I Scored 200 Marathons for BQ-Friendliness. Here's the Method.

Why "flat" is a lazy proxy

Every race-finder slaps a "BQ hotbed" badge on flat courses and calls it a day. We wanted to know whether the label actually means anything, so we did the unglamorous thing: we scored all 200 marathons in our catalog on the factors that genuinely predict a Boston qualifier, using the same data for every race, and looked at what fell out.

The instinct is to rank races by elevation gain and stop there. But flat is only one input, and on its own it's misleading for two reasons. First, "flat" hides net downhill: a point-to-point that loses 350 feet is meaningfully faster than a true-flat loop, but both show up as "flat" on a finder. Those are different courses that suit different runners.

Second, a high published BQ rate is partly a story about who shows up, not how fast the course is. A small race that markets itself to qualifier-chasers will post a gaudy BQ percentage because the field is self-selected and fast, not because the road is special. Strip out the field effect and some "BQ hotbeds" look a lot more ordinary.

So elevation is necessary but nowhere near sufficient. A real score has to combine terrain with the two things that actually decide times on the day, weather and field, and it has to treat net downhill differently from true flat.

The four factors we scored

Every race in the catalog gets the same four inputs, pulled from the same structured data:

• 01Course terrain. Not just total gain, but the shape of it: net elevation change, where the climbing falls, and a cumulative-gain penalty for courses with heavy rolling that the net figure hides. A course with 300 feet of gain stacked into one mid-race stretch races very differently from 300 feet sprinkled evenly. Net-downhill is separated from true-flat, not lumped as "flat."
• 02Race-day weather. The 10-year median low and high for race day, scored against the ideal marathon window of roughly 45 to 60 degrees. This is the factor finders ignore most and the one that decides the most PRs. Weather carries the highest weight in the composite (40%) because the data supports it: a flat course in reliably warm conditions scores worse than a rolling course that's reliably cool.
• 03Field quality and size. A deep, fast field gives you packs to run with on pace, which is worth real time. We use field size as the primary signal and apply a discount to small-field BQ rates, which inflate via self-selection.
• 04Published BQ rate. Included as a field-quality signal, weighted at just 10% and read skeptically. A high rate on a small, ordinary course does not dominate the score. The method is designed so that games-the-stat races score poorly despite their eye-catching percentages.

How the factors combine

The factors use a weighted geometric mean: Terrain^0.30 × Weather^0.40 × Field^0.20 × BQ^0.10. A geometric mean is multiplicative, not additive. A race can't buy its way to a top score on one axis alone: near-zero weather still collapses the composite even if terrain is perfect. This is exactly the profile that produces qualifiers in the real world, a course that gets all four roughly right.

What the data turned up

A few patterns held across the 126:

The famous fast races mostly earn it, but not for the reasons people think. California International Marathon scores second overall globally and first among US races. Its edge is the net downhill (−350 ft) plus reliably cool early-December weather plus a deep qualifier field, not flatness (its first half actually rolls). About 30% of its 2025 finishers ran a BQ. Indianapolis Monumental Marathon earns its score on true flatness, cold November air, and an open, fast field, with roughly 18 to 22% qualifying.

The biggest races aren't the highest BQ rates. Bank of America Chicago Marathon is one of the flattest courses in the world, but its published BQ rate sits around 16%, lower than several mid-size races, because its 52,000-plus field includes every kind of runner, not just qualifier-chasers. Flat and fast, but the rate reflects the crowd, not the course.

Weather quietly reshuffles the list. Several courses that look fast on paper drop once you weight their typical race-day temperature. A flat course that reliably runs warm is not a BQ-friendly course, however good the elevation chart looks. This is why weather carries the highest single weight in the composite.

Global top 15 (126 races scored):

London tops the global list on a combination of a massive fast field, ideal April weather, and a 25% open BQ rate, but its World Major lottery makes that rate a published fact, not an attainable path for most runners. For the US-focused view, the top 10 reorders around accessible races:

Some "BQ hotbeds" are gaming the stat. A handful of races post eye-catching BQ percentages that are almost entirely a field effect or a terrain effect, not evidence of a genuinely fast, open course:

• 01Light at the End of the Tunnel Marathon: 42% BQ rate, net −870 ft, 500 finishers. A railroad tunnel course with extreme descent and a field so small and self-selected that the rate is almost a census of fast runners who chose the race specifically for the descent.
• 02Tunnel Marathon: 39% BQ rate, net −1,000 ft, 852 finishers. Same pattern. The terrain produces fast times for runners trained for steep downhills; the field effect inflates the published rate further.
• 03REVEL Mt Charleston Marathon: 39% BQ rate, net −4,914 ft (nearly a mile of downhill). This is a genuinely different athletic event from a standard qualifier course. The quad damage from that descent is substantial, and the BAA certification doesn't change the fact that most runners who haven't specifically trained for steep descent will not run their best time here.

Where the full data lives

The point of scoring all 200 isn't the list, it's that you can act on it for your own race. The full scored catalog is the race directory, and the recommender turns the same four factors into a personalized top three based on your goal, experience, and how you handle hills.