My Cuyamaca 100k pacing strategy

After signing up for the 2019 Cuyamaca 100k this Spring, I suddenly got nervous. I live in Auburn (northern California), and up to this point all of my races had been local. I'd made a habit of getting out on each course many times in advance, studying each section.
But, Cuyamaca would be way down south, in the high desert east of San Diego, and I wouldn't be able to get out on the trails before the race. How would the elevation affect me? How would these climbs feel?

Creating the course
Using ultraPacer, I started by uploading a GPX file of the course. I plugged in each aid station, important landmarks, and trail junctions. Talking to Cuyamaca veterans, and assisted by Google Earth, I was able to estimate the type of terrain for each section (a combination of dusty singletrack and fire roads). At this point, the pacing affects of altitude, grade, and terrain are all built into the model.
The completed course (with a modification for the 2019 race) is here.

ultraPacer profile and map for the 2019 Cuyamaca 100k

Creating the plan

Now, it was time to make a plan. I went to UltraSignup, and reviewed prior results for the race, looking specifically at athletes with rankings similar to mine. That left a pretty wide range, but I had a target in mind. I chose to target 11 hours 10 minutes, which I hoped would put me in the top 10. I assumed a 15% slowdown in running speed due to fatigue through the course, and a couple of weeks out when the weather forecasts showed up I was able to assume a max heat factor of 10% at the peak of the day. That left me with the target aid station splits shown below.

ultraPacer segment information and aid station arrivals for my Cuyamaca 100k

Training equivalent?
But the big question: could I actually make this work? ultraPacer provides you with details on the modeling, and the key metric I used was the "normalized pace", which essentially takes out all of the factors of altitude, terrain, grade, heat, etc, to normalize against against sea-level smooth flat surface equivalent. My target time left me with an average pace over the race of 10:47/mile, a moving pace of 10:29/mile, and a normalized pace of 7:50/mile.
So, I set up a big training run, with the intention of running a normalized pace of a little less than race pace. I chose to do an 8:00/mile normalized pace run for 45 miles. I created a course using the Strava route builder, exported the GPX file, imported it into ultraPacer, then created a plan using a basis of an 8:00 normalized pace. While I wouldn't be able to get out on the Cuyamaca 100k course, I had, in theory, created an effort-equivalent to the first 3/4 of the course I'd be racing on.

My "equivalent" training run

I underestimated the heat a little and needed to stop and filter water a few more times than I'd planned, but overall the big training day went very well. It was time to taper down for the race with confidence.