Almost every first marathon that goes wrong goes wrong in the same 40-minute window: somewhere between mile 20 and mile 23, the pace that felt easy for the first two hours turns into a shuffle. The runner slows by 30, then 45, then 90 seconds a mile. The clock they were sneaking a look at every checkpoint stops mattering. The goal quietly becomes finishing.
The frustrating part is that the mistake was almost never made in that window. It was made in the first 5K, when the runner banked 90 seconds on the goal pace because they felt fresh and were passing people, or when they picked a goal pace that their training data was never going to support in the first place. Marathon pacing is a data question dressed up as a willpower question, and getting the data right is what separates the people who negative-split their first marathon from the people who spend the last 10K bargaining with a hamstring.
Photo by RUN 4 FFWPU on Pexels
Why the wall lands where it does
The marathon distance is 26.219 miles for a reason unrelated to physiology. The human aerobic system is not designed around that number. Most runners have enough stored glycogen for roughly 90 to 120 minutes of continuous work at marathon effort. After that, the body has to shift to a mix that leans harder on fat, and fat metabolism produces energy more slowly. The gap between the pace glycogen can sustain and the pace fat can sustain is where the wall lives.
That gap is not a fixed constant. Runners who log 50 to 70 miles a week for six months train their bodies to burn fat efficiently at higher intensities, which pushes the wall from mile 18 out to mile 24 or later, or removes it entirely. Runners who log 25 to 35 miles a week and hope enthusiasm will cover the difference reliably find the wall right around mile 20, no matter how much they wanted a 3:45.
The point is not that lower-mileage marathoners cannot finish. They finish all the time. The point is that a lower-mileage runner picking a goal pace as though they trained like a 60-mile-a-week runner has produced a plan that expires exactly when the race stops being polite.
Riegel's formula, honestly
The most cited pace projection in the sport is the Riegel formula, which predicts that your time at a longer distance equals your time at a shorter distance multiplied by the ratio of distances raised to the power 1.06. In practice, that means a 1:45 half marathon projects to a 3:38 marathon, and a 22-minute 5K projects to a 3:32 marathon.
The number is directionally correct for aerobically well-trained runners. It is optimistic to the point of being dangerous for undertrained ones. The formula assumes the runner has enough endurance base to sustain a similar effort level at the longer distance. A first-time marathoner who has never run further than 18 miles is not that runner. Their aerobic system will hold up for the first 90 minutes at a pace consistent with a fast half, and then the wheels will begin coming off as the fuel system runs out ahead of the finish line.
The honest adjustment most coaches make for a first marathon is to add 8 to 15 minutes to the Riegel prediction, based on the runner's weekly mileage and how many 20-mile long runs they got in. That is not pessimism. That is the difference between a plan built for your fitness and a plan built for the version of you that trained twice as much.
Reading your training data instead of your feelings
The training data that predicts race day is more specific than "how did the long runs feel." Three signals matter.
First, your long run's ending pace. Not the average pace, and especially not the pace of the first six miles. The pace of the last four miles of a 20-mile long run, when your legs are honest, tells you a great deal about what marathon pace can look like. A common calibration is that your marathon pace should be roughly 30 to 60 seconds faster than the pace you ended your best long run at, if that long run finished at true easy effort and did not turn into a survival exercise.
Second, your marathon-pace tempo runs. A run in the last eight weeks of training that includes 10 to 14 miles at planned goal pace, followed by an easy cooldown, is the single most predictive piece of data you have. If that run felt sustainable and your heart rate did not drift more than about 5 to 8 beats per minute from mile 4 to mile 14, the goal pace is realistic. If it felt like a race by mile 10 or the heart rate ran away, the goal pace is fantasy.
Third, your race-week resting heart rate compared to your normal baseline. Elevated numbers during taper are a common overreaction to reduced training load, but sustained elevation is a signal to trust the last eight weeks of data and not to sharpen the goal pace based on how fresh you feel.
The Pace and Race Time Calculator is useful here because it lets you go in both directions. You can plug in a realistic time for your 20-mile long run and see what pace that projects, or plug in a goal marathon time and see what per-mile pace it demands. Running both directions and comparing what your training supports against what your ambition wants is usually the moment the plan gets built.
Photo by ahmed akeri on Pexels
What a mile-by-mile split plan actually looks like
Once you have an honest goal pace, the next step is a mile split table. This is where most first marathoners undo themselves in the first 5K.
A workable plan looks something like this for a 4:00 goal, which averages roughly 9:09 per mile:
- Miles 1 to 3: 9:20 to 9:25 per mile. Slower than goal, no exceptions. You are running in a crowd, warming up your legs, and burning as little discretionary energy as possible.
- Miles 4 to 16: 9:05 to 9:12 per mile. Goal pace, held steady. If you feel great here, ignore it. If you feel bad here, ignore it. Hold the pace.
- Miles 17 to 20: 9:05 to 9:15 per mile. Same range, but effort will climb. This is where discipline matters more than legs.
- Miles 21 to 26.2: whatever you have. If the first 20 were honest, this is where a small negative split is possible. If the first 20 were greedy, this is where you meet the wall.
The critical insight buried in that table is that the first three miles are not banked time. They are protection. Every 5 seconds per mile you run faster than plan in the first 5K costs you roughly 30 to 60 seconds per mile in the last 5 miles. That trade is never worth it. Nobody has ever finished a marathon wishing they had gone out faster.
The negative split, and why it is so hard
A negative split means the second half of the race is faster than the first half. It is the pattern that produces every marathon PR ever set at the elite level. It is also the pattern that fewer than one in five recreational marathoners produce.
The reason is not physiological. It is psychological. In the first 10 miles of a marathon, your legs are fresh, the crowd is loud, and your goal pace feels effortless. The temptation to run at a pace that feels natural instead of the pace you planned is enormous. Runners who negative-split are not more fit. They are more disciplined about running a pace that feels too easy for the first 90 minutes.
A useful rule for a first marathon is to aim for a small negative split, on the order of 30 to 90 seconds. Do not aim for a large one. A dramatic negative split usually means you left too much time in the first half and the finish was faster than sustainable rather than the second half being genuinely stronger. A 30-second negative split, achieved by starting 15 seconds a mile slower than goal and running the second half at goal, is a mature plan for a first marathon.
The three sessions in the last eight weeks that tell you the truth
Between weeks 12 and 4 of a training block, three specific workouts predict race day better than everything else combined.
The first is a 20-mile long run with the last 6 miles at goal marathon pace. If you can finish that workout without your form falling apart, your goal pace is defensible. If the last two miles turn into a jog, drop the goal pace by 10 to 20 seconds and re-run the math on the mile-by-mile plan.
The second is a half marathon tune-up race, ideally in weeks 4 to 6 before your marathon. Run it at 15 to 20 seconds per mile slower than your open half marathon PR, not full effort. What you are measuring is not fitness. It is how your body handles race-day logistics: hydration, pace management under crowd energy, and recovery over the following week. A half marathon at goal marathon pace plus 60 seconds per mile is another common variant.
The third is a medium-long run of 14 to 16 miles at goal marathon pace, in weeks 5 to 8 out. This is the specificity workout. If the pace feels manageable for that distance in training, it is the closest simulation of race pace you will have. If it does not, the goal is aspirational.
Higdon, Pfitzinger, and other well-known training plan authors converge on the same idea: it is the marathon-specific work in the final eight weeks that determines outcomes, and it is impossible to fake in the taper. Runner's World publishes a rotating set of these plans, but the specific author matters less than sticking to whatever plan you picked. Underlying every credible plan is the concept of VO2 max and the aerobic threshold, which is what marathon-pace tempo work is training in the first place.
Race week is not the time to be brave
The single most common taper-week mistake is recalculating your goal pace upward because you feel fresh. Freshness is a taper artifact, not new fitness. You will feel that way at the starting line and for approximately 12 miles into the race, at which point actual fitness reasserts itself.
Trust the last eight weeks of data. Trust the tune-up race. Trust the 20-mile long run. Adjust your goal only for weather. Heat and humidity are the two variables worth pacing around: every 5 degrees Fahrenheit above 55 typically costs 4 to 10 seconds per mile at marathon pace, and high humidity compounds it. A cooler-than-forecast race day is not permission to run faster. A hot day is permission to slow the plan by 10 to 20 seconds per mile from the start.
The EvvyTools directory of running and fitness calculators includes tools for these adjustments, and the full tools directory covers pace, splits, and race prediction in one place. The related pieces in the EvvyTools blog go deeper on individual workouts and taper strategies for readers who want the next layer of detail.
Photo by RUN 4 FFWPU on Pexels
What actually happens in miles 20 to 26
A well-paced marathon in the last 10K is not comfortable. It is a controlled exercise in effort management. Your form gets sloppier. Your cadence drops a beat or two per minute. Your breathing gets louder. All of this is normal and none of it means the plan failed.
A poorly paced marathon in the last 10K looks different. Pace collapses by more than 45 seconds per mile within one or two miles. Cramping starts in the calves or hamstrings. Aid stations become long. Runners who were fresh at mile 15 walk through mile 22.
The difference between those two outcomes was almost entirely determined by pace decisions in miles 1 to 3. That is the humbling and useful truth about marathon pacing. You do not run a smart marathon by being tough at mile 22. You run a smart marathon by being patient at mile 2, running the pace on the plan instead of the pace in the crowd, and letting the last 6 miles come to you.
The runners who do this consistently produce times that surprise them mildly to the upside. The runners who do not consistently produce times that surprise them by a lot to the downside. Nothing else in the race is nearly as high-leverage as those first 30 minutes.
One habit worth carrying forward
The single habit that separates people who improve at the marathon over multiple attempts from people who plateau is post-race pace analysis. After each marathon, plot your actual mile splits against the plan. Ask where the pattern broke, not whether it broke. Did the first 5K run too fast? Did miles 16 to 20 drift? Did the last 5K collapse in a way that suggests fueling rather than pacing?
The answers do not always require a new training plan. Sometimes they require better patience in the first 30 minutes, or better fueling in the middle third, or a more conservative goal pace built on the training you actually did rather than the training you wish you had done. A pace calculator is a math tool. The interesting question it makes visible is the honest one: what does the training I actually completed support, and am I willing to race that plan?
The runners who answer that question honestly, and hold the resulting pace for the first three miles when it feels ridiculously slow, are the ones who cross the line looking strong. The runners who do not are the ones who spend the last 5K learning why the wall is where it is.
