“Net downhill” is somewhat synonymous with “fast”. In road running, there are regulations in place that limit how much drop a point-to-point course can have before it becomes invalid for record purposes. Boston, a generally fast marathon—mostly due to its qualification purposes—is aided in its relative speed due to a –447’ net elevation. It’s too downhill for records purposes. When Geoffrey Mutai ran 2:03:02 in 2011 it was the fastest marathon at the time, but not a world record.
Trail running is a different story all together. While trails vary widely, they generally favor the uphill. Black Canyon 100K is not like that.
I underestimated it. Don’t make that same mistake.
The Black Canyon 100K—and its 60K variant which took place the following day— start at the northern end of the Black Canyon Trail and head south toward Phoenix nearly to the end (the trail continues on a few miles longer). It offers 5,190’ of gain with 7,050’ of loss—or that famous phrase, “net downhill”. Of course it’s accurate, but it’s deceptive.
As Jim Walmsley said to Brian Whitfield, who left the Black Canyon City aid station in the lead but would return to it shortly after to drop from the race, “that course doesn’t run like it looks on paper.”
This is a long post that gets deep into the race and what happened. I wrote it so I don’t make the same mistake next time, and maybe it’ll help you in the same way.
Definition of undulate
to form or move in waves : FLUCTUATE
to rise and fall in volume, pitch, or cadence
to present a wavy appearance
to cause to move in a wavy, sinuous, or flowing manner
Take a look at the CalTopo mapping data for the course, especially the elevation profile. The default sampling interval for this map displays data points every 1,092’, and it looks like the following. It’s what you see on the race website.
It’s net downhill and has those two climbs at the end that you need to save yourself for—you’ve maybe read or heard that a million times. A lot of the start, especially through mile 15, looks down and somewhat smooth. Even that part from about 12–15 looks fairly flat. It’s not.
Turn the sampling interval up, which uses more data points to draw the graph. The smallest interval this particular map supports is 50’ samples.
Looks a little different, right? Undulating.
If you download the CSVs of those two profiles, you can see that the top graph is made from 300 data points, 116 of which are uphill from one point to the next. In 1,092’ segments, you spend about 60% of the time going down (relative to the last 1,092’ segment), and 40% going up. Looking at that top image, yeah, that makes sense. However, if you’ve run this race, you know that a lot happens in about fifth of a mile.
Drawing that graph from 6,553 data points tells another story. Percentage-wise the course looks a little more downhill in 50’ segments, at about 70% down and 30% up, but I don’t think the percentage really tells the best story. There are 1,879 times where one 50’ segment is more uphill than the last. That’s the fuzz you see in the second image.
Because the total gain is relatively small compared to a lot of other trail 100Ks at just over 5,000’, that’s a lot of little ups! This destroyed me.
I ended up dropping at the Table Mesa aid station (mile 50.9) after a painful slog from Black Canyon City (mile 37.4).
Where I live, near the foothills of the Rocky Mountains, there aren’t a lot of trails that fit the 83 feet/mile average gain of Black Canyon’s 5,190’ spread across 62.2 miles. A lot of my training takes place on profiles at or above 225 feet/mile, where a climb might gain 1,000’ in two miles—a lot more sustained up’s and down’s. To get ready for what I initially interpreted as a flatter and faster race, I found a stretch of trail that put me closer to 100 feet/mile and made that my new home.
In a sense, this was still the right move, and I’ll probably spend a lot of time there training for my next shot at this race. I did some VO2max and threshold workouts early on in training, and closer to the race I spent a lot of time doing longer efforts including tempo runs on this ~100 ft/mi trail. As you can probably guess from the images above and my result, the flatter little city park trail missed one key element: being undulating.
Because of the relative lack of climbing, I didn’t spend time focusing it. I was averaging a bit more than the course profile in weekly mileage anyway, and the total of those final climbs wasn’t outside of what I would get by going to my usual spots once a week or so. I wouldn’t focus on climbing for next time either, but I do need to focus on how to withstand the undulation. For me, that probably means more work on my hips and glutes. I can run uphill all day, but the pounding is a lot of what killed me.
I am a very salty sweater, and with training for this race mostly taking place in cold Colorado, I was a bit worried how that might work out. I forced myself to drink a bit more water on long runs during training even though I didn’t “feel” like I was sweating as much due to the cold. I’m normally pretty good with water so it wasn’t something I focused on much, and thankfully even though it’d be warmer in Phoenix than in Colorado, being in the low 60’s isn’t too bad.
I like GU products so I stuck with them. GU Chews in the Salted Lime flavor taste great and work well for me, and I trained a bit with their new Liquid Energy gels in order to provide some variation if on race day something wasn’t working. I do one serving of Chews (4 of them) every 30 minutes, or one Liquid Energy. I also like the Roctane drink and usually have a bottle of it during warmer times, but I drink it enough that I was comfortable just taking it as-needed from aid stations.
I planned on wearing a Nathan VaporKrar waistpack and using two Nathan soft handhelds, which worked well for me at previous races, and I did well at Way Too Cool 50K with that setup so I stuck with what I knew. Frankly I think this was a mistake, as this didn’t allow me to adjust for the weather or adjust with my goals. I really should have worn one of my vests so I could pack a windbreaker (because I knew wind could be a factor) and generally be more prepared to be out there later.
I was way off. Or maybe I was right on and just trained extremely poorly for it. You could probably go either way, but for this race I’m taking the former route. If the same thing happens next time it’s the latter.
My A goal was to finish in under 10 hours. I was running my long runs on what I thought was a comparable profile at an average pace around 8:50–9:00 usually for four hours, with several back-to-back weekends, and on most of them I was holding back.
That was partially on purpose as this was my first race back from injury0, so I didn’t want to re-injure myself. And in fact, I spent what should have been my second to last big weekend wearing a plastic boot because I thought I broke my foot again. X-Rays and an exam by my ortho doc showed nothing was wrong, so I took a few easy days and got right back into it. The taper looks a bit weird, but it was fine enough. I had a solid block of consistency heading into that so one down week wouldn’t throw too much off.
My B goal was to finish in under 11 hours. Beyond that, I just wanted to finish.
Seeing someone I know finish in 10:01 makes me realize that was probably a stretch. I think under 11 hours was probably more reasonable for a good day, with 12 a safer bet. A guy I ran a lot of the race and who made it to 50.9 just before me ended up finishing in 12:28. He also ran 5 extra miles because he missed the turn at Black Canyon City.
One thing that really killed my chances of the backup goal of “just finish” is that I was so bought into the A and B goals that I wasn’t even prepared for anything else. I ran this in a thin singlet with handhelds and a waistpack like I was some super fast guy. Starting at 8:15a and finishing in say 14 hours—which would have been the case if I continued on from where I dropped—would have had me running 4 hours after sunset. I did have a headlamp at Table Mesa, but given how I slowed down and how cold and windy it was, it was going to be a harsh time even if my legs worked.
I was basically setup to finish as the sun was setting, or not at all. This was dumb.
Rather than a big mass start, the race was broken up into several waves that started every 15 minutes. I ended up in Wave 6, which started at 8:15a. Aravaipa ran a very safe and well organized event, so getting in and started was super easy. I took off at 8:15a on the dot with about 30 other runners spaced out in the corrals, and then we were off onto the trail to spread out even further.
This stretch does have some legitimately smooth parts! You go around the track, down the street, on some nice packed dirt for a while, and then eventually hit trails. I hit this aid station about 12 minutes up on the sub-10 goal.
In my head, I justified the pace as I barely felt that section. It was smooth, I felt right at home in this pace, and I was ready to spend all day there.
Doing this over again, I’d probably back off here but only slightly. I ran this section at 8:01/mi, but I think if I do this closer to 8:30/mi I’m alright. No need to go out too hard as it’s a long day ahead, but it’s a comparitively easy section so I think it’s reasonable to move through here a little faster.
I didn’t stop at Antelope Mesa. It’s only 5.2 to the next aid from there, and I hit the aid station in almost exactly an hour so it coincided with taking my own GUs that I already had on me.
This is where I started to back off a bit. Coming out of Hidden Treasure is a relatively flat spot that I probably stayed on the gas through ~8:00/mi, which then leads to a good stretch of undulating singletrack. It’s also—to use a word from my wife—serpentine.
This section snakes around a lot to where you’re not running straight for very long. Lots of tight curves amongst these little up and down dips. I ran this section at 8:56/mi, which I’d dial back maybe slightly overall. The first half of it is decently runnable, but I should have slowed down a bit through those dips and swoops.
Heading into Bumble Bee is a nice smooth downhill on a packed dirt road, so you can ease back on it a little bit there, but don’t try to make up any time there. It’s early.
I fumbled my watch here and I see a 27 second lap (I usually hit
lap when entering
lap when exiting), but I got both bottles filled up here. This was probably
a ~60 second stop.
BONUS: My wife was watching the live stream and saw me come into the aid station. She said I looked strong!
This was a tough section for me. Coming out of Bumble Bee brings you to the first sustained climb. It also brought you to the wind. It also really reinforced undulating.
That climb coming out of aid gains maybe 350–400’, and with my hips starting to feel the wear from all of the undulation, I hiked this one. That just lead to more undulating. Once you’re up top, there’s a big valley off to the right and you’re carving the inside of this mountain, just going up and down and up and down, in and out of slots, and then up and down.
I took this stretch around 10:36/mi, for a 47:32 trip to Gloriana. Once there, I stopped for 2:06 to refill bottles and grab a cup of GU Roctane.
At this point my hips and glutes were really feeling it. I’ve had some TFL/glute medius imbalance issues on my right side for a few years now, which leads to a medial knee collapse, but my knees felt comparitively fine. My hips were starting to get trashed.
This stretch sucked. It starts off with some flat-ish parts, and then there’s a long downhill stretch. A lot of that downhill is on some harsh rock formation. Not wanting to burn my quads out and also not having a great hip situation, I took this a bit easy. It was frustrating to not be taking advantage of this part, but I also didn’t feel really confident on that rock. This rock also ripped off the back of my left shoe. Damn this rock.
Then you hit the bottom, tool around in circles for a while, then come back up toward Soap Creek.
In order to hit my sub-10 hour goal I needed to hit Soap Creek in 4:57 elapsed. I sort of knew that was out the window based on the last ten miles, but I was still close. I don’t know what distance my watch thought I was at exactly when I hit aid, but it claims I made a 50K PR at 5:05, so somewhere around there.
I stayed here for 4:04, refilling bottles, grabbing Roctane, eating some watermelon, and just taking a second to breathe and re-frame myself.
I came here the day before the race to run with Bryden and we went about 2.5 miles out and then back, just to get some action on the course and stretch out after a 14 hour drive.
That first 2.5 miles went fairly fine. It’s not too undulating, but does have some sections of softball sized rock that I just wanted to make sure I didn’t break my ankles on. Eventually you get to some smooth rolling dirt road, though it’s pretty aggressively rolling. It’s not like a hill in your neighborhood, but at least it was smooth.
Somewhere in here it became hard to run. My hips were destroyed and it was hard to do much more than hike or shuffle, which started to get demoralizing. Everything else, though, felt pretty good! Digestion was good, energy was good, and my overall mood and confidence were changing but intact. I just had to keep on moving and I’d be alright.
I eventually hit the aid station 1:22 later, slowing down to a 12:58/mi pace through this stretch. It got pretty dark heading into BCC and I briefly thought about dropping.
I ended up staying in Black Canyon City for 9 minutes (!), where I had my first drop bag. I restocked on my GU Chews, put on more Squirrel’s Nut Butter, and read a bunch of notes from my wife. I cried a little bit because she’s so amazing, and also because one of the notes was a list of names we would have to give our (not yet conceived) baby if I drop anywhere after this point. They were all bad names that I don’t want our baby to have, so I had to keep going.
I also borrowed a volunteer’s phone to text her to bring more clothes to Table Mesa, where I’d get to see her. Knowing where I was at pace-wise, I was going to be going on much longer than I initially planned for, and this tiny tank top was not going to cut it as it got colder and windier. Thank you to this person! Normally people just prepare for this, but I was not that person. See the Goal Setting section above.
I had already accepted the fact that the A and probably B goals were gone, was firmly planted in embracing the suck of the upcoming stretch, and was addressing it by focusing on keeping what was working well, working well. Let’s go, on to Cottonwood.
I felt better coming out of the aid station, but still not really in a runnable state. Whatever, let’s just power through and get there, right?
[Narrator’s voice] That is not what happened.
Well, I made it beyond this, so I guess it sort of happened, but this was a brutal stretch. I left BCC in the vicinity of a woman I ran slightly ahead of since Soap Creek at mile 31.2. I never saw her again. Shortly after two people passed me around probably mile 39, I would not see another person behind me—as this section serpentines around this canyon—until probably mile 45. I spent those 6 miles slowly watching these two guys pull ahead, disappear, reappear further away, and ultimately never see them again.
All the while it’s cloudy and the sun is going down, and no one is anywhere behind me. In reality they were probably there, just disappearing around the curves just like the guys in front of me were. It was this, coupled with my diminishing inability to run more than a few steps, that caused me to start panicking a bit.
It would take me 2:10 to get the 8.8 miles to the Cottonwood aid station, and during that time I made up my mind that I had to be done. All my “accept, embrace, address” stuff was not working. I started to worry that if I keep going this slow the rains that I knew were coming and the cold winds that were already blowing as the sun was practically down (rather, behind what appeared to be two hours worth of clouds) were going to leave me cold and wet and out in the middle of nowhere unable to help myself. Unrealistic? Yeah, probably, but that’s what thinking you’re the last person out there will do to you.
The other thing I started thinking was that when I get to Cottonwood, can I even drop there? I’m normally good with spatial awareness and having a reference of where I’m at, but I was totally clueless. I hadn’t seen an actual road in two hours. Nothing ahead looks like it’s heading towards a road, and those jeeps I could hear off in the distance are the kind you drive up rock walls. They’re not passenger vehicles; they’re out here because they’re built to be. They’re built for the undulation.
Thanks to my nifty card, I knew that if I couldn’t drop at Cottonwood, it’s only 4.7 miles to Table Mesa, where my wife was waiting to send me off to the finish (probably for two hours, since I was way behind). Plus, if I could drop there I wouldn’t just be chauffered to wherever I’d like to go (Table Mesa), and where my wife was (Table Mesa) does not have good cell reception anyway. We were there the day before so she’d know where to go1. I had to get myself to Table Mesa.
Probably a mile from Cottonwood, two guys pass me. I hit the aid station after dropping to a 14:38/mi pace throughout the nearly 9 mile stretch, and then two people come in a minute behind me. Where the hell were these people for two hours? Hiding?
I grabbed some water and ginger ale for a 3:52 stop with the fine folks at Cottonwood and headed on my way. I didn’t even bother asking about dropping. It was going to be a hassle and I just had to propel myself. I was still feeling good in all other ways except my legs, so maybe just getting there would do the trick. Plus, that’d get me over 50 miles, so at least I’d have that.
This section had some relatively flat parts that I could shuffle-jog without a lot of pain, and with the sun setting I tried to do that as much as I could. Because I had made peace with this being the final stretch, it was actually kind of fun. Slow, but fun.
At probably mile 48 I saw a familiar orange shirt behind me, and about a minute later he caught up. We started in the same wave and ran a lot of the first half together, and I moved aside to wave him on since I was just walking at that point. He stopped briefly to catch up and see how I was doing since it had been a while, and he’s the one who ran an extra ~5 miles due to missing the Black Canyon City out-and-back. We chatted briefly about how our days were going, then hiked a bit together up a little climb.
It turns out we live near each other and he’s someone I noticed in my wave assignment that I thought, “heh, I wonder if I’ll end up running with that guy.” It wasn’t until the very end that we actually talked to each other, but I let him get on with it and run ahead as I was still mostly walking. I’d see him leaving Table Mesa and told him to go on and get that buckle, which he did in 12:28, roughly 2:30 after leaving the aid station.
After that I was fairly close and the sun was setting, so I picked up the pace a little bit as I heard the commotion of the aid station. As I entered, who was standing there all alone ready to help me out? My wife. It wasn’t even like she was standing with a group of people cheering on runners as they came in, she was just standing out in the middle of nowhere, probably for a very long time.
I gave her a hug and just broke down. That was it. She was ready to get me going for the next segment but I couldn’t. Thankfully she brought some extra stuff based on the text I sent from Black Canyon City, but I was still not prepared enough to do it and I worried that I’d be out finishing those final 12 miles stumbling around in the dark for 5 hours. Should I have prepared better in order to get that done? Absolutely. But I failed even the basic preparation to make sure I could do last possible thing to get myself across the line. It really sucks.
From there we sat in the car for a bit warming up and telling her about the day, then went to the finish line to grab my bag. Then we ate some calzones, tried to get some rest, then drove another 14 hours home.
This seems pretty obvious, but I also thought I had prepared well enough that I was doing the right thing. I know what paces I should fall back to in certain sections, and I think my overall time goals are not too far off of what I’m capable of. I might go slightly more conservative but still want it to be a true challenge.
“Just finish” will be the backup goal, and putting together the necessary gear to support that is going to be a significantly higher priority next time. It was basically an after-thought this time, which means it really wasn’t a goal if I wasn’t really working towards it and preparing for it.
In supporting the goals, having what I need on the course is something that will need to improve. Even some of the actually fast guys who just do handhelds actually started this race with a vest and dropped it at some point.
50K is probably my new limit for getting away with no vest.
The way to get better and faster at running is to run more, not lift weights, but the sustainability problem I had was less with being able to hold the pace but more with the impact that the terrain had on me. Cardiovascularly I felt really good throughout this, so I’m doing something right there. Slowing down would help slightly, but ultimately that makes for more steps and more time which makes for a lot more impact to be taken.
Building up the posterior chain, especially with the known imbalance that I have, is going to be key to building a body that can keep up with the paces that the rest of my body can put down. Hip thrusts, good mornings, RDLs, squats, etc. are on the menu. Banded exercises like clamshells to build up the adductors are something I’ve done in the past that need to come back. Probably some ankle strengthening stuff as well.
Hard days should be hard, so I’ll start mixing some of this in after workouts.
Matt Fitzgerald’s “The Comeback Quotient” is a pretty good read about being an “ultrarealist” and how to “accept, embrace, and address” things that come up in endurance events. After having too many rough runs during my post-injury comeback where I was just a mental wreck, I had to do something about it. I’m almost done with it, but it’s probably worth a revisit before races as it’s a fairly quick read and a good boost.
Addie Bracy, who got fifth at this race, has a book coming out in a few months called “Mental Training for Ultrarunning” that I’m looking forward to reading.
I finished Way Too Cool 50K in March 2020 just seconds under my 4:30 goal, and apparently did so with a broken foot. I felt it a little bit while running but thought it was just a minor tweak and nothing significant. A few hours after the race it was extremely painful, and ended up being a stress fracture of my second metatarsal. I spent probably two months in a plastic aircast and then about a month of regular shoes but no running, then eased back into training around July.
I mapped where to go and then drove there, and guess who was wrong? Me. She was like “do you think it’s that big tent over there?” (it was) and I said something like “this is what the website says” as I drove to this empty dirt road having none of the features I described.