated success.
I also wrote a blog for family and friends, along with some photos, which give an overview of the weekend:
So this post will be primarily a race execution discussion.
Swim:
Goal: 40:00
Actual: 43.27
I liked the rolling start format of the swim, as it eliminated some of the race start craziness. I seeded myself in the 36-45 min group and felt like I was surrounded by similarly paced people most of the swim. The most notable thing on the way out was just how cold the water was, and how long it took me to adjust to this and get over the gasp reflex and settle into a decent swim form. Maybe a neoprene cap might have helped, but from my SCUBA diving days in Monterey, I know that most of that gasp reflex comes from cold water on the face and not so much the head. I chose not to go in the water before the race, as I didn’t want to stand around shivering for the next 30 minutes. Still not sure what the right way to deal with this in the future is.
Another noteworthy thing is just how bad my open water sighting is, particularly swimming with the sun on my right (breathing) side. The GPS plot of my swim shows a pretty zigzag pattern, and I found myself inside the buoy line on the return leg more than once and had to veer back out to the race line. My Garmin 935 recorded over 100 yards more distance than what Strava shows other people swam, and it shows me averaging 1:55 per 100m, as opposed to the ‘official’ pace of 2:15 per 100m. For someone that’s already a slow swimmer, poor sighting is just adding insult to injury and throwing away minutes of time unnecessarily. The last thing that surprised me was seeing an HR in the high 160’s as I exited the water. My tendency on long swims is to ‘settle in’ to a comfortable pace. I remember at one point telling myself that ‘comfortable’ is not a word that belongs in the race day vocabulary, so I guess my inner-self took that to heart, as my avg. HR for the swim was 161. That was a bit concerning to me as I exited the water and headed for T1.
T1:
Goal: 2:00
Actual: 2:57
This was a big transition area as compared to the sprint tri I did a few weeks ago, so much more ground to cover. T1 went pretty smoothly. I had rehearsed my route from swim exit to bike the day before, and after the wetsuit peeler took several hard tugs to get my wetsuit off, I was up and running and found my spot easily. Helmet on, grabbed the bike (shoes were already clipped in and rubber-banded) and off I went. By the time I hit the mount line I felt like my HR had settled a bit and I was good to go.
Bike:
Goal: 2:38:00
Actual: 2:58:51
This kills me. I’m not a fast swimmer or runner, but I consider myself a pretty decent cyclist, so if there was one leg I was looking to shine in, it was the bike. I had run multiple power models in BestBikeSplit and rehearsed them at least 4-5 times in Zwift and TrainerRoad, tweaking the profile each time, so I knew I could ride the power targets entirely in the aero bars for 56 miles. This ride was just piss-poor execution from start to finish (choice of wording intentional, as you’ll see in a moment).
The first issue is that, a few months ago, my old Garmin Edge 810 started not holding a charge for very long so I replaced it with a fancy new Garmin Edge 1030. Well, folks, it turns out that the Connect IQ data field for the 1030 that’s supposed to show me my BestBikeSplit power target for each segment is buggy and doesn’t work on the 1030. I spent a LOT of time over the last several weeks on the phone and trading e-mails with Garmin and BestBikeSplit, posting desperate pleas for help on forums and playing with all the settings in hopes that I could get it to work correctly. I never did fully get it to work, but I finally had a decent workaround established. The reason that this is significant is that I continued to play with the Garmin settings on my road bike setup even after I had already shipped the tri bike off to the race via Tribike Transport. When I arrived in Coeur D’Alene, re-installed my Garmin Vector 3 pedals on the bike and turned on the Edge 1030, it happily found the pedals and said I was good to go. Guess what I forgot to do? 1) reset crank length back to 165mm (road bike is 172.5mm) and recalibrate torque angles on the pedals. There’s a lesson here: whenever removing Vector pedals from the bike, also remove the batteries from the pedals. This will force a recalibration when you re-install the pedals.
During the race, I felt like I was following the BBS power target prompts pretty faithfully, but they all seemed ‘easy’ as compared to all the training rides. HR was where I expected it, so I just chalked it up to freshly rested and tapered legs. Because the ride felt easy I did increase the target power on the descents a bit, never going over the target avg. power for the entire ride, but not dropping as low as BBS instructed. All was good, or so I thought. Post-race examination of the race file shows a left/right balance of power of 75%/25%, which those of you that own Vector pedals know is an instant sign that the data is bad. There were no dropouts in the file (I’m running the new battery compartment doors and fresh batteries), so this was purely a calibration issue. The difference between crank lengths accounts for about a 5% lower power than shown, and not recalibrating the pedals resulted in some unknown additional error. Long story short, I inadvertently under-paced the bike despite following the BBS model almost spot-on.
Now for the second problem. I have spent a significant amount of mental energy over the last six months trying to come up with a workable nutrition and hydration strategy for the bike. My main concerns were taking in enough fluids, calories, and salt. After some experimentation and lots of input from folks on the EN forums, I had what I thought was the winning formula. One bottle of Gatorade Endurance every 45 minutes, one sleeve of Clif Bloks at the 1 and 2 hours marks (spread out over 20 minutes and chased with water), for a total of between 950-1,000 calories. I had rehearsed this many times, and it was easy to stomach, simple to follow, and everything was readily available on the bike course. That’s all fine and well, except every time I rehearsed this at home I would have to stop for bathroom breaks every hour. No worries, I thought to myself… this won’t happen during the actual race. It’ll be warmer, I’ll be sweating more, blah, blah, blah. Guess what. What happens in training happens in racing. I am not exaggerating when I say I stopped at EVERY. SINGLE. AID. STATION to use the porta-potty. Plus once more in T2 for good measure. Yes, that’s right folks, for those counting along at home that’s a solid 8 minutes of porta-potty time, according to my watch. Are you KIDDING me?! Oh well, back to the drawing board on the hydration/nutrition strategy. Long story short on the bike, I followed a bad plan pretty faithfully.
PS: I know that someone will recommend that I just ‘go’ on the bike, but this is a skill that I have not yet acquired (not for lack of trying). I don’t know if I need a new saddle, or more practice, or what. This is compounded by the fact that I do most of my training indoors on the trainer all winter/spring, so this is not something I can practice at home (ewwww).
Now that that’s out of the way, on to the highlights… the run!
T2:
Goal: 2:00
Actual: 3:57
Not much to say here, other than another porta-potty break.
Run:
Goal: 1:55:00
Actual: 1:55:12
Hey! What?! Something went right!
If you’ve read this far, I thank you! What I want to share next is the outcome of some self-experimentation. I’ve read the race pacing wiki a few times and understood the rationale behind pacing by HR. I was an early adopter of training and racing with HR, but since I started using a Stryd footpod a year ago, I will admit that I see this as a far more reliable datapoint *for me* than heart rate. In training I’ve been pacing entirely with power and RPE, sometimes not even having HR in the data screen on some runs, just looking at it afterward in the download. My race sim runs have all been paced with power, so I felt like I had a decent strategy:
start at z1 power (~190W) and ramp to TRP (210W) in the first 1.5 miles. (RPE “easy”)
Hold TRP until mile 3 (RPE “I got this all day long”)
Ramp to z2 (220-225W) for 8 miles (RPE “steady, not flashy”)
Ramp to z3 (235-240W) for final 2 miles (RPE “not a sprint, but it’s work”)
I’ve rehearsed this several times and could execute it well. A few weeks before the race in my coach thread @coach Patrick cautioned me against pacing with power and pointed out that the EN guidance is to pace with HR, per the Wiki. I reread the wiki, and we traded a couple of messages on this. The analogy I used at the time was that I saw HR as the car’s temperature gauge and power as the tachometer. I suggested that I pace by power, but not allow HR to exceed the avg. HR I saw on the bike leg in the first three miles of the run. Coach P. thought this was a reasonable approach (or maybe just wanted to see what would happen ;-).
My average HR for the entire bike leg was 146 (I had a field programmed to show this), so as I started the run I had this number in mind, along with my power zones. I also knew I was well-hydrated :-( so off I went. Now would be a good time to mention that the local community in Coeur D’Alene is VERY supportive of this race. Why is this relevant, you ask? Well, what do you think happens to HR when you step out onto the run course and are greeted by thousands of spectators cheering, ringing cowbells, high-giving you and generally being awesome? I remember looking down at my watch about 1/4 mile into the run and seeing ridiculous power (245w) and HR in the high 150’s. Whoa! Time to chill out. Except that wasn’t happening. I eased up to “easy” RPE, and power was still in the 230’s. I eased up to “ridiculously easy” RPE, and power was still in the 220’s and HR in the low 150’s. It honestly took me nearly a mile and a half to get power down to TRP, and HR never got below 150, so I had a choice to make… do I run at what feels like a fast walk, or do I run at target power (by this time I was at TRP power) even though HR is way above what I wanted it to be? I chose the latter. So, after finally settling into TRP power I chose to ignore HR for the rest of the run and pace entirely by power and RPE. The net result was that I hit all the power targets and hit my goal time. For the entire run, HR was roughly 10bpm higher for each target power zone than what I’d seen in training, and I have no idea how to explain this. I did glance at it from time to time, and it did cause me to wonder what would happen when I got to the final two miles? Would I blow up? Usually the HR that goes with z3 power is in the low 170’s, but I was already in the high 170’s at z2 power. I had no idea if I could sustain HR in the 180’s for the final two miles. It turns out I can.
I’d love to hear everyone’s thoughts on this, as I realize that it goes against the EN orthodoxy on run pacing, but for this n=1 it seems to have worked out for me, so I’m inclined to continue with this run pacing approach and see where it takes me going forward.
Overall, I’m thrilled with the race. My swim was ‘meh,’ my bike was awful, but the fact that I put together what was, for me, a very solid run at the end more than compensates for those. The bike performance is just fixing stupid mistakes, so I’m looking forward to the next race.
Thanks for reading!
-John
1
Comments
Nice report. Just a note on race day HR. My HR in a race is always 5-8 bpm higher than what I typically see in training. That number is consistent, for me, at any given power on the bike or pace while running. You had noted how high your HR was exiting the swim. Don't underestimate the cost of that high HR as it usually shows up later the race.
This was a fun report to read. CDA is a party town for the athletes!
How did you figure out TRP power? I've been collecting data with Stryd and almost ready to try it as a focused metric in a race.
PS - the picture of you and your family is my favorite. Beautiful!
I will admit that when I first started running with power a year ago I was just coming off a run injury and had no desire to do a 5k test for fear of re-aggravating my injury, so I sort of 'reverse-engineered' my TRP pace and power. I'd go do a long-ish steady run by RPE and HR and then come back and look at average power and pace and play with the vDOT calculator to see if what I was seeing passed the 'sniff test'. I then played with the run FTP number in the online calculator, which generated all my running power zones from there. Later, as I started doing some more intensity and racing, I actually had some 5k data to use in the online calculator.
If you've been collecting Stryd data, I'd start by looking at runs where you had some good steady blocks of TRP pace, and look at the corresponding average power. Alternately, you can do a 5k test, per the protocol in the run testing with power section of the wiki, and then plug the data into the run power zone calculator.
Honestly, the core of my weight loss approach was "calories in < calories out". The central element of this was (still is) logging everything I eat every single day, and weighing myself every day. I use Myfitnesspal to log what I eat, and sync it to my Fitbit app, which shows me calories burned and remaining calories to eat for the day. I manually enter workout summary data in Fitbit to get 'credit' for the calories burned from training, and shoot for 10,000 steps a day on top of training volume.
Everything else derives from this. It's amazing how quickly you make diet modification choices when you need to log that you ate 1,200 calories of pizza and beer and that put you over your budget. Suddenly, a mixed greens salad with some grilled salmon starts to sound a whole lot better
My other big learning (and the cause of one of the aforementioned crashes) was to not, under any circumstances, try to skip fueling during long runs/rides in an attempt to speed up the weight loss. I had a couple of disastrous long rides early on where I tried to get by with mostly water + a Clif bar, and ended up having a lousy ride and felt horrible afterward. So now I always follow my nutrition strategy on long rides and runs, and then manage the caloric deficit with what I'm eating outside of workouts. (Mind you, even if I drink/eat per my nutrition strategy, I'm still burning more than I'm taking in during a workout.)
Sorry for the long-winded post. It really is a pretty straightforward approach... what gets measured gets managed.
Dang, that's tough luck with the bike setup. I see you have a 165 crank length on your tri bike as opposed to 172.5 on your road bike. I have 170 on my tri bike and 172.5 on my road bike. As a person with very, very tight hips, would you suggest looking at 165mm cranks for my tri bike?
Yes, shorter cranks will definitely help open up the hip angle at the top of the pedal stroke. I went with shorter cranks on the tri bike primarily to be able to get lower on the front, as 172.5's were resulting in what felt like a very closed hip angle. Attached is a shot of me from my Retul fit appointment. This is with the 165's. You can see that the hip angle on my right leg is still pretty open, so I could drop the front end even further if I wanted to. If you have very tight hips, this could be a good solution for you, even if you're not looking to get lower in front. There is endless debate on what the 'right' crank length is. I went with 165's because it's the shortest length that Shimano makes, so I could easily find Dura-Ace cranks in that length. I could have gone with Rotor or some other brand and gone as short as 150, but for my purposes it turns out that I didn't need cranks shorter than 165 to achieve the fit goals I was looking for. If you're interested in going with shorter cranks and are not sure what size to go with, you may be able to find a fitter that has a fit bike with adjustable-length cranks so you can try out various lengths and see what the resulting hip angle is at the top of the pedal stroke. Don't select based on how they feel. They will all feel weird compared to your current 170's or 172.5's. But, whatever you choose, even something as extreme as a 150 or 155, will begin to feel normal after a couple of rides.