In my previous article I sought on the journey to predict when the 2 hr marathon would be broken. What I had forgot to mention that it was already broken by Eluid Kipchoge unofficially on October 12th, 2019 in Vienna Austria. Since this was a sponsored event and not an official race that could be ratified by the IAAF or legal governing bodies of athletics, it does not count in the record books but it does prove that this 2 hr barrier can be broken. Lets take a quick look at this historical performance:
Looking at those splits back in 2019, Kipchoge was able to average 2:50 min/km or 4:33 min/mile. Also, you can’t help but notice that Kipchoge did run a slightly faster second half (10 seconds) meaning he negative split the challenge. This challenge had organized pacers, nutrition, course organization and plenty of other advantages that most record eligible races do not offer. Although record breaking, these advantages are not allowed in the official history books.
Going back to my last article, I laid out the groundwork in showing the men’s marathon world records over the course of history before building my model. This 26.2 mile/42.2km race has had its world record progression fall from 2 hrs and 55 minutes back in the 1900’s to now 2 hrs and 35 seconds in the 2000’s. The quest to break the 2 hr barrier has been a historical challenge:
As you look at the spread and the trend line, you can see that post 2000’s show “underfitting” of the data points alluding to the notion that the 2 hr barrier could be broken sooner than later. It is from this assumption that a form of regression modeling would be best for this type of prediction. In general terms, regression analysis is a statistical analysis aimed at explaining…