Setting a race goal can be a daunting process, particularly for marathons. You don’t want to see yourself short and set a goal that will leave you feeling like you could have pushed harder on race day, but you also don’t want to pursue a time so speedy that you crash and burn.
So you take a recent race time (bonus points if it’s a half marathon!) and enter it into one of the many race predictors available online. And now, Runner’s World promises that their new, Better than ever!, predictor will render an even more accurate prediction to help you set your expectations right for race day.
But just how accurate and practical is this new race predictor?
The Other Contenders: McMillan and VDOT Calculators
The new Runner’s World Race Predictor, while it has shortcomings that I’ll discuss below, offers a significant improvement over other race predictors such as the McMillan and VDOT calculators. So let’s first examine these calculators and how they predict your race times.
The Jack Daniels VDOT Calculator relies upon the VDOT formula, which Daniels developed throughout years of research and statistics. Based on your VDOT score from a recent race time, the calculator produces your training paces and equivalent race times across the popular distances. Likewise, the McMillan calculator relies on statistics to produce a conversion formula for equivalent race times. This formula relies on averages, meaning that some runners ran faster marathons for a given half marathon time and some slower.
The VDOT and McMillan calculators rely on slightly different formulas and thus produce different marathon finish times for the same shorter distance race times. For 1:45:00 half marathon time, the VDOT calculator predicts a 3:38:xx marathon (8:19/mile), while the McMillan estimates a 3:40:xx finish time (8:26/mile). While the difference isn’t significant, the two minute difference can make or break a BQ time for some runners. Even for the average marathoner, those minutes matter.
Both the VDOT and McMillan calculators often overestimate one’s ability in the marathon. Why? These calculators assume that you are training around 70 or more miles per week and use statistics from elite runners to project the change in pace from one distance to another. So for the average marathoner, the statistics were slightly skewed. Most runners run 30-60 miles per week in marathon training, which does impact efficiency, aerobic capacity, and fatigue resistance.
Knowing what type of runner (speed vs. endurance) is also important for choosing the right race predictor calculator as well. For the same 1:45:00 half marathon time, the McMillan calculator predicts a 7:18/mile average pace for the 5K, while the VDOT suggests a 7:22/mile average pace. Thus with the faster marathon time and slower 5K time, Daniel’s formula favors long distance runners, while the McMillan is a better bet for those who excel at running short and fast.
For the average runner, then, it’s helpful to use a variety of race distances when attempting to predict your marathon finish time. If your 5K PR is the best out of all of your races, the McMillan calculator would be a more reliable choice for you. For marathoners and half marathons, the Jack Daniels calculator may be more realistic, if you are running the mileage that it accounts for.
While I use them for training paces all of the time and find them fairly accurate for predicting shorter races, both these calculators always predicted marathon times a good 5-10 minutes faster than I knew myself capable of running.
How Does the New Runner’s World Calculator Stack Up?
The biggest pros for the new calculator is that, instead of relying on the results of highly trained, professional, super human runners, this new race predictor offers a more realistic prediction for the non-elite runner. The calculator is based upon a model presented in a recent statistical study of recreational runners, their training, and their race times. The study, “An Empirical Study of Race Times in Recreational Endurance Runners,” surveyed statistical evidence from recreational runners, rather than elite runners. Beyond this, however, it fails to address factors that the other calculators ignore as well.
Mileage quantity vs. quality.
A major con this new tool has is that it only allows you to input your total weekly mileage, and it’s not just the quantity of miles that matters; the quality of those miles matters as well. One could run 50 miles per week of all easy miles, or one could run 50 miles a week including a tempo run and an interval workout. Although there are always exceptions to the rule, the runner including threshold training and speed work will improve their race times more than someone running the same mileage but without variation of pace. The study upon which this race predictor was based did affirm the value of threshold training and interval training on race times, yet fails to take that into account.
The new Runner’s World Race Predictor also fails to account for diminishing returns. As Tim Noakes argues in his Lore of Running, the average recreational runner (read: non-elite) will start seeing diminishing returns once they increase their mileage to over 70-75 miles per week. Past that point, and the miles are less effective in aerobic development and more likely to cause injury.
The calculator, however, does not reflect this concept. The predicted race time keeps increasing as I input 100, 120, and 150 miles. Of course, that’s because this calculator is using an algorithm, and an algorithm can never fully account for the factors of real life. If someone like myself, a 1:38 half marathoner who runs 30-50 miles per week depending on the season, increased my mileage up to 100 miles per week, I wouldn’t see a 3:11 marathon as a result like the calculator suggests. Instead, overtraining, injury, and mental burnout would happen.
Don’t forget the race course! Another problem that the new calculator does not address is the race course itself or the courses of the races you input to predict your goal time. Most calculators assume you are racing on a similar course profile with the same weather, which means if you are using a flat half marathon in perfect temperatures to predict your time at Boston in the heat, you will overestimate your finish time.
A race predictor calculator can be valuable in setting realistic goals, but it can never fully take into account everything that influences race time: the fine balance of training vs. overtraining, quality vs. quantity, weather, and mental toughness.
For an average marathoner or someone aiming to BQ, the VDOT and McMillan race predictors may be overly optimistic. While still not accounting for a variety of factors, a calculator that accounts for average weekly mileage, such as the new Runner’s World Calculator, offers a more realistic prediction for us non-elite marathoners, but still falls short of being a sure-bet.
Which calculators have you used? How accurate was the prediction?