Build the finish before chasing the pace.

9:09 per mile, or 5:41 per kilometre, held for the whole 26.2 miles. That means passing halfway in 2:00:00. If you cross the half in 1:55 you have banked five minutes you will hand back with interest after mile 20 — the marathon is the one distance where going out fast is almost never recoverable.

Even, or one to two percent slower in the first half than the second. Analysis of large marathon fields consistently shows negative and even splits correlate with personal bests, while positive splits of more than three percent correlate with the classic late-race collapse. Practically: run the first three miles feeling like you are holding back.

About 10 weeks for a beginner, peaking near 17 miles a week over three running days; roughly 7 weeks for an intermediate runner peaking near 25 miles a week; and 5 weeks for an advanced runner already carrying 36 miles a week. All three assume you can already run continuously for about 20 minutes on day one.

Twelve weeks from a beginner base of about 9 miles a week, ten weeks from an intermediate base of about 19, and eight weeks from an advanced base of about 31. The plan compounds weekly volume — 10% a week for beginners, 8% for intermediate, 6% for advanced — through the build phase and then tapers.

Do not increase weekly running volume by more than about 10% over the previous week. It is a ceiling for the tissue that adapts slowest — tendon, bone and fascia — which lags cardiovascular fitness by weeks. It is not a schedule: plenty of weeks should be flat, and every third or fourth should drop 20 to 30%.

Poorly at both ends. At 8 miles a week it permits less than a mile of increase, which is needlessly slow. At 70 miles a week it permits a 7-mile jump in a single week, which is far more than most runners absorb. Below about 15 miles a week, add a mile or two; above about 40, cap the increase at 3 to 5 miles regardless of percentage.

Good if you have done the long runs, optimistic if you have not. Cameron’s formula uses an exponent that rises with your 10K time, so it predicts a larger slowdown for slower runners than fixed-exponent methods. Its blind spot is the same as every equivalence formula: it measures your speed, not your endurance, and the half exposes endurance.

Riegel raises the distance ratio to a fixed 1.06 for everyone. Cameron makes the exponent 1.07 + 0.0065 × ln(t), so a 35-minute 10K runner gets an exponent near 1.093 while a 70-minute runner gets about 1.098. The gap is small in the exponent but grows over a doubled distance, and it tracks real results better across a wide range of abilities.

Roughly 21 easy days by the one-day-per-two-kilometres rule, and about 26 by the older one-day-per-mile version. Either way it is around three weeks before hard training resumes. Easy means walking, gentle jogging or cross-training — movement helps, intensity does not.

Not inside the recovery window without paying for it. Blood markers of muscle damage stay elevated for one to two weeks after a marathon even when the soreness has gone, and immune function is measurably suppressed in the days immediately after. Book the next hard race at least four weeks out, and a short one before a long one.

Naismith’s rule, converted to US units, says every 528 feet of gain costs you about one flat mile of effort. A 13-mile race with 2,640 feet of climb therefore runs like an 18-mile flat race. On very steep or technical ground the real cost is higher, because the rule was written for hill walking and credits nothing for a difficult descent.

Because it sets the grade, and the grade decides whether you run or hike. Under about 100 feet of climb per mile a course is essentially rolling and runnable; over about 300 feet per mile most runners are power-hiking the ups, which changes your pacing, your fuelling and often your footwear.