How to Train for Your Best Marathon PR: What We Learned Coaching Katy Walmsley to Boston

Test, don't guess. We anchor every run plan on Critical Speed, a working number built from a 400, an 800 and a 3200. It reflects both speed and endurance, and it sets every zone, every workout and her race pace. Katy tested three times before Boston.

Reading a test honestly. In January her 3200 improved, but her 400 and 800 were run on a small, icy indoor track, and slow short reps make the fit read fast. In February, on a proper surface, Critical Speed came in at 6:48. Her 3200 got faster at every test, so her sustainable speed kept rising. February became the anchor for Boston.

We don't keep an outdated number because it's convenient. One test is a snapshot.

Katy stayed a triathlete. The bike carried a large share of her aerobic work while the run built steadily, because every running mile carries an impact cost a riding mile doesn't. See how to split your hours when life is full.

Her total fitness (CTL) climbed from about 93 in early December to around 107 through February and March, then eased to about 99 on race day.

Run only numbers tell a different story. Her run fitness held near 50 from January through March while total fitness rose on the bike and in the pool. In April her run fatigue fell from about 42 to under 18, and she started Boston with her freshest legs of the build.

It's never all easy. We measured her real time in each zone from her 30 second run and ride data on 88 of 88 runs and 111 of 113 rides. Her run build was 73% easy, 19% moderate and 8% hard.

As race pace work took over, the hard share grew. Easy work still did most of the lifting: it built her durability and let her recover around a full family calendar.

A standard plan runs base, build, peak, the same for everyone, timed off the race date. We don't build that way. We pick each block by what is missing now, each sport gets its own decision, and blocks end when the gains flatten, not on a date.

Every dashed line is a moment we checked the truth: a test, a retest, a tune up race. Each one decided what came next.

The bike made its own decisions. It stayed in base and durability from Dec 9 to Feb 28 (Low Z1 rides, LT1 high torque intervals, Top Z1 intervals) while the run took the new load, then moved to build to threshold from Mar 4 to Apr 14. In the pool she kept steady sessions, 1 to 3 a week.

We read every test before using it. Her bike power setup changed more than once, and in February her 6 minute result came in below her 12 minute result, so no Critical Power could be calculated. The run was the clean story.

We wrote three Boston plans, each tied to weekly run volume. Fewer miles means less durability late in a marathon, so each plan asks for a smaller share of Critical Speed.

Plan C was written for 30 to 37 miles a week, or a matching amount of stress through multisport training. That was Katy. Her race file is labeled “Boston Marathon (C)”.

She went through halfway right on Plan C pace. Miles 2 to 4 were her quickest, between 7:31 and 7:35, because Boston opens downhill. From mile 6 to mile 15 she sat within a few seconds of 7:46, every mile.

Then the Newton hills. Mile 21, with Heartbreak Hill, was her slowest at 8:28. She held 8:13 to 8:17 for the last five full miles, and her final stretch came back to about 7:42 per mile.

Her middle miles were exactly what the plan asked for. The early miles were a little hot, and the late course took more than her durability could cover that day. Her graded pace of 7:56 against 7:54 actual shows the hills cost her. Want to learn from your own race? Here is how to analyze it.