Boston University: The Fastest Track in the NCAA
A Statistical Analysis of Men’s NCAA Track and Field Distance Results at Boston University
Introduction
Breaking 4 minutes in the mile was once thought to be impossible, but at a small track in Boston, Massachusetts, they are routine. Despite being over 20 years old, the best runners in the world still flock to the Boston University Track & Tennis Center to break records. For example, at the David Hemery Valentine Invite in 2023, an unprecedented 52 athletes broke 4 minutes in the mile, including 10 under 3:56. A few weeks prior, Northern Arizona athlete Drew Bosley broke the NCAA 3000m record with a time of 7:36.42, yet was overshadowed by Yared Neguse, a professional athlete who broke the American record with a time of 7:28.23. Galen Rupp also famously owns the 2 mile American record at 8:07.41, run in 2014, the same year that Ethiopia’s Yomif Kejelcha indoor mile world record of 3:47.01.
The times run at Boston University are undeniably impressive, but is this 20 year old track actually the fastest in the world? After all, hydraulic banked tracks pop up year after year in the US, but still athletes return to BU.
This study looks at Division 1 NCAA Men’s results run at Boston University between 2018 and 2023 and compares them to similarly banked tracks across the country. There are 3 main questions to be answered…
- Does the Boston University (BU) Indoor Track deliver faster times, on average, than other 200m banked indoor tracks in distance events?
- Are more top NCAA times being run at Boston University than expected?
- Is the percentage of top NCAA times being run at BU increasing as time goes on?
This third question is particularly interesting because it could give some insight into why BU is so fast. If the amount of top marks being run at BU is increasing over time, then perhaps runners only run fast at BU because they expect the track to be fast.
Summary Data
First, we gather some basic summary statistics for 4 events, the 800m, the mile, the 3000m, and the 5000m. These are the most widely run indoor distance track events, and the ones BU is notoriously fast at.
Does the Boston University (BU) Indoor Track deliver faster times, on average, than other 200m banked indoor tracks in distance events?
The average race at BU seems to be faster than the average at other 200m banked tracks in the NCAA. In fact, the difference in means for every event in every season (except for 2021, when BU did not host races) are statistically significant at both the 5% and 1% levels. Although only around 10% of 200m banked track races are run at BU every year, we still have significant evidence that, on average, Men’s NCAA distance races are faster at BU than other 200m banked indoor tracks.
Are more top NCAA times being run at Boston University than expected?
Significant evidence that Boston University is faster on average than other banked tracks is a strong start, but this could merely be the product of stronger runners traveling to BU than other banked tracks. Instead, we look to the highest caliber competitions, including national and conference championships, by limiting our analysis to the Top 100 marks run on banked tracks for the entire NCAA season.
We will look at the difference between actual and expected Top 100 marks. For example, in 2023, 334 total NCAA Division 1 Mile marks were set at BU, or 13.81% of all mile marks run on 200m banked tracks. Therefore, we would expect 13 or 14 of the Top 100 NCAA times to have been run at Boston University. If, as it was in reality, 53 Top 100 marks were set at BU, we would mark a difference of 39.19 between actual and expected top 100 marks.
The following tables plots the difference between actual and expected Top 100 results by event and year.
While the difference for some events, namely the 800, is fairly inconsequential, other events, like the 5000, have jarring results. For example, in 2023, 27.2% of 100 top marks were expected to be run at BU, but an incredible 77 marks were run in reality.
Using a chi squared test for goodness of fit, all differences in expected values for events greater than or equal to a mile in length are statistically significant at the 5% level. In addition, all but 1 event (3000m marks run during the 2018 season) are significant at the 1% level.
The 800, on the other hand, does not have strong evidence for extreme improvement. Only in 2020 was the goodness of fit test significant at the 1% level, and all other years were not significant at even the 5% level. While the average 800 race is faster overall, the very fastest collegiate races are equally fast at BU vs other banked tracks. The 800m is often considered a mix between sprinting and endurance, verses the other events in this study which are almost purely distance events. Looking more closely at 800m marks at BU and across the country may be an interesting area of farther study, but for now, we cannot conclude that BU is the best track in the NCAA to run the 800m.
Is the percentage of top NCAA times being run at BU increasing as time goes on?
To analyze whether the Top BU times are increasing over time, we analyze the correlation between difference and season
There is not strong evidence that BU track times are improving over time. Although most of the correlations are positive and linear, the R values are not particularly strong and none of the linear models are statistically significant at the 5% level. It is possible that there may be a relationship between time and faster results if more years are included, which is an area of farther study. However, we cannot draw any conclusions from the data presented here.
If anything, this is even more impressive evidence of BU being the fastest track in the country. It was theorized that perhaps the reason for BU being fast was because athletes expected the track to be fast, a self fulfilling prophecy in a way. However, despite world class marks being set at BU every year, reinforcing it’s claim as a fast track, the quantity of those top times is actually not increasing over time, indicating that the placebo effect may not be as strong as initially thought.
Conclusion
There is convincing evidence that the Boston University track runs faster in distance races than other banked tracks in NCAA. Both average and Top 100 times are faster than other banked tracks, and there is not a notable time relationship that would indicate a placebo effect.
There are many areas of farther study, most notably looking at women’s results. BU could also be compared to different types of tracks, especially 300m flat indoor tracks or even outdoor tracks. Also, it is unclear if sprinting events, particularly the 200m, 400m, or arguably the 800m, have any significant difference at BU compared to other notably fast sprinting tracks, such as the University of Arkansas track.
Appendix
