Unfinished Business in Boston: Projecting Non-Finishers' Finish Times

It's about time. Literally. It's about declaring projected finish times for the thousands of Boston Marathon runners who were not allowed to finish after the terrible tragedy suspended the race.

Three weeks have passed since the tragedy, and yet there is no resolution for those thousands who have every right to declare themselves Boston Finishers, given the circumstances of the day. Time for closure. Time to define your projected finish time. The BAA's resolution isn't ready yet, which is reasonable, given the myriad of other issues they face.

Until then, it's time to step up with projected finish times, and we've taken on the assignment (unofficially). We ran the numbers, slicing and dicing the data every which way (see notes on Methodology, below), and after hours and hours of detailed analysis, I'll get to the punchline.

Projecting Your Boston Marathon 2013 Finish Time

What might your finish time have been?  It may be as simple as this:
  • If you were stopped after 35k, your effective finish time may = 35k time multiplied by 1.23
  • If you had crossed the 40k, your effective finish time could =  40k time multiplied by 1.06
Is this 100% accurate? No. But I'd expect our projected results will be within a minute or two in a wide majority of the cases.

But: One exception will be the runners who went out too fast, too early, and had hit the wall, slowing to a crawl. See our Correlation Between First and Second Half Split Times to gauge if you fit in that group.

Will the BAA's resolution be different? It very well could be. We don't know what direction they will go. But one thing is clear: our analysis here is unofficial; BAA's will be the official resolution. 


Background: our perspective on this topic comes from several dimensions. We know the course intimately; 13 consecutive Boston finishes. We've been presenting in-depth coverage of the Boston Marathon for years, see www.RaceBoston.com. We've developed and posted dozens of analytics and diagnostics about Boston; see our Stats/Results/Analysis archive for more than 200 posts analyzing top marathon and triathlon events. We're not new to projecting finish times based on incomplete data; see many examples in our Race Statistics/Results Analysis archives. Most recently, we projected finish times of the new Ironman 70.3 St. George within 1% of actual.]

We noted the multipliers above were 'deceptively simple' because it took a great deal of work to confidently calculate the appropriate value. 

Our approach is illustrated in an example of our analysis of all Illinois finishers in the 50-54 age group. This included 53 men, 19 women. We did extensive analysis of this sampling of runners, tabulating their 5k splits from start to finish, comparing an contrasting splits from sub-3 hour finishers to 4+ hour runners, from Boston re-qualifiers to those who, well, didn't come close to requalifying. 

See data below, from drill-down detail for each finisher, up to average splits as a percent of total, which we used to calculate the 35k and 40k multipliers.

Male Finishers Aged 50-54 from Illinois

Female Finishers Aged 50-54 from Illinois

For more on our extensive Boston Marathon analysis, see our Complete Boston Marathon Coverage. For even more, see our archive of 200+ Race Stats/Results Analysis posts